Amiga Tools 3/Amiga Tools 3.iso/grafik/raytracing/imagine/tips/arc14.txt

Text File | 1993-03-12 | 134KB | 3,264 lines

IMAGINE archive: collected off of [e-mail]

ARCHIVE XIV
Oct. 17 '91 - Nov. 8 '91

If you have questions or problems with this file, email Marvin Landis
at [e-mail]

note: each message seperated by a '##'

&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

Subject: Re: brush maps
Date: Thu, 17 Oct 91 9:46:59 EDT
From: [e-mail] (John J Humpal)

> locally to the top.  Is the move supposed to be in local mode too?  I am

    Yes.  If you're doing a WrapX and/or WrapZ, make sure you have
the brush axis in the middle of the sphere.  That is, the axis should
extend from the bottom to the top of the sphere, and it should pierce
right thru the middle of the object.  Also, make sure the y-axis size
is 1.
    IFF wraps work fine in scanline.  It sounds like your problem
is related to axis placement (the infamous "digital bounce" problem, whatever
that means).

-John

John J. Humpal -- [e-mail] -- short .sig, std. disclaimer


##

Subject: Converting grey-scale textures
Date: Fri, 18 Oct 91 11:25:42 -0600
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

    On converting those grey-scale textures what was the procedure again?
Load them up with Sculpt RAW and convert them to 8-bit grey-scale? I tried this
out a pfft! You can only see whats there by turining up the contrast & it just
looked like repeative junk to me (I tried image06 & image07 off tape3), I must
be goofing this up somehow.

    Chris
    [e-mail]


##

Subject: bug?!?!
Date: Fri, 18 Oct 91 13:42:49 EST
From: [e-mail] (Ramin Taraz)

Somthing rather funny happend to me  when I rendered a picture
I had a table with a lamp hanging from the top.
ambient light was around.
I rendered it and every thing was as I expected, the table was shiny and
around it was dark.  but the problem was that under the table was bright too
which should have been.  because the light was above the table and the table
should have blocked all the light.  When I said it was bright, I didn't mean
that there was light, I meant that there was actually a bright point like
there was no table at all.
I hope I explained the situation well,
writing was never my favorit subject! :)


rtaraz

##

Subject: real world lighting ?
Date: Fri, 18 Oct 91 16:26:51 -0700
From: [e-mail]

mark thompson made a comment a while back about being unable to model "real
world optics" (i think it was in reference to the cool CD he created in
Lightwave...)

my question is...do the values in the light intensity have
ANY correlation to the real world ? i.e. lumens, footcandles, etc....
i am an architect trying to use the amiga to model rooms, etc. and would
like to be somewhat accurate with my lighting if it is at all possible.
There is a public domain program called "radiance" which does real world
light models, in the unix world...

and why can't i use florescent (long tube) as a light source ? and why doesn't
imagine support a true sunlike source like Lightwave ? inquiring architects
need to know ! :)

i am currently teaching arch. design using the Mac, at UC Berkeley. The studio
has 12 Mac II's, plotter, and NTX,at least partly courtesy Apple computer.
Of course, the arch. department also has SGI's, Suns,and DECs, so my class
is considered "the low end" :) no amigas yet, although i may demo toaster...
if the modeling packages on the amiga could support more realistic modeling
tools, it could fit right in...

kevin

##

Subject: Light Ray spikes.
Date: Sat, 19 Oct 91 12:14:33 MDT
From: [e-mail] (Steven Lee Webb)

    Has anyone ever rendered anything on Imagine, and gotten some weird
    spikes coming from the light source?  On mine, I rendered the pumpkin
    that was on hubcap, with a face carved out of it, and I got some of
    those black "Light Rays" coming out of the eye sockets, but not the
    mouth.  (To tell you the truth, it looks really cool, but it's not
    really what I had in mind when I made the scene. Oh well.)

    -- Steven.
--
     _________  _________  _________  ___   ___  _________  ____   ___
    /  ______/\/__   ___/\/  ______/\/  /\ /  /\/  ______/\/     \/  /\
   /  /\_____\/\_/  /\__\/  /\_____\/  / //  / /  /\_____\/  /|  /  / /
  /_____   /\   /  / /  /  ___/\   /  / //  / /  ___/\   /  / |    / /
 _\____/  / /  /  / /  /  /\__\/  /  /_//  / /  /\__\/  /  / /|   / /
/________/ /  /__/ /  /________/\/________/ /________/\/__/ / |__/ /
\________\/   \__\/   \________\/\________\/\________\/\__\/   \_\/
         ___   ___  __ ______  _ _______   ________
        /  /\ /  /\/  ______/\/  ___   /\/  ___   /\          Guess
       /  /_//  / /  /\_____\/  /\_/  / /  /\_/  / /       /\_  Who
      /  /  /  / /   __/\   /  ___   / /  ___   / /        \\ /\ Jr
     /  /  /  / /  /\__\/  /  /\_/  / /  /\_/  / /         / / \
    /________/ /________/\/________/ /________/ /         // \\  _
    \________\/\________\/\________\/\________\/         //\\   (_)
          ([e-mail])               ~~~~~~~~~~~~

##

Subject: Re: Light Ray spikes.
Date: Sun, 20 Oct 91 19:22:57 CDT
From: [e-mail] (Randy Carris)

Steven,
I got the same thing on a project I was working on a while back
involving an eye.  (refering to "weird spikes" or "Light Rays")  I still
don't know what caused it but it made my trace times on an 030 to
be about 16-20 hours per frame, which really bites.  I also got it to
a much lesser extent on some extruded letters.  The project was
uploaded to hubcap as ovp.imp.lzh in the projects directory if
anyone wants to look in detail at it.  The "artifacts" only appear in
trace and not scanline renders.  I would appreciate any input on
what causes it.
 Thanks,
Randy Carris

##

Subject: Well, where are they?
Date: Sun, 20 Oct 91 22:17:31 CDT
From: [e-mail] (Steve W Davis)

I accidently mis-named a file in a post to alt.binaries.pictures.d and
alt.binaries.pictures.misc.  I have thus far received about 20 mail messages
asking why the file I said on hubcap.clemson.edu isn't there.

WOULD SOMEONE PLEASE DOWNLOAD augjpeg.lzh FROM STUDIO AMIGA BBS AND
PUT IT ON THE HUBCAP?

I would do it myself, but I can't get a a file transfer to work between
my Amiga and the K-State network.  (I can download, but not upload).

As far as JPEG files on hubcap.... I suggest the following format.
filename.jpe
filename.readme

As soon as I get file transfers working on my end, I have some guilt-ware
textures (736x480x24 bit) to put on HUBCAP.

I can also set aside a directory on my BBS for these files if someone
is interested.  At the moment, our modem is a v.32/v.42, so HST people
would be out of luck...

Stratocaster

-------------------------------------+---------------------------------
Steve Davis     [e-mail]    | "Sir, I protest:  I am NOT a
                                     |    merry man!"   -- Worf
-------------------------------------+---------------------------------

##

Subject: BIG ANIM
Date: Sun, 20 Oct 91 20:40:10 CDT
From: [e-mail] (Michael Linton)

I rendered an animation the other day, and to my surprise it was 7 megs
rendered!  Since I only have 6 megs in my A3000, I don't have enough ram to
play it.  Does anyone have any ideas as to how I can get it to use the
hard disk to load off constantly instead of RAM?

Thanks.


--- (Michael Linton) a user of sys6626, running waffle 1.64
E-mail: [e-mail]
system 6626: 63 point west drive, winnipeg manitoba canada R3T 5G8

##

Subject: Re: real world lighting ?
Date: Sun, 20 Oct 91 20:46:36 CST
From: telepro![e-mail] (James Hastings-Trew)

In a message dated Fri 18 Oct 91 19:47, [e-mail] wrote:

 > and why can't i use florescent (long tube) as a light source ? and why
 > doesn't
 > imagine support a true sunlike source like Lightwave ? inquiring
 > architects need to know ! :)

I don't know what you mean by a "true" sunlike source, but you can get very
close by creating a lamp that does not diminish intensity with distance, and
setting it VERY far from your scene. The rays will not be visibly coming from
any particular point, just a direction.

As for a florescent lamp source, you gotta do a bit of fiddling. First,
create the tube as a "bright" object. Second, arrange a series of axies along
the top, bottom, or side of the tube (where you want the majority of the
light to eminate from), and set the axies to be lights, diminish intensity,
fairly low intensity, cast shadows. Group these to your tube object. It will
look somewhat like this:

  z    z    z    z    z    z    z    z
  |    |    |    |    |    |    |    |
  +--x +--x +--x +--x +--x +--x +--x +--x   (these are the lamp axies)
 -----------------------------------------
[                                         ] (this is the bright tube object)
 -----------------------------------------

Reasons for doing this: if you set an object to be a "light source", only
it's axis is actually emitting light, not the surface of the object. (If you
have the axis inside the object, and set it to cast shadows, it will be
invisible and not light anything!) The bright tube is just a "stand in"
actor, to be the "visible" part of the light source. The multiple axies will
cast overlapping shadows, simulating the diffuse shadow look of the
florescent tube. The more axies you have, the closer the simulation (and the
more lengthy the rendering time.)

Problems with this method: wasteful of rendering time, and the tube will cast
it's own shadow. This is usually not a problem, as floresecnt lights are
usually mounted very close to a ceiling, or in a fixture, so you can put the
lamp axies facing the direction that the light needs to be cast in, and the
tube should hide it's own shadow.

If you need to simulate a florescent lighting panel, simply use a plane, set
it to "bright", and place a grid of lamp axies just under the panel, to cast
the same kind of diffuse light.


-- Via DLG Pro v0.975b

--- James Hastings-Trew    telepro![e-mail] ---
->  If everyone in China jumped at the same time <-  subliminalmessagewith
->      would they bring back TWIN PEAKS?        <-  subtlebutdeviousplan.

##

Subject: Re: BIG ANIM
Date: Mon, 21 Oct 91 09:36:20 EDT
From: Mark Thompson <[e-mail]>

> Does anyone have any ideas as to how I can get it to use the
> hard disk to load off constantly instead of RAM?

This has been discussed on comp.sys.amiga.graphics to some extent in the
past couple weeks. Basically there are a couple hard disk based animation
players but they have all been included with comercial products and are
not freely redistributable. There is one included with Director 2.0,
another with Digimate III I believe (the animation companion to Digipaint 3),
and there was one that worked with TASS which was released on a disk with
a magazine ages ago (which one I forget). Actually, I think the Digimate III
and TASS HD animators are one and the same. Any way, hope this helps.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: lighting
Date: Mon, 21 Oct 91 12:14:51 EDT
From: [e-mail]

Kevin asks:

>and why can't i use florescent (long tube) as a light source ? and why doesn't
>imagine support a true sunlike source like Lightwave ? inquiring architects
>need to know ! :)


I don't know what relation light intensity has to "real world" units. However,
the lights do allow you to set them up in a "sunlike" manner. In fact,
this method is the default for lights. The intensity is constant, even
at a distant point. To become truly sunlike, you should probably use a
spotlight with a very large radius to get parallel light rays instead of
a point source that would have divergent rays.

You can't have a "tube"  distributed light source. This is annoying, but think
of the amount of effort it would take! (Certainly possible, though). I have
always approximated distributed sources (like a flourecent light) by
a set of point lights (in this case maybe 4 or 5) of low intensity. The
shadows aren't exact butt the low intensity blurs them together quite
nicely. You can make a "bright" tube to actually represent your source.

If you are really looking for accurate lighting, you might check out
3D Professional 2.0 when it comes out. They claim it will have full
radiosity support (!!!) but I'll beleive it when I see it.

-Steve

##

Subject: Thanks Loads
Date: Mon, 21 Oct 91 12:48:29 CDT
From: [e-mail] (Steve W Davis)

Thank you very much to everyone who falls in one of the following
categories:

1.  Backed up my statements on 24-bit graphic files in alt.binaries.pictures.*
2.  Responded to my queries about available JPEG software.
3.  Found sources for JPEG test files.
4.  Has, is, or is considering putting JPEG graphic files/software on
    hubcap.

It now looks as though we can use alt.binaries.pictures.misc to send
UUENCODEd JPEG files to one another, without risk of being flamed to
death by GIF fanatics.

I have some guilt-ware textures/background images to distribute,
as soon as I can get some files from my Amiga to this campus UNIX
machine.  I am also setting up a file area for JPEG graphics files
on the BBS listed below.


Stratocaster

----------------------------------------------------------------------
Internet: [e-mail]      | Steve Davis
 Fidonet: Steve Davis @ 1:295/3  |
     BBS: The Boarding House BBS | Video Animator
          9600 Baud, v.32/v.42   |
          (America) 913-827-0744 |  Have you hugged your Amiga today?
----------------------------------------------------------------------

##

Subject: BigAnim
Date: Mon, 21 Oct 91 13:27 MST
From: "Vax Headroom (Dave E Martin)" <[e-mail]>

If you have FTP access, there is a file:

/incoming/amiga/BuildAnim.lzh

on ab20.larc.nasa.gov.  It contains biganim, which plays anims off the disk.
I havent used it enough to properly determine its speed yet.  Also included
is buildanim which does not seem to work properly under 2.0.

The whole archive is only about 4600 bytes so it could probably be uuencoded
and sent to people without ftp.

----
Between the streets of Dallas and the Beaches of Miami...

[e-mail]
[e-mail]
     (NOT .mit.edu)

##

Subject: Some suggestions for imagine (2.1?)
Date: Mon, 21 Oct 91 14:31 MST
From: "Vax Headroom (Dave E Martin)" <[e-mail]>

Unless these are already in 2.0, for 2.1 I would like to suggest the following:

The priority be a parameter in the configuration file and/or a menu item.
It would be even nicer if I could specify that the program run at -1 while
rendering and at 1 at all other times.

The following are suggestions from a friend of mine and are his words:

I suggested that the index of refraction be setable for each wavelength of
light.  This will enable prisms and other interesting real-world transparent
solids (like diamonds) to be better ray-traced.  It would be nice if it would
interpolate between given wavelengths (such as R, G, and B), but be able to
produce a prismatic rainbow.

DKB-trace has a greatly variable texture generator.  This generates a "three
dimensional" terrain which is then normalized to a 0.0 to 1.0 range.  Your are
allowed to assign RGBA (Red-Green-Blue-Alpha) values to a height, and DKB-trace
uses linear interpolation to get the value in between.  You can also suddenly
change the RGBA at a height range.  There is no arbitrary limit set on the
number of changes that can be made.  This method of creating textures allows for
some very natural-looking clouds (with silver linings), and more natural-looking
marble and wood than Imagine, as well as some other very interesting textures.

##

Subject: prismatic refraction
Date: Mon, 21 Oct 91 17:37:31 GMT-0500
From: Scott Matthew Krehbiel <[e-mail]>

In response to the Vax Headroom ( Dave E. Martin )'s suggestion for
real refraction, I have a question: When you set the Red Green and Blue
values for an object, does Imagine know what frequency light that object
reflects?  I would suggest that it doesn't.  My guess would be that most
ray-tracing programs treat Red, Green, and Blue separately, just as our
eyes do.  If you had true color refraction in say a clear glass ball
that was bending orange light, my guess would be that you'd get separate
red and green beams exiting the ball.

It's interesting to think that our computer screens may not be emitting
orange or purple or any other combination-of-pixels color, but it
just seems that way to our eyes, because we see RGB.  If your monitor
was emitting an orange glow, and was lighting a piece of white paper,
how would a spectrometer see that paper?  Again I'm guessing, but I
would tend to think that it would see both red and green, but not the
orange frequency.

Think about this:  if imagine had to refract white light, what frequency
would that light be?  ( Perception is such a strange beast )

Scott Krehbiel

##

Subject: Hubcap, oh Hubcap.
Date: Mon, 21 Oct 91 18:31:57 EDT
From: [e-mail] (Alan Price)

I worked hard all weekend to finish a render of a new room I made.
It has columns and arches and paintings and a pool and all kinds of
stuff....

I wanted to ftp it to Hubcap to show off, but Hubcap has temporarily
denied write access to anonymous ftp users (sniff sniff).

Maybe later!

AP.

##

Subject: PlayIanm
Date: Mon, 21 Oct 91 17:14:31 CDT
From: [e-mail] (Michael Linton)

Is the progam PlayIanm freely distributable?



--- (Michael Linton) a user of sys6626, running waffle 1.64
E-mail: [e-mail]
system 6626: 63 point west drive, winnipeg manitoba canada R3T 5G8

##

Subject: Textures
Date: Mon, 21 Oct 91 22:55:20 -0600
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

    I've converted over Images6-12 on tape3. Only about 3 of these really
look to be of any use image6- a tile like image, image7- snakeskin or beads?,
image10- a marble, image11- a wood, maybe a brick. The rest appear as random
pixels, unless of course thats what you want....
    I propose that short descriptions are written for each texture and put i
in one text file on hubcap. Sure make one more facet of life easier.
    I'll upload the ones that I have to hubcap, once I can figure out the
school's machine on recieving files. Shall we have a seperate sub-dir under
TEXTURES for them?

    Chris
    [e-mail]

##

Subject: Re: real world lighting ?
Date: Tue, 22 Oct 91 11:56:58 +0200
From: [e-mail] (Helge Egelund Rasmussen)

kevin (lands.ced.berkeley.edu!kevink) wrote:

> my question is...do the values in the light intensity have
> ANY correlation to the real world ? i.e. lumens, footcandles, etc....
> i am an architect trying to use the amiga to model rooms, etc. and would
> like to be somewhat accurate with my lighting if it is at all possible.
> There is a public domain program called "radiance" which does real world
> light models, in the unix world...

I don't think that Imagine is able to do what you want, so you probably
should take a closer look at Radiance.

I've been porting Radiance to the Amiga. Furthermore, I've written an Imagine
to Radiance object converter.

When the port is finished (real soon now), the necessary patches will be placed
on hobbes.lbl.gov (in pub/ports/amiga).

Helge
---
Helge E. Rasmussen  .  PHONE + 45 36 72 33 00  .  E-mail:  [e-mail]
Compel A/S          .  FAX   + 45 36 72 43 00  .
Copenhagen, Denmark

##

Subject: Re: prismatic refraction
Date: Tue, 22 Oct 91 15:29:29 EDT
From: Mark Thompson <[e-mail]>

> I suggested that the index of refraction be setable for each wavelength of
> light.  This will enable prisms and other interesting real-world transparent
> solids (like diamonds) to be better ray-traced.  It would be nice if it would
> interpolate between given wavelengths (such as R, G, and B), but be able to
> produce a prismatic rainbow.

The problem with this is that ray-tracers and other "light simulators"
are programmed and run in a digital domain. Light is an analog phenomena
with a continous range of wavelengths across the entire spectrum (visible
and invisible). It is not really a simple matter of interpolation from
say red to green. What you really need is to trace MANY rays per pixel
(for every one ray you would normally trace) in an effort to simulate
the full spectrum. There are a few ways this process could be sped up.
First of all, only shoot ray spectrums when a single ray intersects a
refractive medium. Second, coherence optimizations can be made since each
sequential spectrum ray will deviate only slightly from its previously
traced 'sibling' ray. It will be an extremely compute intensive rendering.
From a forward trace perspective, this all sounds somewhat intuitive but
I question how effectively such techniques could be implemented in a
backward tracer (the only kind of tracer any sane person would implement).
In any event, it is most likely WAY beyond the realm Imagine to handle
this kind of stuff.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re: prismatic refraction
Date: Tue, 22 Oct 91 16:07:14 EDT
From: Mark Thompson <[e-mail]>

Scott Matthew Krehbiel writes:
> My guess would be that most
> ray-tracing programs treat Red, Green, and Blue separately, just as our
> eyes do.  If you had true color refraction in say a clear glass ball
> that was bending orange light, my guess would be that you'd get separate
> red and green beams exiting the ball.

This is not the case. Most if not all ray-tracers (Imagine included) consider
all light wavelengths to be bunched together in a SINGLE ray representing
the entire spectrum of that light.

> It's interesting to think that our computer screens may not be emitting
> orange or purple or any other combination-of-pixels color, but it
> just seems that way to our eyes, because we see RGB.  If your monitor
> was emitting an orange glow, and was lighting a piece of white paper,
> how would a spectrometer see that paper?  Again I'm guessing, but I
> would tend to think that it would see both red and green, but not the
> orange frequency.

It is important to remember that the R,G, and B phosphers of a monitor
each produce a spectrum of wavelengths and not monochromatic outputs.
The spectrometer you speak of would show a range of wavelengths centered
around orange and not any one or two frequencies. Also, the RGB model is
one of conveniance and in NO way represents the way light actually works.
We do not see RGB.

> Think about this:  if imagine had to refract white light, what frequency
> would that light be?  ( Perception is such a strange beast )

Imagine would refract white but real life would produce a spread spectrum.
White has no frequency (like white noise in the audio domain) and Imagine
does not understand frequencies anyway.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Pixel 3D 2.0
Date: Tue, 22 Oct 91 19:13:22 EDT
From: [e-mail]

A mini-review of a very useful program for you Imagamaniacs.

---------

At the Expo in Oakland I picked up a copy of Pixel 3D 2.0. This
program is a "third party" 3D support program for use by owners of
almost any Amiga 3D software. Pixel 3D has two major functions; it can
load brush maps to produce faced and extruded objects, and it can also
load in objects (in many formats) and convert them to different
formats (like Imagine's).

Pixel's brushmap manipulation abilities are absolutely brilliant. In
Imagine, you can use the built in brushmap converter to make objects
from brush outlines, but Imagine's ability is crude as compared to the
control and options that Pixel offers. It can load standard 1,2, 3, or
4 bitplane pictures (not IFF24 or HAM, though!) and use the outline of
the picture to form an extruded 3D object. This is a useful ability,
especially since you can type out a logo in a fancy font in DPaint
then convert it to a 3D object for rendering.

This basic extrusion is just the simplest manipulation Pixel can
handle, though. Remember a few months ago when we determined how to
make beveled text? It was certainly possible, but it involved point
editing every letter to make a bevel that was able to "skin" to the
extruded form. One of Pixel's extrude options is to automagically add
a bevel to its extrusions!  This bevel is a true bevel, not a wimpy
cone-like facing, so hollow letters and concave shapes are handled
correctly. The depth and angle of the bevel cut is completely user
definable, so subtle additions can be made to letters, or titles can
be given an aggressive edge.

An aside to give you guys a quick hint. If you want to Phong shade an
object with a beveled extrusion in Imagine 1.1, you have to perform
some tricks to get the bevel to be sharp and not just a blurry edge.
You want the front face to be flat (unPhonged), the bevel to be
Phonged around its edges, but NOT Phonged where it joins the extruded
sides of the object. You can't pick and choose independent edges to
Phong shade in 1.1, so how can you do this?

The answer is to make three objects that are all grouped together and
independently Phonged.  You want to take your beveled object, copy it
into the clipboard (right-Amiga-C) and then delete all of the points
that aren't in the very front face.  Edit the attributes and turn
Phong OFF. Save this object.. it's just the 2D front "plate".
Right-Amiga-P will paste a copy of your original object back. Delete
the front layer of points, which will leave you with just the sides of
the object. Make sure Phong is ON for this version.  Paste a final
copy of the original object.  Delete the BACK set of points, leaving
you with the front face and beveled portions. Enter "pick points"
mode, then "hide points" and set the select method to "drag box". From
a side or top view, select the back set of points to hide them,
leaving only the front plate visible. You might want to hit "redraw".
Enter "select faces" mode, select the entire object, then use
right-amiga-d to delete the faces. (The edges will still be there but
won't render). Go back to select object mode, make sure that Phong is
turned on for this object. Finally, move all three of the objects
together to their proper places, and group them. Tahdah! This group
will now render with the proper Phong shading.

Back to Pixel 3D.

Beveling is a very professional touch, especially for lettering, but
Pixel has many more abilities. Multiple colors can be extruded to
different depths, allowing nice background effects.  The brushmap
outlines can also be spun as they extrude, so snake-like trails can be
left behind letters and shapes. And, of course, they can
simultaneously be beveled.

Most of these effects are easily reproducible in Imagine, but it's
convenient to have them all included in one package. Also,  people
who are using other renderers might not even have some of the tools we
take for granted.

One trick that really isn't available in Imagine is the ability to load
an image and output a grid of (connected) points with each point having
an altitude that corresponds to the average intensity of the pixels in the
image beneath it. I had written a routine using Glenn Lewis' TTDDD to
perform this very feat.. I wanted to make an easy way to output a true
height field from a C program to test my ocean wave stuff long ago.

What can you do with this intensity->altitude conversion? Quite a bit!
I took a greyscale image of myself and ran it through Pixel and got a
very interesting object, indeed! When I also applied a color brushmap
to it in Imagine, it gave my portrait a whole new dimension. [dodges
rotten tomatoes.] At the show, Axiom was selling T-shirts of their
logo in a raised 3-D grid style made with this technique.

Pixel's strong point is definitely converting brushmaps into 3D
objects.  It has excellent control over the way they can be
manipulated, and the author, Scott Thede, deserves a lot of credit for
making the program so robust and easy to use. One point I forgot to
mention is that the brushmaps, once converted to 3D objects, are
actually previewed on the screen with user controllable viewing angles
and size control, and options of wire frame or a solid view. The
preview is in color, so face attributes can be easily determined. This
display isn't a renderer, it's just a flat shaded preview, but it is
superbly adequate for previewing objects you are manipulating before
you finally save them.

Pixel can obviously save its objects in Imagine format, but it
actually supports a wide variety of renders. It can save in Sculpt,
Imagine, Turbo Silver, Lightwave 3D, VideoScape, 3D Pro, or DXF
format. For triangle based modelers like Imagine, all polygons are
saved as triangles, and polygon based modelers like Lightwave are
saved as optimized polygons.

The second major ability of Pixel is object conversion. It can
obviously write all of these different formats, and it can read them
as well (except DXF). Files can be selected from disk and the object
format detected automatically.  The object will show up in the preview
window, and some useful operations (such as duplicate point removal)
can be performed. The model can then be saved in any of the other
formats, giving Pixel general 3D object conversion capabilities.

I had no problems loading in VERY large objects (a 250K Boeing 767
model) and saving them to other formats. However, there is a MAJOR
flaw with the object conversion aspect of Pixel, as it does not
recognize hierarchies (grouped objects). If you load a grouped object,
only the parent of the group is loaded; the children are not kept at
all.  Nearly all complex models in Imagine are grouped objects, not
single masses of fused polygons, so importing these complex objects is
impossible. The only workaround is to join all of the objects into one
master object, losing all of the hierarchical information, then
loading the single macro-object into Pixel. This method works very
well, but obviously a lot of useful (critical!) information is lost.

Even though this makes the conversion abilities of Pixel 3D 2.0
somewhat stunted, you CAN convert any object by joining it first and
trying to sort out the organization in your destination renderer.
Another option would be to save each child as a separate object then
re-group them in the new modeler.

Obviously advanced information such as textures and brush maps are not
preserved for any object.

Note this opinion is slanted to how useful Pixel 3D is with Imagine.
Most other renderers lack Imagine's built in brushmap converter and
might not have hierarchical objects; in these cases Pixel 3D would
serve as a CRUCIAL piece of software, especially for the simple task
of making logos from brushmaps.

A last minute addition to Pixel 3D was a complete AREXX port. Scripts
can now be written to automate brushmap or object conversion, and some
examples are given in a file on the disk.

Overall, the program and its abilities really are pretty sleek.  The
object conversion limitation is not insurmountable, and for people who
do not use groups often this limitation won't even exist. The brushmap
extrusion abilities of Pixel 3D 2.0 are more than enough on their own
to make this program a useful utility for Imagine users.  The list
price of Pixel 3D is a fairly steep $100; but if you are constantly
making logos and other bitmap-based objects, this program is well
worth the price. For the casual (or poor!) user who has the patience
to use Imagine's built in bitmap converter, the price may be a bit
high for the abilities it offers.

An upgrade for current owners of Pixel 3D exists, though I do not know
the price.

Axiom Software
1221 E Center St. SE
Rochester MN 55904
[phone]

-Steve

--------------------------------------------------------------------------
Steve Worley                                       [e-mail]
--------------------------------------------------------------------------

This text is Copyright 1991 by Steven Worley but may be redistributed in
an unedited form on any media.

##

Subject: PSU Update
Date:    Wed, 23 Oct 91 09:33 EDT
From: "Marc Rifkin" <[e-mail]>

Thank you increduluously to all who sent in replies to my Amiga SOS call!!

However, the battle has just begun...
Just as I was starting to turn heads, make waves, etc.  Penn State pulled
the demo Amigas that I had previously convinced them to show in their
student System Evaluation Lab.  Then I discovered that orders came down
to departments that had Amigas (two were loaned by Commodore) to get rid
of them.  The demo Amigas have been replaced by Apples.


This is intolerable. (Further sentiments implied.)

..Still alive and kicking,
Marc Rifkin
Integrative Technologies Lab, Penn State University
5F Mitchell Building, University Park, PA 16802
814-863-8062 or for you normal people, [e-mail]
"Say, that's a nice bike!"

##

Subject: Textures
Date: Wed, 23 Oct 91 12:58:50 CDT
From: [e-mail] (Steve E Tietze )

I've been recently messing around with the textures in Imagine... I was wondering if anyone has any good settings for the wood,camo,brick... I can't see to
get the right affect I want..

Also has anyone tried Map Master yet?

Steve Tietze
[e-mail]

##

Subject: Re: Pixel 3D 2.0
Date: Wed, 23 Oct 1991 17:24:21 GMT
From: [e-mail] (Stephen Menzies)

[e-mail] writes:

>A mini-review of a very useful program for you Imagamaniacs.

>---------
> [much good info deleted]

>-Steve

>--------------------------------------------------------------------------
>Steve Worley                                       [e-mail]
>--------------------------------------------------------------------------

>This text is Copyright 1991 by Steven Worley but may be redistributed in
>an unedited form on any media.

Thanks for the review and the time you take to do things like this, Steve.
One question though: does the bevelling take place at the front *and* the
back of the object ?

-stephen
--
Stephen Menzies
#Internet: [e-mail]
#Fidonet : Stephen Menzies @ 1:167/265

##

Subject: Tie Fighter
Date: Thu, 24 Oct 91 00:04:30 PDT
From: [e-mail] (Rob Borsari)

I just put Darth Vaders Tie fighter in the uploads dir on Hubcap
(Hubcap.clemson.edu)  I hope it gets moved to the right place.
I have no idea who to tell ;)  Let me know what you think.  It is very
simple but (i think) clearly a tie fighter. -R-

[e-mail]  Rob Borsari  "Bourne to be Wild"

##

Subject: Re: Textures
Date: Thu, 24 Oct 91 00:34:44 CDT
From: [e-mail] (Michael Linton)

[e-mail] (Steve E Tietze ) writes:

> I've been recently messing around with the textures in Imagine... I was wonde
> get the right affect I want..
>
> Also has anyone tried Map Master yet?
>
> Steve Tietze
> [e-mail]
>
>
If you want a good wood texture, try rotating the texture axis at different
angles.  Also, try wrapping more than one wood texture on the same object with
slightly different settings.


--- (Michael Linton) a user of sys6626, running waffle 1.64
E-mail: [e-mail]
system 6626: 63 point west drive, winnipeg manitoba canada R3T 5G8

##

Subject: Re: Textures
Date: Thu, 24 Oct 91 04:39:09 -0600
From: Kiernan Holland <[e-mail]>

SOrry MIKE, didn't mean to bother you.

I'm doing a post to this listserv for the first time.

Lots of great objects on HUBCAP, love em.

Does anyone want to compile that JPEG compression program and
upload it to AB20 or something. I'm still in PASCAL so it'll
be two semesters before I learn how to compile that file.

I've got a 68030 based AMiga 3000 and a 68881.

I'm uploading a 514K file full of DMCS files to AB20 later on.
I configured them to work on the MT-32, for those MIDI freaks.

That had nothing to do with graphics, sorry, I still think I'm
in USENET.

##

Subject: yooo
Date: Thu, 24 Oct 91 04:43:32 -0600
From: Kiernan Holland <[e-mail]>


I've been doing a lot of stills with Turbo SIlver 3.0,
I was wondering if any body here could tell me how to
make a set of objects move along paths for an animation.

Actually just how to do animations.

I'm probably going to upload some of my pictures later on.
I have 1meg and a 50 meg HD, this is hell.

;-)

Later

##

Subject: Radiance
Date: Thu, 24 Oct 91 12:57:25 +0200
From: [e-mail] (Helge Egelund Rasmussen)

Several people has asked me about Radiance, so here is a short description
taken from RTNews.


RADIANCE was written as a lighting simulation more than a renderer, so it does
not allow some of the tricks that are permitted in other ray tracing programs.
(For example, all light sources fall off as 1/r^2.)  It has some of the nicer
features of advanced renderers, such as textures and bump maps (I've always
argued for calling them patterns and textures, respectively), octree spatial
subdivision, several surface primitives and hierarchical instancing.  It's
main distinction, however, is its ability to calculate diffuse interreflection
like a radiosity calculation.  (See the article by Ward, Rubinstein and Clear
in the 1988 Siggraph proceedings.)

Summary Description of RADIANCE:

    Lighting calculation and image synthesis using advanced ray tracing.
    Scenes are built from polygons, cones, and spheres made of plastic,
    metal, and glass with optional patterns and textures.

Detailed Description:

    RADIANCE was developed as a research tool for predicting the
    distribution of visible radiation in illuminated spaces.  It takes as
    input a three-dimensional geometric model of the physical environment,
    and produces a map of spectral radiance values in a color image.  The
    technique of ray tracing follows light backwards from the image plane
    to the source(s).  Because it can produce realistic images from a
    simple description, RADIANCE has a wide range of applications in
    graphic arts, lighting design, engineering and architecture.

    The reflectance model accurately predicts both diffuse and specular
    interactions, making it a complete simulation applicable to the design
    of unusual electric and day lighting systems.  Scenes are described by
    boundary representations with polygons, spheres and cones.  Materials
    include plastic, metal, glass, and light.  Textures and patterns can
    be described as functions or data.  Additional programs (generators)
    produce descriptions of compound objects, and allow regular
    transformations and hierarchical scene construction.  A 3D editor is
    being developed.

    The software package includes some image processing software and
    display programs for X, SunView, and the AED512, and comes with
    converters to Sun rasterfile and Targa formats.  Code is provided for
    writing other drivers, and the list is expected to grow.

Interface Description:

    The software is well integrated with the UNIX environment.  Many of
    the programs function as filters, with a reasonable degree of
    modularity.  An interactive program provides quick views of the scene,
    useful for debugging and view determination.  Scenes are described in
    a simple ascii format that is easy to edit and program.  Generators
    are provided for boxes, worms, surfaces of revolution, prisms, and
    functional surfaces (eg.  bicubic patches).  A small library of
    patterns and textures is included.  In general, the software is
    sensible but not mouse-based.

Overall Goals of Developer:

    The primary goal of the program is the accurate simulation of light in
    architectural spaces.  Secondary goals are image synthesis and
    geometric modeling.  Efficiency is an important concern in any ray
    tracing method.


To pick up release 1.4 from anonymous ftp, connect to hobbes.lbl.gov
(128.3.12.38) with ftp using the "anonymous" account and enter your e-mail
address as the password.  Everything is in the directory pub, and the main
distribution is called "Radiance1R4.tar.Z".  This file is about 3.5 Megabytes,
so please do your transfers in binary mode the first time!  Also, you will
probably experience less network traffic in the morning, when most computer
scientists are asleep.

Lighting Systems Research Group
[e-mail]

Here's a few comments about the Amiga Port:
Most of the Radiance package has been ported to AmigaDos 1.3 (SAS/C), another
person here in Denmark is working on a 2.0 port, but it might first be
available when 2.0 is available for the A2000.
The only programs which hasn't been ported in the 1.3 version is the programs
which use the Unix Pipe system call, these programs will be included in the
2.0 version.
The following Amiga specific programs are included:
  Drivers for the amiga display (32 colors, and HAM-E 256 colors).
  A radiance picture file to iff24 converter.
  A Imagine object to Radiance object converter.

The port isn't available yet, but should be finished within a couple of weeks.
It can then be found on hobbes.lbl.gov as patch files to the Unix Radiance
distribution (which has a size of about 3.5 megabytes compressed!).

Furthermore, I'm working on a spline based patch modeller which will be
able to generate both Imagine (TTDDD) and Radiance objects.
This modeller will probably be available within a month (in a beta version).

---
Helge E. Rasmussen  .  PHONE + 45 36 72 33 00  .  E-mail:  [e-mail]
Compel A/S          .  FAX   + 45 36 72 43 00  .
Copenhagen, Denmark

##

Subject: Water??
Date: Thu, 24 Oct 91 12:10:52 CDT
From: [e-mail] (Steve E Tietze )

Im new to this mailing list of Imagine And i dont know if this has been posted
yet.  But I was wondering how to make water with Imagine..  For example I would
like to make a Water ball (drop of water) anyone have any suggestions...


Steve Tietze
[e-mail]

##

Subject: Re:  yooo
Date: Thu, 24 Oct 91 12:13:54 -0600
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

    I've been doing a lot of stills with Turbo SIlver 3.0,
    I was wondering if any body here could tell me how to
    make a set of objects move along paths for an animation.

    Actually just how to do animations.

    I'm probably going to upload some of my pictures later on.
    I have 1meg and a 50 meg HD, this is hell.

    ;-)

    Later

Heh, well at least you have a 3000 to render on. Lots of us would like to deal
with that hardship. :) How come you only have 1meg, I thought the min. was 2 on
a 3000?

As for Silver animating I could help you out some. I did _a_little_ with it
before I moved onto Imagine. Just mail me.

    Chris
    [e-mail]

##

Subject: Re:  Water??
Date: Thu, 24 Oct 91 12:23:12 -0600
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

    Im new to this mailing list of Imagine And i dont know if this has been
    posted yet.  But I was wondering how to make water with Imagine..  For
    example I would like to make a Water ball (drop of water) anyone have
    any suggestions...


    Steve Tietze
    [e-mail]

---------------------
     Welcome Steve! I suggest the forms editor if you are going for a
teardrop shape, although spinning will work just as well. Then just set the
attributes occurdingly, or better yet ftp Steve Worley's from hubcap.clemson.edu
! :)

    Chris
    [e-mail]

##

Subject: Hubcap
Date: Fri, 25 Oct 91 06:34:28 PDT
From: [e-mail] (Action Jackson)

Loooks like hubcap.clemson got cleaned up a little.
The amiga JPEG binaries are located in the /pub/amiga/incoming/uploads
as augjpeg.lzh. The source code is in /pub/amiga/incoming/graphics.

Hey noticed some new pics in IMAGINE/pictures directory.  There's one
called skyscrape or soomething like that... looks awesome and is in
24 bit iFF as well as IFF 16 color greyscale.

-JC-

##

Subject: Teardrops
Date: Fri, 25 Oct 91 10:18:55 PDT
From: [e-mail] (Rob Borsari)

Imagine has a built in teardrop generator.  Take a sphere and conform it to a s
sphere.  I don't know why but you get a teardrop.  you may have to play around
with the default settings, I can't remember.  If you want, mail me and I
will look up the settings I used.  (hint log those really good settings or you
will not remember them if you need them again ;)  -R-
(if this is all garbled, mail me and I will re post
im having line problems)  -R-
[e-mail]  Rob Borsari  "Bourne to be wild"

##

Subject: Re: JPEG
Date: Fri, 25 Oct 91 10:21:54 EDT
From: Mark Thompson <[e-mail]>

HURTT CHRISTOPHER MICHAEL writes:
>     JPEG has already been compiled (both cjpeg & djpeg) and thrown together
> with the necessary PPM/IFF24 conversion utils and '030 verisons.

My understanding from Tom Lane, the group leader of this JPEG implementation
is that the version you are refering to is quite out of date. The very latest
version has only just recently been released.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re: Hubcap
Date: Fri, 25 Oct 91 10:57:12 EDT
From: Mark Thompson <[e-mail]>

> The amiga JPEG binaries are located in the /pub/amiga/incoming/uploads
> as augjpeg.lzh. The source code is in /pub/amiga/incoming/graphics.

The source is the new version but the binaries are old.

> Hey noticed some new pics in IMAGINE/pictures directory.  There's one
> called skyscrape or soomething like that... looks awesome and is in
> 24 bit iFF as well as IFF 16 color greyscale.

I don't know how they got there (IMAGINE/pictures directory) but they
they have been on hubcap for sometime in another directory, the LightWave
picture directory, since LightWave was used to create them. One is of a
skyscraper/skyline and the other is the cobra helicopter with the motion
blur method I described a couple months ago.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re:  Water??
Date: Fri, 25 Oct 91 12:17:09 -0600
From: Kiernan Holland <[e-mail]>

    Im new to this mailing list of Imagine And i dont know if this has been
    posted yet.  But I was wondering how to make water with Imagine..  For
    example I would like to make a Water ball (drop of water) anyone have
    any suggestions...


    Steve Tietze
    [e-mail]


-----------------------------
I take a sphere in detail editor, set the attributes to
be a little reflective and filter a lot of light,
index of refraction around 1.25 I guess,
use disturb texture with long waveforms (around 200)
and a minimal amplitude. Ray-trace and there is your
water-ball. I haven't worked with Imagine enough to
say anything exact, but you get the idea, huh??!!

##

Subject: Re:  yooo
Date: Fri, 25 Oct 91 12:31:15 -0600
From: Kiernan Holland <[e-mail]>

        I've been doing a lot of stills with Turbo SIlver 3.0,
        I was wondering if any body here could tell me how to
        make a set of objects move along paths for an animation.

        Actually just how to do animations.

        I'm probably going to upload some of my pictures later on.
        I have 1meg and a 50 meg HD, this is hell.

        ;-)

        Later

Heh, well at least you have a 3000 to render on. Lots of us would like to deal
with that hardship. :) How come you only have 1meg, I thought the min. was 2 on
a 3000?

As for Silver animating I could help you out some. I did _a_little_ with it
before I moved onto Imagine. Just mail me.

    Chris
    [e-mail]


---------------------------------------------
The 3000's do not have their kickstarts in rom form.
They are on the hard drive (in the system directory "s" in
both 1.3 and 2.0 partitions. I do have 2.4 but they
are not roms. I'm going to send hate mail to Mehd Ali.)

So..... I have 600K of my ram taken up by kickstart
and another 1.4 megs free. Workbench takes up 80K.
Then you add Virus programs, cpublit, and various
utilities plus a interlaced workbench. I have 1,030,000
bytes left of ram.

The 3000's are not all that great. If I were you,
I'd stick with the computer you have, get a flicker fixer
and a CSA Midget Racer (or Mega Midget). I'm going
to sell this one, get a 500 with a accelerator and
a 486-33 (package #1, ZEOS), that should get me more
for the money. I'm selling the 3000 (with Multi-scan monitor)
(and  bunches of PD/SW [that I haven't looked at as usual])
for about 2,600 dollars.

 Later

I'll send you mail.

(oh great, I bet my english teacher is spying on me.)

Later

##

Subject: PIC on HUBCAP
Date: Fri, 25 Oct 91 18:43:38 EDT
From: [e-mail] (Alan Price)

Howdy, I just uploaded a recent rendering to the amiga/incoming/uploads
directory on Hubcap. I hope it eventually makes its' way to the proper
imagine/pictures directory. Please check it out!

The name of the file is Corinthian.LZH and Corinthian.readme. I have a
24bit version as well, but I'm waiting to check out a freshly downloaded
version of JPEG.

Corinthian is not yet complete, I left out some statues that are to go
into recessed spaces on the walls. I have plans to use this scene in my
next animated film. BTW, if anyone from NY is reading, my last film will
be showing at the 25th NY EXPO at the New School on Friday, Nov.1 at
7:30 (better confirm that time). It is 8:30 minutes long, created with
Turbo Silver and Imagine (and DPaint) and has already shown at Ann Arbor
and Humboldt film festivals.

See Ya!

AP.

##

Subject: Conical/cylidrical lights ?
Date: Sat, 26 Oct 91 19:51:12 EET
From: Juha Kallioinen <[e-mail]>

 What's the secret between conical and cylindrical lighting ? I have tried
to use them in my renderings, but all I get visible is total blackness.

 Am I missing something here ? Do these lights even work ? When then put a
spherical light there, I got everything rendered, so it can't be wrong with
the other settings.. and I even pointed the con/cyl lights in the right
direction.. help please, anyone ?

--
o------------------------o-----o-------------------------------------------o
| [e-mail] /   __ |_______/ Tomorrow will be canceled \_______|
| Juha Kallioinen      / __/// |       \  due to lack of interest. /       |
o---------------------o  \\X/  o-------------------------------------------o

##

Subject: Re: Conical/cylidrical lights ?
Date: Sat, 26 Oct 91 14:48:02 PDT
From: Daryl T. Bartley <[e-mail]>

Aha! So I'm not the only one who gets this! I thought it was just me all this
time! Is this some kind of horrendous bug, or what?


Oh, and while I'm 'here', I had another question. Has anyone ever had a problem
with brushes, or parts of brushes, disappearing? I had an altitude map, and
approx. 1/3 of the brush didn't show up! I had it in front of the object's
surface, so no digital bounce, and everything else was set right, so I don't
know what it is. It was done on a 6 meg 3000, with a very simple object and a
global brush map. I just can't figure it out.

Thanks in advance,



Daryl Bartley
[e-mail]
Survivor: AnimeCon '91 staff
Thank you for your hard cooperation.

##

Subject: Punching Holes?
Date: Mon, 28 Oct 91 14:43:43 CST
From: [e-mail] (Dave Wickard)

In that this is my first submission
to the Imagine Group, let me take a moment to say
"Hello" from Minnesota, Home of the World Champion TWINS.  :-D

My question for you Imagineers is a simple one and probably in the
category of FAQ (fool asking a question).

I am designing as part of an anim, a doorknob on an old
fashioned backplate and want to put a keyhole through the
backplate. HOW in the world do I punch a hole through the thing
without removing the surrounding faces ON the backplate?

Simple? Yeah, but no doubt there are SOME people who are
at a beginners level like me. I tried lookin through the
tutorials but couldn't find anything applicable.

Thanks in advance for any advice in either where to get the
answer...or the answer itself.


-Dave Wickard   [e-mail]  [phone]
                Coming to you LIVE from Unisys where our motto is...
                "Nobody gets out of here alive"

##

Subject: Use Slice
Date: Mon, 28 Oct 91 16:40:53 EST
From: [e-mail] (John J Humpal)

Dave Wickard writes:

> I am designing as part of an anim, a doorknob on an old
> fashioned backplate and want to put a keyhole through the
> backplate. HOW in the world do I punch a hole through the thing
> without removing the surrounding faces ON the backplate?

Use SLICE!  Make your keyhole objects (use DETAIL editor, or draw one in
Dpaint and use Imagine's or Pixel3D's IFF->object function.  Select the
keyhole object and the backplate and then do a SLICE (from the Object
menu, I think.  I know it's Right Amiga-X).  Wait a couple of minutes,
and you'll have a whole bunch of objects named "Part.x" (x=some number).
They will be grouped, so you want to Ungroup them and discard them one
at a time until all you're left with is the backplate with the keyhole
punched through it.  You might have to tweak the number of faces on your
two original objects and/or the placement of the keyhole on the object
before Imagine accepts the Slice command.  Good luck.

-John

John J. Humpal -- [e-mail] -- short .sig, std. disclaimer

##

Subject: Hmmm
Date: Mon, 28 Oct 91 19:03:45 -0700
From: Kiernan Holland <[e-mail]>

The Backplate question,
try using the JOIN command in detail editor. Then you just
delete the parts that are in the key-hole. I think you have
to make one of the objects a child of the other.
Later

##

Subject: Re: Pixel 3D 2.0
Date: Mon, 28 Oct 91 16:28:37 MET
From: marco pugliese levi <[e-mail]>

Thankyou Steve for the help you give to all of us with these tutorials.....

In the Real3D 2.0 one you said:

> One trick that really isn't available in Imagine is the ability to load
> an image and output a grid of (connected) points with each point having
> an altitude that corresponds to the average intensity of the pixels in the
> image beneath it. I had written a routine using Glenn Lewis' TTDDD to
> perform this very feat.. I wanted to make an easy way to output a true
> height field from a C program to test my ocean wave stuff long ago.
>
> What can you do with this intensity->altitude conversion? Quite a bit!
> I took a greyscale image of myself and ran it through Pixel and got a
> very interesting object, indeed! When I also applied a color brushmap
> to it in Imagine, it gave my portrait a whole new dimension. [dodges
> rotten tomatoes.]

That's really terrific, but there's just one problem:
                                HOW TO DO THAT????!

I'm not sure you can answer me, but any help will be very appreciated.

Oh one last thing: Pixel3D 2.0 has a big bug!
When you try to manage with memory allocation under Kickstart 1.3, it ALWAYS
takes to a Guru!

                                        Bye.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Marco Pugliese                                    [e-mail]
via Roncaglia 13                                  [e-mail]
20146 MILANO                                      [e-mail]
ITALY                                             [e-mail]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##

Subject: RE: DCTV ANIM frame rates
Date: Tue, 29 Oct 91 22:04:09 -0800
From: [e-mail]

I use both CPUblit and the aforementioned View program by Cryogenic on a
16/3000 with 6 megs and get frame rates around 24/30 fps in 3bit overscan or
4bit non-overscan usually.  Once, on a wide camera pan (i.e. completely re-drawing the screen in every frame) the frame rate visibly decreased ( I don't know
exact frame rates).  This is a limitation of the DCTV unit.

I still believe that for it's price DCTV is an excellent graphics enhancement
device.  The only other product which I believe can be fairly compared to it as
competition is the HAM-E box.  This has it's advantages, such as RGB output,
although for my purposes, as I like to dump my animations to tape, this would
be no advantage to me.  It also has it's disadvantages, such as being
incredibly slow (at least when I saw it, several months ago).  I have also not
seen HAM-E animations, though I know they are possible.  I was also under the
impression that HAM-E was not able to produce as many colors on screen at once
as DCTV, having 256 or 200,000+ color modes, but I understand this may have
changed.

In any case, I believe this area is for discussion Imaging and related topics (
such as using DCTV with Imagine), not for suggesting alternatives to the Amiga
(as IMHO there aren't any), so I'd ask that the flood of Amiga bashing end


  ___________________________________________________________________________
 /  [e-mail]                                       |     //         \
|                                                           |    //Only Amiga |
|'We're all in it together' -Tuttle from 'Brazil'           |\\ //  makes it  |
 \__________________________________________________________|_\X/___possible_/

##

Date: Wed, 30 Oct 91 12:00:41 PST
From: [e-mail] (Rob Borsari)

 When both were just coming out, Dr. Gandalf bought a HAME to test.  At that
time the problem with animation on the HAME was that there was no way to lock
the pallette.  I think the HAMe people said that they couldn't fix that because
of the way the unit worked.  I don't know if this is still true but you can ask
them.  We found the quality of the unit to be good but not in the class of
the firecracker or other true 24bit boards.  If anyone is interested I will
ask Eric to tell me why he went with dctv over hame.  -R-
[e-mail]   Rob Borsari  "Bourne to be Wild"

BTW.  just got off the phone.  Some dealers have been shipped 2.0 roms
and upgrade kits for 2000's (and I assume 500's)  I am getting mine tonight.
for $99.   have fun and try not to call all at once ;);)

##

Subject: Re: prismatic refraction (non-Imagine related)
Date: Wed, 30 Oct 91 13:38:23 EST
From: Mark Thompson <[e-mail]>

For those interested in human color perception, I have a correction to
my previous post. Otherwise, stop reading now.

In a past reply to Scott Krehbiel, I talked briefly about light and color
as well as human color perception. In fact I stated that "We do not see
RGB". Scott questioned this and with good reason. It is true that we do
not see RGB in the way computer representations use RGB, ie. three
individual frequencies whose ratio of amplitudes determines the color.
What we do have are receptors called "cones" that are sensitive to three
broad bands of frequencies. The spectrums of three bands are overlapping
and are centered at red, green, and blue. Together, they cover the entire
visible spectrum rather than three distinct frequencies. Anyone wishing
more detailed info on human vision and color perception could refer to
"General Opthamology" by Daniel Vaughn and Taylor Asbury. While this is
not the best source on this topic, I could not locate all my papers from
my machine vision days.

In any event, the statements by Scott and myself were BOTH correct but
lacked sufficient detail. Hope this clarifies things...if you were even
interested.
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~%
%      `       '        Mark Thompson                 CONCURRENT COMPUTER  %
% --==* RADIANT *==--   [e-mail]        Principal Graphics   %
%      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   %
%     Productions       [phone] [phone]   & General Nuisance   %
%                                                                          %
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Anim Players We'd Like To See
Date: Wed, 30 Oct 91 15:20:03 EST
From: [e-mail] (KENNETH DYKE)

One thing that I haven't seen anyone take advantage of yet with Anim players
is the fact that the A3000's CPU has a 32-bit data bus to chip ram.  As far
as I know, the Anim formats used today use a variation of the Short Delta
mode described in the Anim specs.  I would think that using the Long Delta
mode would give much better frame rates on an A3000 (by being able to cut
the number of accesses to chip ram in half) at the expense of a bigger anim
file.

I would like to see programs that generate (and play back) anim files give
the user the option of using either short or long word compression. Bit #
0 has been set aside for this function already (in the mode longword of the
anim header), so it's just a matter of someone using it.


Kenneth C. Dyke                                    | Internet: [e-mail] |
MTCA Amiga SIG President and CS major at MTU       | BITNET:     kcdyke@MTUS5 |
A3000-25/50                                        | Phone:  [phone] |

               MS-DOS: The official operating system of Hell

##

Subject: TV-unit
Date: Wed, 30 Oct 91 22:21:45 EET
From: Juha Kallioinen <[e-mail]>

 If I made a tv-unit, which would have a flat plane as the screen, and I would
have made the planes attributes to be white and that it would be a 'light',
then if I attached an animation to the plane with multiple brushes, would the
TV-unit (or actually the plane) illuminate a dark room when put in one..

Anyone got that ? I'm not even sure if I did :)

--
o------------------------o-----o-------------------------------------------o
| [e-mail] /   __ |_______/ Tomorrow will be canceled \_______|
| Juha Kallioinen      / __/// |       \  due to lack of interest. /       |
o---------------------o  \\X/  o-------------------------------------------o

##

Subject: TV screen with brush maps and light
Date: Wed, 30 Oct 91 18:57:46 -0500
From: [e-mail] (Michael B. Comet)

    With regards to brush maps and lighted objects, I am pretty sure
that the screen would still illuminate a light (whatever color you set
in the light requester, not the object color - if you are brush mapping
over the whole screen, then object color isn't important).

    I have made a similar animation where the camera flys up to a
computer and computer screen, then the monitor 'turns on' and then play
a sequential brush map created from a previous animation.  really cool
effect to have an animation in an animation.  I did not make the screen
light though, I just added another light in front of it.

    Note : I have found a bug in imagine 1.1 and have notified
impulse.  It seems in pick faces mode you have the ability to make
specific faces "bright" and "light".  Try it.  After you do either of
them and then select OK from the attributes requester, go back into the
requestor and look!  Lo and behold, even though the same faces are still
picked, the two little boxes Bright and Light are no longer marked.
This means to make light and bright objects you need to group them.

    Another use I have tried for sequential brush mapping is
digitizing every other frame fro a video laser disc player, and then
sequentially mapping them on to objects.  Instant video in animation
just like the 'Change myself' (ug -lightwave) animation.

    I hope this helps.


                    Mike Comet
                    mbc
                    [e-mail]

##

Subject: New Image on Hubcap
Date:    Fri, 1 Nov 91 00:10 EST
From: "Marc Rifkin" <[e-mail]>

I just uploaded my first finished rendering to hubcap uploads.  I usually
don't get to satisfactorilly complete a pic like this one.  I'm a heavy
critic of myself: it is "good."

STELLARSHIP.LZH

Marc Rifkin
Integrative Technologies Lab, Penn State University
5F Mitchell Building, University Park, PA 16802
814-863-8062 or for you normal people, [e-mail]
"Say, that's a nice bike!"

##

Subject: JPEGged files via this mailing list?
Date: Thu, 31 Oct 91 21:27:29 PST
From: [e-mail]

Say, dudes.. with the availability of JPEG now, we can get those
24-bit images down to an easily-mailable size.  How about sending
some of your hot images to this mailing list as a uuencoded JPEG
file? They shouldn't run more than 40-50K each. Would this cause
any undue strain with mailing systems and such?

Further, any images I received this way, I would gladly post into
the "Hi-End Graphics" library in the Amiga Zone on Portal (of which
I'm the Sysop), the commercial system from which I'm posting this.
We have hundreds of Amiga users online, we have the JPEG binaries
in our library, we have Perry K online, constantly talking about
the new ADPro 2.0 (which has JPEG loader and saver) so all bases
would be covered.

Good idea? Bad idea? You tell me. :)

Regards, Harv

##

Subject: Re: JPEGged files via this mailing list?
Date: Fri, 1 Nov 91 7:23:01 EDT
From: [e-mail] (Dan Drake)

>
> Say, dudes.. with the availability of JPEG now, we can get those
> 24-bit images down to an easily-mailable size.  How about sending
> some of your hot images to this mailing list as a uuencoded JPEG
> file? They shouldn't run more than 40-50K each. Would this cause
> any undue strain with mailing systems and such?

I would much prefer havine a seperate mailing list for pics.  It's that
we are always close to 100% full here, and a bunch of 50k files sitting
in my mail directory would suck up alot of unnecessary space.

just my .000002 millionths of a ruble.

dan.

##

Subject: Re: jpeg pic
Date: Fri, 01 Nov 91 09:45:54 EST
From: Mark Thompson <[e-mail]>

> Does JPEG work better on 24-bit pictures than HAM pictures cause it really
> pixelized some of my ham pictures.

JPEG definately works best on 24bit images. In many cases, the dithering and
quantization errors inherent in GIF, HAM, and other color pallete based image
formats can cause very poor results when JPEG compressed. This is due to the
fact that the JPEG algorithm was designed specifically to work with full
color imagery.  The beauty of it though is that a JPEG compressed 24bit image
is not much different in size than a JPEG compressed GIF image, so there is
really no reason to use the color pallete based image format in the first
place.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re: JPEGged files via this mailing list?
Date: Fri, 01 Nov 91 10:30:33 EST
From: Mark Thompson <[e-mail]>

> How about sending some of your hot images to this mailing list as a
> uuencoded JPEG file?

Not a bad idea Harv but besides the fact that a few people may not be able
to handle this much data, I have one major problem with that: EVERY
message I post to this list bounces back at me a minimum of 4 times.
It is a minor nuisance for messages like this but all that bounced image
data would be a burden.

I will however be more than happy to send you some of my 24bit images
directly to [e-mail]. My gallery of 3D artwork numbers
around 20 images at this point and a couple I consider exceptional.

> we have Perry K online, constantly talking about the new ADPro 2.0

Next time you talk to Perry, I have a couple requests for the next rev of
ADPro..... Add a ppm Loader/Saver (this would be utterly trivial) or
make it possible for people to add their own Loader/Savers. And second,
I would really like to see an Operation module that allows you to apply
painterly effects to your image very much like Paul Haeberli's Siggraph '90
paper "Paint By Numbers: Abstract Image Representations". Once again, if
this is beyond their capabilities, providing a method for users to write
their own modules would be very useful. My goal is to be able to render
a 3D animation and then process the image sequence (via ARexx with ADPro)
to make the animation look like a moving oil painting (as well as other
effects). I have been experimenting with this a bit but lack the time to
create a program to handle the entire task. ADPro would fit very well into
enabling these effects.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Stuff
Date: Fri, 01 Nov 91 12:14:05 EST
From: [e-mail]

Some quick notes; I haven't had much time to post lately since that long
Pixel 3D post.

---

Lights: Cylindrical and conical lights work great, they aren't buggy. The
important part is to understand how the axes control the way the light
come out. For a cylindrical light, the x axis controls the radius of the
beam, AND THE Y AXIS CONTROLS HOW MUCH DISTANCE THE LIGHT ILLUMINATES!
Thus, having an axis thats 10 units in X and 1000 in Y is not unusual.
Same goes for a conical light- the Y axis must be huge if you want to
illuminate a large scene. If you have a small Y axis, you'll think the
light is broken because you don't see anything; in reality the light just
doesn't extend that far.

I might have gotten the X & Y directions backwards, but I think that's right.

-----

Nope, Pixel 3D does NOT bevel the back of extrusions. Howvwever some surgery
in the Detail Editor can fix that. A pain, but possible.

-----

Someone noted that the signal-to-noise level on the list suddenly
dropped about a week ago; some posts started to appear which rerally had no
relevance to 3D or Imagine. Please think about what you're posting: if it
doesn't belong, don't post.  Also, be courtious! Don't have enormous sigs,
Don't quote other articles in their entirely. A final note: if you post
a one or two line message, odds are you're not saying much and you can save
on bandwidth if you just save it up. For example, this post makes several
short unrelated points.

Thanks guys.


BTW- a lot of the noise may come from the sudden exposion in the size of
the list... Colleges are in session and the list is getting publicity..
I've added about 50 people in the bast two weeks!  (now at 240 or something)

--------

Last, I'll upload someof my new pix as soon as Hubcap straightens
itself out.  I haven't been able to upload there for weeks. Doug?

-Steve

---------------------------------------------------------------------------
Steve Worley                                        [e-mail]
---------------------------------------------------------------------------

##

Subject: those damn lights
Date: Sat, 2 Nov 91 1:33:52 EET
From: Juha Kallioinen <[e-mail]>

 AAgh.. I can't get the conical/cylindrical lights to work at all..

I have tested it on a verry simple setup like this:

|
| -O
|
    \
     C

Where that -O is cylindrical light, with that little 'beak' (or whatever)
pointing to the left. And that vertical line is a simple plane. And \ is
the camera lookin up and left.                                       C


Now the space between the plane and the light is about 60 iu, and if I set
that light, so that it's size is e.g. (32,100,z) (z doesn't matter, I quess)
Shouldn't that produce some light on the plane, well it doesn't. No matter
what the values are for the size of the light, I can't get anything but
darkness.. (perhaps it's black light :) )..

Frustration, frustration..

Suggestions ? Example projects to hubcap ? Anything helpful is appreciated,
and I've just read what Steve W. wrote about this subject, but it didn't
help..

uhh.. bye for now.



--
o------------------------o-----o-------------------------------------------o
| [e-mail] /   __ |_______/ Tomorrow will be canceled \_______|
| Juha Kallioinen      / __/// |       \  due to lack of interest. /       |
o---------------------o  \\X/  o-------------------------------------------o

##

Subject: Black-lights
Date: Sat, 2 Nov 91 01:36 MST
From: Oberon Kenobi <[e-mail]>

     I'm that annoying person who wanted to have simulated refraction for
different wavelengths of light.  (I had Dave Martin post it for me, so don't
blame him.)

     Anyway, I have another annoying future feature for Imagine.  Blacklights!!!
I know.  I know.  black-light is an oxymoron.  (There will be those who dispute
that statement too.)

     I would like a light source to be able to be specified as "black."  Objects
could be specified to emit a given RGB value in proportion to the amount of
blacklight that is hitting it.  Again emulating reality.  (After all, isn't that
the purpose of ray-tracing; to emulate the effects of the electromagnetic
spectrum?)  I can't use a "glowing" object because I want shadows from the
blacklight.

     Before you get me for not knowing the physics behind THIS feature that I
want, I'll explain it superficially.  (I'm sure that there will be others who
NEED to provide more information, and will.)  Ultraviolet (and some violet) is
emitted from the black-light.  Certain substances absorb, and then re-emit this
energy in the visible spectrum.

     What do you guys/gals think?  Once that is done, different frequencies of
blacklights can be made.  (I can hear the groans now.)  Maybe my next suggestion
could be alternate frequency "film."  (I know, it can be simulated.)
--
____  ____  ____ |=============================================================|
\   \/    \/   / |   "vi?  Because I don't have a real editor." -- U*x user    |
 \Weber  State/  |=============================================================|
  \University/   |Oberon| Internet: val@{alpha|beta|gamma|csulx}.weber.edu     |
   \___/\___/    |Kenobi|                                                      |
Computer Science |=============================================================|

##

Subject: Re: Conical/Cylindrical light difficulty
Date: Sat, 2 Nov 91 20:22:11 -0500
From: [e-mail] (Michael B. Comet)

    In regards to the problem with the cylindrical light and plane :

        |
        |  -O
        |
        |
           \
                       C


    First, make sure the light color (RGB) intensity isn't at 0,0,0
eg : no light being emitted.

    Also, make sure that the light is truly aiming at the plane.
And that the camera is looking at the section at the part of the plane
the light is aiming at.

    In regards to lights, the x size is the radius and the y is the
length.  This means that your light should have a radius of 32 and should
illuminate up to 100 iu's I think, feel free to correct me if I'm wrong.
Anyways, you should be able to see something, if not, try making an
animation where you either move the light closer and farther to the
plane, or tween the x or y values of the light.

    Hope this helps.

                Mike C.
                [e-mail]

##

Subject: The Art Dept Proffesional 2.0
Date: Sun, 3 Nov 91 19:13:35 -0500
From: [e-mail] (Michael B. Comet)

    In regards to the release date of 2.0 for adpro - it's already
out.

    $75.00 upgrade
    $30.00 upgrade for adpro if you bought it after sept 1,1991

    ADPro 2.0 includes

    everthing from adpro 1  plus........


    IFF, JPEG (JFIF standard), PCX, GIF, QRT, DPIIe, Digi-View 3.0,
Sculpt, Imagine, MacPaint, BMP, Postscript   - formats fro load/save
etc.....
        except postscript which is write only and dv 3.0 and
macpaint which is read only.

    direct support for CBM A2410, GVP IV 24, ACS Harlequin, HAM E,
Mimetics FrameBuffer, DCTV, Firecracker 24.

    Digitize directly in ADPro with your : PP&S Framegrabber or GVP
IV 24.

    New WYSIWYG functions

    2.0 compatible

    and the list goes on...........................




                Mike Comet
                [e-mail]
##

Subject: Imagine Lights
Date: Mon,  4 Nov 91 08:01:18 PST
From: [e-mail]

If you're having trouble with lights here is a little clue.  Imagine will
allow you to set the brightness of the lights much higher than 255.  Try
setting them to 1300 (in each RGB) and then after testing with a scanline
render to see if the brightness works adjust them down.  Also set the
Ambient light in your global settings to around 100 (in each RGB).  I have
found that light brightness settings between 500 and 800 usually work best.

##

Subject: Hubcap change
Date: Mon, 4 Nov 91 10:09:01 PST
From: [e-mail] (Steve Worley)

I finally discovered why nobody has been able to upload to Hubcap! Apparently
there is a new procedure for uploading to the hubcap.clemson.edu site.
The README:
----------------------------------
Uploads shall be done the following way:
        1) upload file to incoming/uploads directory
        2) upload a text file along with your upload files so
           that I may know what is being uploaded.
        3) email me a message and become a registered uploader.
           I would like to keep track of who uploads and what
           they upload.
        4) if you upload ray-tracing files (Imagine, etc.), indicate
           that in the email message so I know how many people are using
           Imagine and the like.
        5) As soon as uploads startup again, a log of all uploaders
           and Imagine/raytracing users will be kept on file.

MAKE SURE YOU REGISTER WITH [e-mail] WHEN UPLOADING!!!

Any problems/comments/suggestions/money/etc. should be emailed to:
        [e-mail]  John Maurice
-------------------------

I for one am not pleased by this change: First, there are now a lot of
misfiled files in the Imagine directories, and more importantly we
don't have any control of uploads- I can't say "Hey, there's now a
demo version of Pixel 3D 2.0 on hubcap in the IMAGINE/misc
directory.", since there might be several days or more delay before
things are refiled. Also, some of the uploads I make are temporary
files (a dump of raw posts that Marvin Landis graciously assembles
into our archive) and this new uploading procedure makes this very
difficult.

I think this new upload procedure would be a good way to control misuse
of the FTP site, but I haven't seen ANY previously-  all files seemed to
be properly filed, nobody deleted files... do we need this control now?

Is it time to find a new site, or am I overreacting to this change?
It could quite possibly be the latter...

-Steve

--------------------------------------------------------------------------
Steve Worley                                       [e-mail]
--------------------------------------------------------------------------

##

Subject: New Site?  Hubcap change seems less than neat...
Date: 4 Nov 91 13:49:00 EST
From: "SYSTEM MANAGER" <[e-mail]>

I just read with a bit of disgust the change at hubcap.  I suppose
if we could change the minds at hubcap it would be worth it, but if
not...

I am offering hard disk space on vger.nsu.edu for this mailing list.

-mark=
[e-mail]

##

Subject: New Site?  Hubcap change seems less than neat... (fwd)
Date: Mon, 4 Nov 91 16:30:21 EDT
From: [e-mail] (Doug Dyer)

Forwarded message:

> Message-Id: <[e-mail]>
> Date: 4 Nov 91 13:49:00 EST
> From: "SYSTEM MANAGER" <[e-mail]>
> Subject: New Site?  Hubcap change seems less than neat...
> To: "imagine" <[e-mail]>
>
> I just read with a bit of disgust the change at hubcap.  I suppose
> if we could change the minds at hubcap it would be worth it, but if
> not...
>

Now now... being a mind at hubcap I can tell you I disagree as well.
Perhaps if we politely emailed [e-mail] the situation,
what we liked, and suggest what we want now he will see that imagine
users dominate hubcap activity.   Actually I am sure he would replace
the write permissions if he knew how many use the imagine area.


> I am offering hard disk space on vger.nsu.edu for this mailing list.
>
> -mark=
> [e-mail]

--------------------------------------------------------------------------
Doug Dyer     *      Clemson University     *     [e-mail]
Quote 'O Fun: DOS is a bike without the seat.           skydive naked
--------------------------------------------------------------------------

##

Subject: 24bit render problem
Date: Mon, 4 Nov 91 16:15 PST
From: [e-mail] (Scott Busse)

Another quirk arises :(
I'm rendering a sequence of frames that involves a segmented tube and 5
lights (I had to use 5 lights because the phong shading is so wimpy it wouldn't
smooth the object. Oh to have Sculpt's phong shader...).
  When I render the image in 320x400 (as a test), it looks the way I have set
it up. When I try to render it in 24bit (scanline) at 640x480 (for conversion
to PCX), and load it into ADPro to look at it (by converting it to HAM), the
object is rendered in a grey silhouette, as if there was nothing lighting it
but the global illumination. The same thing happens if I try to render a HAM
640x480 version and look at it in ADPro.
   The clincher is that when I do either the 24bit or the HAM in trace mode (at
640x480), it works fine.
   Any ideas folks?
--
* Scott Busse email:           O    O   O_     _      ___ .....
* CIS 73040,2114              |||  /|\  /\   O/\_     /         O    )=|
* [e-mail]    l   | |   |\    / \   /\                _\
* [e-mail]                  Live Long and Animate... \

##

Subject: Imagine 2.0 Demo
Date: Tue, 5 Nov 91 10:32:21 PST
From: [e-mail] (Steve Worley)

For those of you who are in the San Francisco area, Imagine 2.0 (well,
a late beta version anyway) will be shown at the FAUG (First Amiga
Users Group) meeting in Palo Alto TONIGHT. (Tuesday, November 5). You
too can find out about the fun features of Imagine and heckle me in
person as I uncomfortably lecture in front of you.

The meeting will be held at the Hyatt Rickey in Palo Alto CA, at 4219
El Camino Real. The meeting starts at 7:30 pm, and will probably last
about 2 hours. Note that there are TWO Hyatts in Palo Alto on El
Camino; the Rickey is the smaller one with a brown wood exterior.
Normally meetings are held in the large multi-story Hyatt, but they
decided to move the location for some reason or another.

AVID Magazine (Amiga Video) will be printing the debut article on Imagine
2.0 in this month's issue, out next week. The article lists all of the
features and new toys in the update, so if you miss the demo, you might
pick up a copy to get the scoop. Better yet, subscribe! It's a great
magazine.

Tomorrow I'll post a _short_ note on the biggest new features. Why not
a comprehensive list? Because I already wrote one for AVID. :-)

-Steve

--------------------------------------------------------------------------
Steve Worley                                       [e-mail]
--------------------------------------------------------------------------

##

Subject: NEW new hubcap changes :)
Date: Tue, 5 Nov 91 17:37:07 EDT
From: [e-mail] (Doug Dyer)

John and I had a discussion and decided that imagine should remain open.
follow these steps so we can keep track of things:
(with my comments like this)
--------------------------------------------------------------------------------
Due to space problems, there have been several changes.

    1) You may upload to whatever directory you want to
       in the /pub/amiga/incoming section.
        (actually, some dirs. will remain read only)

    2) BUT, files will only exist for 2-3 weeks.
        (imagine pic files will last a month.. text has no
         limit...)

    3) Please register with [e-mail] when
       uploading so that we can keep track of who is
       uploading what and when. thanks.
        (for every upload, just a readme that explains the file,
         who did it, etc.)


Any problems/comments/suggestions/money/etc. should be emailed to:
    [e-mail]  John Maurice
    (or me.  I also take visa,mc)
-------------------------------------------------------------------------

if the group decides what pictures are "valuable examples" I will let
the powers that be know.  I think this is a great solution because we
will hit space limitations eventually, and with the group growing it is
important to prepare for cycling pictures (coming and going).

the first thing to do is to delete all those lightwave pics.  then
those of you who uploaded pics in the past and think they have had time
to circulate (or have more than one version uploaded) decide if they
have to go for now. That will show a concern on our side that we as a
group appreciate the free service and are willing to work with it.

The imagine directory is by far the most active on hubcap, and has a
local lobbyist to boot :)

thanks,

--------------------------------------------------------------------------
Doug Dyer     *      Clemson University     *     [e-mail]
Quote 'O Fun: DOS is a bike without the seat.           skydive naked
--------------------------------------------------------------------------

##

Subject: LA Int'l Anim Celebration & Imagine
Date:    Tue, 05 Nov 91 09:35 PST
From: Ivan I                               <[e-mail]>

I thought you all might be interested to know that there were two
films at the LA International Animation Celebration's "State Of the
Art Computer Animation" screening which were Amiga produced this year.
One, "Dot in Duggie Sic Diggie" seemed to be a DPaint animation, or
some similar 2D product, and was produced by Wayne White.  The other,
"Curiosity (Almost) Killed The Cat", was an Imagine animation which
also employed The Director and a host of other software, but was
entirely Amiga produced; it surprised me in that it was one of the
first lengthy character anims I've seen from Imagine so far (at about
5 minutes running time).  My friend Nina and I almost immediately
recognized it as an Imagine animation because 1) there was that
"crawling ants" effect on a concrete floor, 2) a wooden box exploded
into triangles, and strangely enough 3) we each of us seemed to
recognize the wood grain texture (used on the cat to replicate
stripes) as peculiar to Imagine.

Anyway, both films were very well received & quite funny.  Commodore
was a sponsor this year, and will hand out an Amiga each to the best
student film & best first film for public exhibition in what I
consider to be an excellent marketing move.  They also were quite well
received when the reel of sponsors came up before the screening.

I don't have the slightest idea where or how either of these will be
distributed, but I recommend keeping an eye out for them both.


##

Subject: filter maps
Date: Tue, 5 Nov 91 22:12:16 GMT-0500
From: Scott Matthew Krehbiel <[e-mail]>

I seem to be having a rather frustrating time with filter brush maps
in Imagine, and I wonder if anyone could help me out.  I know that
my brush axis is placed correctly, 'cause I get a beautiful color
map of my brush, but no transparency.  I have shinyness off, and
have messed around with the filter values for the object, and got
the results you'd expect if it was supposed to be a color map.

Just to see if my technique is totally brain-dead, I tried
reflection mapping, and got the same result - a rather nice
color map, but no reflection.

Is there some magic setting that I'm forgetting in attributes?
Thank you much for your help
- Scott Krehbiel
[e-mail]

##

Subject: Re: filter maps
Date: Tue, 5 Nov 91 20:57:23 MST
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

>
> I seem to be having a rather frustrating time with filter brush maps
> in Imagine, and I wonder if anyone could help me out.  I know that
> my brush axis is placed correctly, 'cause I get a beautiful color
> map of my brush, but no transparency.  I have shinyness off, and
>
    Scott, quick question. Are you rendering in ScanLine or Trace? One
guy I know drove himself nuts trying to get a filter to work in ScanLine,
which it doesn't. Most likely the same for reflection maps too.

    Chris


##

Subject: Complete TeX Roman font available on hubcap
Date: Tue, 5 Nov 91 22:31:46 PST
From: [e-mail] (Glenn M. Lewis ~)

    For those that are interested, I have placed the complete TeX Roman
font (cmr17.1255pk) on hubcap.clemson.edu [130.127.8.1] in the directory:
pub/amiga/incoming/uploads.  It will probably be moved to the imagine/OBJECTS
directory sometime soon.

    There are three files:
total 7591
-rw-r--r--  1 glewis      14441 Nov  1 14:17 TeXRoman.Readme
-rw-r--r--  1 glewis    2283478 Nov  1 14:45 TeX_Roman_Font_Object.lzh
-rw-r--r--  1 glewis    5455054 Nov  1 17:33 TeX_Roman_Font_Preview.lzh

which you can see are rather (!) large.  The first describes the details of
the font, how I made it, how you can make others, and how to use it.
The second file contains 127 objects of the letters that make up the TeX font.
The third file contains 127 PostScript wireframe previews of the letters so
that you can see what you can expect to get.  (I would imagine that the
maintainers of hubcap would appreciate it if you limited your downloads to
off-hours.)

    Are there any strings attached to this font?  (Pardon the programmer's
pun :-)  Yes and no.  You are free to redistribute these objects or convert
them to other formats at will.  If you find the font useful, I would very
much appreciate it if you register for the TTDDD shareware package (at which
point you can make objects from any TeX font you want).  Thanks!

                            -- Glenn Lewis

[e-mail] | These are my own opinions... not Intel's

##

Subject: spherical brushmaps
Date: Wed, 6 Nov 91 8:30:17 EDT
From: [e-mail] (Dan Drake)

Here is the problem:
I made a brush in DPAINTIII which is a 640x400 medium blue square, then
took the color white and the airbrush and sprayed some pattern on it.  I then
wrapped it on a sphere to simulate the clouds on a planet.  Everything came out
pretty much as I expected, except for two things.

1:) the sphere is no longer smooth, you can see the angles.  I tried using the
sphere primitive, and also the sphere object.  I made sure that the primitive
sphere had phong smoothing on, but results were the same.

2.) when My planet is rotating, at one point there is a white line that goes
from the top of the planet to the bottom.  North to south pole.  The color
of the sphere before I did the mapping is white.  Does this mean that the map
did not go all of the way around the planet?  I guess I can solve this by
changing to color of the planet to blue instead of white.


thanks ahead of time,

dan.

##

Subject: Re: spherical brushmaps
Date: Wed, 6 Nov 91 07:24:20 -0800
From: [e-mail] (Edward Chadez)

On Nov 6,  8:30am, Dan Drake wrote:
} Subject: spherical brushmaps
}
}
}
} 1:) the sphere is no longer smooth, you can see the angles.  I tried using the
} sphere primitive, and also the sphere object.  I made sure that the primitive
} sphere had phong smoothing on, but results were the same.

You've probably enlarged the sphere or zoom'd in really close.  I've seen
this, myself, when I've enlarged the sphere (in the stage editor).  My
advise?  You might want to use the non-primative and enter in a larger
number for sections.

Another solution is not to enlarge the sphere, and don't zoom in too close
so that you can see the angle edges.

}
} 2.) when My planet is rotating, at one point there is a white line that goes
} from the top of the planet to the bottom.  North to south pole.  The color
} of the sphere before I did the mapping is white.  Does this mean that the map
} did not go all of the way around the planet?  I guess I can solve this by
} changing to color of the planet to blue instead of white.

Yes, the brush map didn't make it all the way around...but I don't think
it's because the brush map is too small.  Instead, you might want to make
sure that your wrap axis is properly aligned and wider than the radius of
the sphere.

I've made dozens of planets, using both the color map and the altitude map.
Somethings that point to not aligning the wrap axis correctly are the one
you described as well as the north pole and/or the south pole not getting
wrapped.  A simple adjustment of the wrap axis should take care of the
problem.

}
}
} thanks ahead of time,
}
} dan.
}
}-- End of excerpt from Dan Drake


-->Edward Chadez  =Amiga3000=
   Computer Animation/Pixel Pusher and
   Intuition WB2.0 NewLook Programmer.

--
  //                                                               ()__()
\X/ [e-mail]/Edward Chadez     CARL Systems(303)861-5319    (  )
                                                                    M.M.

##

Subject: Re:  spherical brushmaps
Date: Wed, 6 Nov 91 11:00:25 CST
From: Wayne Haufler 283-4160 <[e-mail]>

[e-mail] (Dan Drake) writes:
> 2.) when My planet is rotating, at one point there is a white line that goes
> from the top of the planet to the bottom.  North to south pole.  The color
> of the sphere before I did the mapping is white.  Does this mean that the map
> did not go all of the way around the planet?  I guess I can solve this by
> changing to color of the planet to blue instead of white.
>
I had the same problem quite a while ago.  It turns out that I had forgotten
to paint underneath the horizontal menu bar and vertical toolbox bar in
DPAINT III.  It is easy to forget these regions.  A Fill with these bars
visible does not fill underneath them.  Imagine wraps the entire 640x400
image, including those possibly unpainted regions.

This may or may not be your problem.  That unpainted region on mine appeared
more like a 10 degree black wedge at the 'back' of the sphere.  BTW,
a rendered frame of the camera aimed at the north or south pole of that
sphere resulted in a pretty cool and strange 'tree ring'-like image.
Probably the results of severe distortions of the brush at those
regions.

Hope this helps :)


         __         Wayne A. Haufler   [Christian/SW Engineer/XLib'er/Amigan]
\\ /\\ /\\ //_       [e-mail]          MDSSC - Houston
 \/--\// \//__       Hobby:"Exploring the Use of Computer Graphics and
     //                     Animations To Support Christian Endeavors"

##

Subject: Re: spherical brushmaps
Date: Wed, 6 Nov 91 10:36:09 MST
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

> }1:) the sphere is no longer smooth, you can see the angles. I tried using the
> }sphere primitive, and also the sphere object.  I made sure that the primitive
> }sphere had phong smoothing on, but results were the same.
>
> You've probably enlarged the sphere or zoom'd in really close.  I've seen
> this, myself, when I've enlarged the sphere (in the stage editor).  My
> advise?  You might want to use the non-primative and enter in a larger
> number for sections.
>
    Hmm, I thought the primative wouldn't do this. Hence, the whole point
of the primative! Maybe 2.0 will offer some splined based stuff like spheres
& lathes.

    Chris

##

Subject: Re: filter maps
Date: Wed, 6 Nov 91 9:40:36 PDT
From: [e-mail] (John T. Grieggs)

>
>
> I seem to be having a rather frustrating time with filter brush maps
> in Imagine, and I wonder if anyone could help me out.  I know that
> my brush axis is placed correctly, 'cause I get a beautiful color
> map of my brush, but no transparency.  I have shinyness off, and
> have messed around with the filter values for the object, and got
> the results you'd expect if it was supposed to be a color map.
>
> Just to see if my technique is totally brain-dead, I tried
> reflection mapping, and got the same result - a rather nice
> color map, but no reflection.
>
> Is there some magic setting that I'm forgetting in attributes?
> Thank you much for your help

Yup, but not in attributes.  Make sure you are ray-tracing.  Filter maps and
reflection maps don't work when scan-line rendering...

> - Scott Krehbiel
> [e-mail]
>
_john


##

Subject: New pic on Hubcap
Date: Wed, 6 Nov 91 09:45:57 PST
From: [e-mail] (Steve Worley)

Hubcap uploads are back!

I uploaded a new rendering of mine in both hires and JPEG formats. The
picture is called "Strike."  Enjoy!

-Steve


##

Subject: Re:  Complete TeX Roman font available on hubcap
Date: Wed, 6 Nov 91 09:53:34 PST
From: [e-mail] (Glenn M. Lewis ~)

>>>>> On Wed, 6 Nov 91 11:46:41 PST, [e-mail] said:

Viet> Glenn,

Viet>    Would it be possible to digitize Compu-Graphic/Adobe fonts with
Viet> your program?  These are more predominate on desktop machines.

    Not at present, but if I am given the format of Compu-Graphic
files, I could whip something up.

    Also, I can forsee that it would be quite easy to convert
PostScript fonts to TTDDD, using GhostScript and it's "pstopbm" program.

    In fact, when I get a chance, I will add a PBM2TTDDD program to
TTDDDLIB.  That should really be neat.  For any PostScript page, be able
to create an Imagine object that is basically the extrusion (in depth)
of the 2-dimensional PostScript page.  Would anybody like this?  I know
I would have fun with it!  :-)

                            -- Glenn

[e-mail] | These are my own opinions... not Intel's

##

Subject: Re: spherical brushmaps
Date: Wed, 6 Nov 91 10:11:19 -0800
From: [e-mail] (Edward Chadez)

On Nov 6, 10:36am, HURTT CHRISTOPHER MICHAEL wrote:
} Subject: Re: spherical brushmaps
}
Someone else wrote...

}  1:) the sphere is no longer smooth, you can see the angles. I tried using the
}  sphere primitive, and also the sphere object.  I made sure that the primitive
}  sphere had phong smoothing on, but results were the same.

And I replied...

}  You've probably enlarged the sphere or zoom'd in really close.  I've seen
}  this, myself, when I've enlarged the sphere (in the stage editor).  My
}  advise?  You might want to use the non-primative and enter in a larger
}  number for sections.
}

And Chris noted...

}     Hmm, I thought the primative wouldn't do this. Hence, the whole point
}  of the primative! Maybe 2.0 will offer some splined based stuff like spheres
}  & lathes.
}
}     Chris
}

Well, Chris is exactly right-->the primatives shouldn't pull this kind of
stuff.  But, as an example, create a sphere in Detail and save it.  Then go
into stage and position the camera in front of the sphere, enlarge the
sphere so that it's framed nicely in the perspective view (ie, takes up
almost 95% of the screen).  Save your stage settings...

Now trace the picture.  With scanline, you will definatly see angle edges
on the sphere.  I believe this is ditto for full TRACE...but I don't recall
at this time.

It remains to be seen if Imagine 2.0 has improved the sphere primative.


-->Edward Chadez  =Amiga3000=
   Computer Animation/Pixel Pusher and
   Intuition WB2.0 NewLook Programmer.

--
  //                                                               ()__()
\X/ [e-mail]/Edward Chadez     CARL Systems(303)861-5319    (  )
                                                                    M.M.

##

Subject: The scoop on Imagine 2.0
Date: Wed, 6 Nov 91 10:13:52 PST
From: [e-mail] (Steve Worley)

Last night I demonstrated Imagine 2.0 to members of FAUG- presenting to
an audience containing both Allen Hastings and Jay Miner makes a person
quite nervous... but it went really well. I met 8 list members at the
meeting as well.

Well, anyway, here is a quick summary of the features of Imagine 2.0.
This is not an complete list, but covers maybe 80% of the new
features. (These are the features I personally saw in my late beta
copy of 2.0) Please do -NOT- ask me "Will XXXX make it in?" because I
don't want to promise features that I haven't seen myself, and I don't
even know for sure what will be added in the next week.

Anyway, Imagine 2.0 should be released on November 15, (nine days!),
barring any calamity. Impulse will send out a newsletter to all
registered users, at which time you can send away for the upgrade. The
upgrade fee is a hefty $100, but includes all new software and a new
manual. Do NOT send in your money now; Impulse will return any checks
you send in before the official release. So wait for your newsletter
and/or more definative posts on the list before you send in your
money.

Again, a complete article explaining these changes will appear in this
month's AVID magazine, which should be at newsstands and computer
stores by Saturday.  I'll post AVID's phone number and address for
those who want to call them directly. (I heartily suggest you
subscribe! Imagine articles every month!)


-------------

A quick summary of the new features in Imagine 2.0 over Imagine 1.1:

I General

    2.0 interface look
    User defined gadgets in each editor for often used commands
    Preferences Editor added for editing config file and user gadgets
    Action Editor now independent from Stage
    ESC-escape picture viewing bug fixed (finally!)
    Infinite planes visible in preview windows
    All-new manual by Mike Halvorson ( > 200 pages )

II Renderer

   ``Quick Render'' option - scanline render whatever is shown in the
        preview window in any editor at any time
    Altitude mapping bug fixed: dramatic quality increase
    Direct DCTV support
    (No major rendering speed or RAM requirement changes either way)

III Detail Editor

    Brushmap axes default to span object
    Picked points can be interactively scaled, rotated, and moved
    New invert and genlock options on brushmaps
    ``Subgroups'' of an object define collections of faces for easy
       attribute definition; textures and brushmaps can be limited to only
       certain subgroups
    Phong shading selectable for independent edges
    Spline paths (editable in Detail) used for extrusion and for conforming
        objects
    Objects have a fog attribute- not just ``transparent edges'' but true
        light-absorbing volumes. Clouds, smoke, and fire are a snap, as well
        as spotlight beams and glowing warp drives.
    ``Cycle Transform'' command allows rotating cycle links around the ENDS
        of their axis
    Attribute loading now loads brushmaps and textures as well.

IV Forms Editor

    Major change: any number of key horizontal cross sections may be defined.
         ``Skin'' is now completely outclassed.
    Different options for radial cross sections: ``spacer view'' for
         simple skinned cross sections, or a two or four ``former view'' as
         in Imagine 1.1
    Symmetry in horizontal cross sections
    Automatic sealing of top and bottom points

V Stage & Action

    Camera tracking controllable from Stage
    Quickdraw of objects completely definable from Stage
    Multiple F/X applied to an object simultaneously
    Global fog (with altitude ranges)
    Backdrop pictures supported (animatable)
    World reflection map animatable

---------------------

Some of these features are really understated; The user gadgets are
terrific time savers, the fog attribute is absolutely incredible, and
the utility of interactive point manipulation is amazing. And Forms
just increased its abilities by an order of magnitude!

Most of the new toys are in the Forms and Detail editor, so you
modellers should have a field day!

-Steve

--------------------------------------------------------------------------
Steve Worley                                       [e-mail]
--------------------------------------------------------------------------

##

Subject: Reflection & Filter in scanline
Date: Wed, 6 Nov 91 10:28:03 PST
From: [e-mail] (Steve Worley)

> I seem to be having a rather frustrating time with filter brush maps
> in Imagine, and I wonder if anyone could help me out.  I know that
> my brush axis is placed correctly, 'cause I get a beautiful color
> map of my brush, but no transparency.  I have shinyness off, and
> have messed around with the filter values for the object, and got
> the results you'd expect if it was supposed to be a color map.
>
> Just to see if my technique is totally brain-dead, I tried
> reflection mapping, and got the same result - a rather nice
> color map, but no reflection.
>
> Is there some magic setting that I'm forgetting in attributes?
> Thank you much for your help


A couple of people have brought up some explanations that are
misleading.  A couple of facts: The perfect sphere will render as a
phong-smoothed but faced sphere unless you use trace mode, where it is
perfect. Don't know why... But as someone suggested, if you are using
scanline, you can use a primative with many points.  Also, the faces everyone
is talking about aren't a failure of the Phong shader: the place where
the faces are seen are right at the edge of the sphere. Phong doesn't
change the object's sillouette at all, it just smooths the face-to-face
edges.

Also, reflection and filter maps most definately work in scanline! There
will be no refraction of transparent objects, and the reflection will be
confined to showing grounds, sky,  and the world reflection map, but it
will certainly work.

If you are trying to use a relection/filter map on you object, make
sure your object has something to reflect,  and something behind it
so the transparency is shown. It seems like you certainly know
how to apply brushmaps (your colormap worked), so I can't see any
reason why the filter and reflection mapping would fail unless you
just didn't notice the effect. If you can change the reflection and
filter values of the object from the attribute requestor (with no
filter/ref maps) and you see the effects clearly, then there probably
is something wrong with the maps. This would be a good test to try.
If you still have problems, and standard reflection/filter work,
then I can suggest some other tests to see what is wrong.

You mention that you made sure shininess was off. Good check! But you
might double-check, since a shininess of even 1 will disable standard
reflection and filter.

Hope this helps.

-Steve

--------------------------------------------------------------------------
Steve Worley                                       [e-mail]
--------------------------------------------------------------------------

##

Subject: I need help with JPEG programs
Date:    Wed, 06 Nov 91 11:19 PST
From: Andy Kohler                          <[e-mail]>

I retrieved the NewAmyJPEG archive from Hubcap, extracted
cjpeg and djpeg, and found no documentation.  I assume
cjpeg is for compressing, and djpeg for decompressing; I got
djpeg to give me a template listing its options; but I don't
know what those options mean, and I've been unable to get it
to decompress the mr2nite.jpeg file (also from Hubcap).
When I tried:  djpeg mr2nite.jpeg,  it gave the dimensions of
the picture, then sent gibberish to the screen.  I tried
djpeg >mr2nite.i24 mr2nite.jpeg; it spent 15 minutes writing
a 1.1 MB file, but DCTV wouldn't load the file (saying bad iff file
I think).  I have an A2000HD, plenty of free space on the HD, and
4 MB fast, 1 MB chip memory.

Any and all advice will be greatly appreciated (but I won't send money).

Thanks.

Andy Kohler / [e-mail]

##

Subject: Re: I need help with JPEG programs
Date: Wed, 06 Nov 91 15:44:56 EST
From: Mark Thompson <[e-mail]>

> I retrieved the NewAmyJPEG archive from Hubcap.
> I tried djpeg >mr2nite.i24 mr2nite.jpeg; it spent 15 minutes writing
> a 1.1 MB file, but DCTV wouldn't load the file

That is correct, the problem is that the output of djpeg is a ppm file
as used in the pbmplus toolset. You will need PPMto24 which was included
with the augjpeg.lzh archive. You do not want to use the JPEG binaries
in that archive because the code is outdated and may not work with images
compressed with the latest code.

Here is the usage documentation included with the jpeg code:

On PC, Macintosh, and Amiga systems, you say:
        cjpeg [switches] imagefile jpegfile
        djpeg [switches] jpegfile  imagefile
i.e., both input and output files are named on the command line.  This style
is a little more foolproof, and it loses no functionality if you don't have
pipes.  You can get this style on Unix too, if you prefer, by defining
TWO_FILE_COMMANDLINE in jconfig.h or in the Makefile.  You MUST use this style
on any system that doesn't cope well with binary data fed through
stdin/stdout.

Currently supported image file formats include raw-format PPM, raw-format PGM
(for monochrome images), and GIF.  cjpeg recognizes the input image format
automatically, but you have to tell djpeg which format to generate.

The only JPEG file format currently supported is a raw JPEG data stream.
Unless modified, the programs use the JFIF conventions for variables left
unspecified by the JPEG standard.  (In particular, cjpeg generates a JFIF APP0
marker.)  Support for the JPEG-in-TIFF format will probably be added at some
future date.

The command line switches for cjpeg are:

        -I              Generate noninterleaved JPEG file (not yet supported).

        -Q quality      Scale quantization tables to adjust quality.
                        Quality is 0 (worst) to 100 (best); default is 75.
                        (See below for more info.)

        -a              Use arithmetic coding rather than Huffman coding.
                        (Not currently supported, see LEGAL ISSUES.)

        -o              Perform optimization of entropy encoding parameters.
                        Without this, default Huffman or arithmetic
                        parameters are used.  -o makes the JPEG file a tad
                        smaller, but compression uses much more memory.
                        Image quality is unaffected by -o.

        -d              Enable debug printout.  More -d's give more printout.

Typically you'd use -Q settings of 50 or 75 or so.  -Q 100 will generate a
quantization table of all 1's, meaning no quantization loss; then any
differences between input and output images are due to subsampling or to
roundoff error in the DCT or colorspace-conversion steps.  -Q values below 50
may be useful for making real small, low-quality images.  Try -Q 2 (or so) for
some amusing Cubist effects.  (Note that -Q values below about 25 generate
2-byte quantization tables, which are not decodable by pure baseline JPEG
decoders.  cjpeg emits a warning message when you give such a -Q value.)

The command line switches for djpeg are:

        -G              Select GIF output format (implies -q, with default
                        of 256 colors).

        -b              Perform cross-block smoothing.  This is quite
                        memory-intensive and only seems to improve the image
                        at very low quality settings (-Q 10 to 20 or so).

        -g              Force gray-scale output even if input is color.

        -q N            Quantize to N colors.

        -D              Use Floyd-Steinberg dithering in color quantization.

        -2              Use two-pass color quantization (not yet supported).

        -d              Enable debug printout.  More -d's give more printout.

Color quantization currently uses a rather shoddy algorithm (although it's not
so horrible when dithered).  Because of this, the GIF output mode is not
recommended in the current release, except for gray-scale output.  You can get
better results by applying ppmquant to the unquantized (PPM) output of djpeg,
then converting to GIF with ppmtogif.  We expect to provide a considerably
better quantization algorithm in a future release.

Note that djpeg *can* read noninterleaved JPEG files even though cjpeg can't
yet generate them.  For most applications this is a nonissue, since hardly
anybody seems to be using noninterleaved format.

On a non-virtual-memory machine, you may run out of memory if you use -I or -o
in cjpeg, or -q ... -2 in djpeg, or try to read an interlaced GIF file.  This
will be addressed eventually by replacing jvirtmem.c with something that uses
temporary files for large images.

Hope this helps.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re: I need help with JPEG programs
Date: Wed, 6 Nov 91 13:47:18 MST
From: [e-mail] (Marvin Landis)

Andy Kohler writes:

> I tried djpeg >mr2nite.i24 mr2nite.jpeg; it spent 15 minutes writing
> a 1.1 MB file, but DCTV wouldn't load the file (saying bad iff file
> I think).

This is the right idea, but djpeg does not create iff24 files, it creates
24 bit ppm files.  This is why DCTV thinks it is a bad iff file (it is! :-).
In the jpeg archive I downloaded from portal, the ppm to iff conversion program
was also included (so was the iff to ppm program).  I know the ppm programs
were on ab20 at one time, so check for them there if they are not on hubcap.

Marvin Landis
[e-mail]

##

Subject: no filtering
Date: Wed, 6 Nov 91 18:10:28 GMT-0500
From: Scott Matthew Krehbiel <[e-mail]>

This problem with the filtering is really strange... I have a
blue sphere behind a plane divided into 4 color sections
( white, black, red, and blue )
I would think that the white section would dissapear, giving me
a view of the blue sphere behind it.
By the way, this is in quarter-screen trace.

Could I possibly have the map as Color AND Filter by accident?
Maybe the colormap is outweighing the filtermap?
I am getting predictable results from playing with the
object's filter sliders, so I know that nothing is
interfering with standard transparency.

I'll have to get a scene from someone that does filter
properly, and let you all know what I find out!

Oh well... thanks for the help!

Scott
[e-mail]

##

Subject: Color vs. Reflect attribute
Date: Thu, 7 Nov 91 8:13:35 CST
From: camelot![e-mail]

        I'm a fairly new  Imagine user.  What is the difference between an
        objects color and the color it reflects?  I am under the impression
        that an object has no color of its own and merely appears to have a
        color due to the electro-magnetic frequencies that are "bounced"
        off of its surface and reach my eye.  Some frequencies get absorbed
        by the object; others bounce off.  Therefore, what definitions is
        Impulse assigning to these physical characteristics.  I'd like
        definitions for ALL of the attributes.  The manual was especially
        ambiguos about color and reflect.

        Thanks in advance.


        ____________________________^__________________________________

        dale r. rogers               Email:  ingr!b24a!camelot!dale
                                     Internet:
                                     ingr!b24a!camelot![e-mail]

##

Subject: Re: Reflectivity and other Attributes
Date: Thu, 7 Nov 91 10:44:06 -0500
From: [e-mail] (Michael B. Comet)

    In regards to attributes this may or may not help in
understanding what each exactly is, but it may help:

Color]  This is the overall color of your object.  If this is the only
thing that is set, then you object will be this color at the points
where light hit it the most, and then fade gradually to black where
there is no light hitting it.  eg to make a red ball, bring up the red
slider.

Reflect]  This is more similar to mirroring.  To make an object have no
color at all, bring up all of these sliders.  This means that even if
you had a red ball, bringing reflectivity to a max, would make you ball
totally mirror like.  NOTE: if you are only rendering say one object, on
a perfectly black sky, you will not see anything, since the object will
be prefectly reflecting black!.  Actually reflect values >150 may cause
an object to be invisible.  Also,  I would reccomend if you are trying
to understand these attributes such as reflectivity, that you do it in
Trace mode because scanline only approximates it.  eg: I spent over an
hour trying to figure out why this chrome ball I made wouldn't reflect
this primitive plane I had, then I did it in trace, and it was fine.

Filter]  This sets up which colors an object is to filter.  If you bring
up all three sliders, an object should be perfectly clear, and if you
bring them say around 3/4 of the way, it should be clear, but with some
'fogginess'.  This is usually used with the index of refraction value.
NOTE: filter will not work properly in scanline, AND it will not work at
all if the shininess value is anything but 0,0,0.

Specular]  This attribute sets up the color that a shiny object will
shine with.  eg,  usually objects just shine a bright white.  But you
may want the hotspot of an object to be a different color such as say
green.  In that case,  bring up the green specularity slider all of the
way.

Dithering] This simply sets up how smooth color transitions will be in
an object in ham mode and other modes other than 24 bit.  For example,
if you have a light illuminating a sphere from a side, you will see the
sphere will be whatever color it was set to on one side, going to dark
on part where no light hits it.  If there is a small dithering value,
then rather than seeing a smooth transition, you would see what looks
like bands or stripes, where as a higher value whould yield a smooth
fade across the object.  NOTE: they say that for 24 bit, objects should
have no dithering.  I have a firecracker, and have not seen this to be a
problem.  no matter whether the object has dithering or not, in 24 bit
it is always a smooth fade.

Hardness]  This really relates alot to the specular setting.  Let's say
you had the specular all the way up to white, and the color set to red
for a sphere.  You will now have a red ball and a white hotspot on the
ball.  The question becomes, how big is the hotspot?  Will it be a small
point, or greatly spread out over the object.  This is what hardness
sets, how the hotspot affects your object.  A small value will make the
hotspot spread out over most of the object.  A large value will make
small bright highlights.

Roughness]  This puts in a grainy look to an objects appearance making
it look rough.  This is good for making stones and rocks.  I have found
however that a little bit goes a long way.  eg: a high roughness can
make you object look like mush.

Shininess]  This basically sets up how shiny your object will be.  This
will make an object have a plastic or glossy look to it.  eg: you have
speular and hardness set, but now how shiny is the object as whole?
Higher values are more shiny.


    Well, that's about it.  I have to admit, it is not easy to
master these setting, I know I have not truly done so yet either.  Some
of it is very difficult, such as the fact the shininess can't be used
with filtering.  The best to way to learn attributes, is to experiment.

    Heres another idea I used to help me learn how attributes
interacted with each other.  If you have a fast enough computer, try
makeing small animation (100x100) or so in raytrace with a ground, a
gradient sky and an object such as a sphere.  Then save two spheres one
with a basic staring attribute, and another with the same attribute but
with one of the selections in a different position.  Then tween objects
say over 30 frame and play back then anim.  You should see how it
affects your object.  If you like the attrbutes for the objects satat
frame 14.  Goto it in the stage editor, and you can snapshot it to save
your object as it is with those attributes.  Then you can do some more
anim tests.


                Mike C.
                [e-mail]


will NOT work with the filtering setting.  An example is makeing a
plastic looking object.  Leave the specular alone, and bring up the
shininess.  This will create

##

Subject: RAM Limitations
Date: Thu, 07 Nov 91 11:10
From: [e-mail] (MOLDFIEL)

All too often I get the message "Not enough RAM" when attempting to render
a scene or animation.  My current setup includes 3MB of 16-bit RAM and the
wimpy 1/2 MB Fat Agnus.  My question is, which will Imagine most benefit
from, upgrading to the Fatter 1MB or 2MB Agnus or increasing to 5 or 9MB of
RAM?  Perhaps my question should really be phrased, in what situations does
Chip RAM become crucial to Imagine (i.e., brush maps)?

- Mark Oldfield

##

Subject: Re: Color vs. Reflect attribute
Date: Thu, 07 Nov 91 11:15:54 EST
From: Mark Thompson <[e-mail]>

dale r. rogers writes:
>What is the difference between an objects color and the color it reflects?

Well this is the second time I have heard this question in the past week or
so. Color is the object color and it affects the diffuse and ambient light
reflected from the surface. The easiest way to think of reflect is as
a settable color filter for environmental reflections. Basically it determines
what colors will be seen in the surface of the object as a result of
reflections from surrounding objects. Say for example you have a sphere
with the following surface attributes:
color: 255 100 10
specular: 0 200 100
reflect: 100 100 0
Around it are four objects, one each of red, green, blue, and white.  Assume
a white light source. The sphere will be an orange red, but it will have a
spectral highlight from the lightsource that is blueish green.  Also, there
will be a reflection of the four other objects, but the reflection will only
pass red and green light, therefore the red object will be red, the green
object will be green, the blue object will be invisible, and the white object
will be yellow. These will all be blended in with the surface color.
Transparency works in a very similar fashion.  It is important to remember
that Imagine DOES NOT behave as real light does so the methods in which
color, reflection, spectral highlights, and transparency are handled are all
approximations that are conveniant for the computer.

>Therefore, what definitions is
>Impulse assigning to these physical characteristics.

Since I didn't program Imagine (and I don't even use it for that matter),
the following may not be 100% accurate. However it is based on typical
computer graphic rendering implentations which Imagine basically follows.
The following Phong lighting model calculates intensities for each of R,G,
and B. (NOTE this has nothing to do with Phong normal interpolation or
smoothing)
                                                      n
 I = Ia x Ka + { Ip x Kd x (N * L) + Ip x Ks x (R * V)  } / d
                                     ^^^^^^^^^^^^^^^^^^specular
Ia: ambient light intensity
Ka: ambient light refectance coefficient
Ip: point light intensity
Kd: diffuse light refectance coefficient
Ks: specular light refectance coefficient
N:  suface normal
L:  light vector
R:  reflection vector
V:  viewer vector
n:  specular focusing coefficient
d:  distance attenuation (assumed to be 1 for infinite light sources)
*:  dot product

In the above equation, Ka and Kd are derrived from the surface Color setting.
Ks is from the Specular setting. Ip is your light color and intensity.
And n is a function of the hardness setting. Once the above calculations
are performed, transparency and reflection may be factored in. I cannot say
for certain how Imagine calculates these into the final color, but it
is approximately something akin to:

Final Intensity = (I * (1 - R) + (O * R)) * (1 - T) + (B * T)

I: intesity calculated above
R: object reflection coefficient
O: color of reflected object
T: transparency (filter) coefficient
B: color beyond this surface

This is also done for each of R,G, and B. Don't ask me how Shininess fits
into the equation because I am not entirely sure.
Hope this helps.
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
|      `       '        Mark Thompson                 CONCURRENT COMPUTER  |
| --==* RADIANT *==--   [e-mail]        Principal Graphics   |
|      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   |
|     Productions       [phone] [phone]   & General Nuisance   |
|                                                                          |
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Re: Color vs. Reflect attribute
Date: Thu, 07 Nov 91 13:16:51 EST
From: Mark Thompson <[e-mail]>

> Final Intensity = (I * (1 - R) + (O * R)) * (1 - T) + (B * T)

Ooops, I didn't mean to imply dot products here. It should be

Final Intensity = (I x (1 - R) + (O x R)) x (1 - T) + (B x T)
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~%
%      `       '        Mark Thompson                 CONCURRENT COMPUTER  %
% --==* RADIANT *==--   [e-mail]        Principal Graphics   %
%      ' Image `        ...!uunet!masscomp!mark       Hardware Architect   %
%     Productions       [phone] [phone]   & General Nuisance   %
%                                                                          %
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

##

Subject: Lighting tips
Date: Thu Nov 07 10:19:51 1991
From: [e-mail]

Can someone give some basic lighting layout settings, and even better
would be a detailed tutorial?  Thanks.

Scott

##

Subject: Re: Lighting tips
Date: Thu, 7 Nov 91 17:16:50 MST
From: HURTT CHRISTOPHER MICHAEL <[e-mail]>

    This was shone to me by Don Whitaker (a list member) and works
quite well for me. I know it's been explained before in the ImagineComp but
I wasn't able to grasp it until Don showed me at a meeting. So here goes my
try.
    A good way to get even light is to have 2 lights. One pre-dominant
light and another fill light. The settings I use are 300 300 300, and 75 75
75. Varying either way depending on the overall scale of the scene. Both
lights I use are spherical, with the brighter one casting shadows, and the
lighter one (sometimes both) with diminshed intensity. Try setting up your
lights similar to this: (crude ascii follows)

____________________________________
|              top                 |
|                                  |  [] = box obj
|                                  |  O  = camera tracked to obj
|              []                  | /
|                                  |   * = bright light
|                  \   *           |   @ = fill light (dimmer)
|       @           O              |
|__________________________________|________________________________________
|                                  |                                       |
|                   O  *           |                     *O                |
|                  /               |                     /                 |
|       @      []                  |                 []   @                |
|                                  |                                       |
|                                  |                                       |
|                                  |                                       |
|__________________________________|_______________________________________|
        front                   right

    The basic thing is just to have 2 (or more) lights, where one light
is bright enough to cast some shadows and get some shading going, while
having another light that fills in the other areas enough so that your entire
object is lit.
    Try this out and let me know if it works for you! Happy rendering!

    Chris


##

Subject: more 'bout lights
Date: Fri,  8 Nov 91 08:04:52 PST
From: [e-mail]

setting ambient light to high will wash out shadows, you're right.  setting
at 100 is a good place to start (I didn't mean that would be where you would
nessisarily keep it. 5-20 seems a little too low...try turning down your
lights and rendering with ambient set that low and you probably won't see
anything which defeats the purpose of ambient light.  I ussually end up
setting ambient around 40 or 50 on the highest gun.  I also vary the color
depending upon the scene.  I was basically trying to give a setting that
would show something and give a good starting point...it is imposible to say
what the exact setting should be without know the scene or the mood that is
intended and as far as that goes what I see as one mood someone else will see
as a different mood.  Anyhow it is harder to completely wash out the entire
frame because of too much light than it is to be completely black so I always
set the guns high and then lower them.  Also a tip for everyone...keep a
notebook and write down the settings you like and you can always refer back
to it when you are doing a similar scene.
RMills

##

Subject: Re: RAM Limitations
Date: Fri, 08 Nov 91 16:06:50 CST
From: [e-mail] (Michael Linton)

[e-mail] (MOLDFIEL) writes:

> All too often I get the message "Not enough RAM" when attempting to render
> a scene or animation.  My current setup includes 3MB of 16-bit RAM and the
> wimpy 1/2 MB Fat Agnus.  My question is, which will Imagine most benefit
> from, upgrading to the Fatter 1MB or 2MB Agnus or increasing to 5 or 9MB of
> RAM?  Perhaps my question should really be phrased, in what situations does
> Chip RAM become crucial to Imagine (i.e., brush maps)?
>
What you need to buy is more fast ram.  Imagine, like most programs, uses fast
ram first, and then uses chip ram later.  It dosn't matter how much ram you
have, you still never have enough.  I have 8 megs of fast ram, and I sometimes
run out.  A Fatter Agnus chip is still a nice thing to have.  Imagine only uses
chip ram when your fast ram becomes very low.  I have had my fast ram down to
4 K while my chip ram was still at 1 meg (I have 2).  More fast ram is
certainly a necessity when render large files, and playing back large anims.

Hope this helps.


--- (Michael Linton) a user of sys6626, running waffle 1.64
E-mail: [e-mail]
system 6626: 63 point west drive, winnipeg manitoba canada R3T 5G8
system 6626: 63 point west drive, winnipeg manitoba canada R3T 5G8