May 24, 1995
12 people in the audience have done 100 CDS in last year
James Watson
Senior Analyst
Doculabs - comparing CD Authoring - in newsletter
Craig Hansen
Gen Manager of Microboards
since 1980
in CD 1985 CD Interactive
Pete Jenny, Dataware Technologies
CD
Royce White
President dataDisc - 5 yrs.
all had booths
Not specifically on products..
ISSUES:
Frame of reference on applications.
We tend to forget about business problem tie back to problem.
Try to ask about applications.
The focus is going to be on the market itself on where it is going, inhibitors, future tech, standards
Where will CDR be SIGCAT 98
Questions to address:
effect of falling pricing - how change prices
What if MS introduces a product?
Label printers
Media writer compatibility
Any other questions you might have.
Hot in here. A bit burnt out.
As James started out with the chart here.
We wanted to talk about market share, pricing, units sold
75% have recorders in here
in marketplace about 75 000 units world wide
1987 first ones - people paid 6 figures for their systems
projection by next year 200 000 additional units in market
26 million readers - will double in next 5 years
overall growing dramatically, in terms of recorders, readers and titles available
Why developing?
A dozen different applications
The one that will drive some of the most
archiving backup - will become bigger as price drops down
a number of people buying for regular computer type purposes - prototyping , publishing one-offs, distributions - light replication
applications wise - every market. Every body. Some markets haven’t been touched.
Medical first
Legal coming in fast
Insurance
just about every industry as you look down list
every agency
every association
someone in each industry
doesn't matter what you are doing
class here showed mm aspects.
text retrieval on market
adobe stuff
facilitating it
almost as fast as coming up to speed, something else is coming up too.
And - 5000 pieces of media under $7 a piece -
premastering has started to crumble
Pete Jenny Dataware - low cost software provider -
what effects on market?
Well, CDR software has taken turns over several months.
$2 500 in 1989
remained consistent - 1/2 around 1993
recently hit the floor on the street $120 - 150 for recordable software
many following suit, others not
The trajectory of a falling rock
Hardware pricing is about matching that also
$120-150 000 in late 1989 heck of unit - worth every penny paid for it.
Mid year - under $1 000 complete with bundled software
trend down. floor is $2-300 bucks on level of current readers
12 premastering vendors now
if this much fall - fallout of companies - how to select vendor -
Basically, attrition on developer side. The current crop of software covers about everything you need to do - only certain CD formats out there. Not likely to change drastically. Not need for new software dev. Along those lines, people chugging along. New hot high tech development.
How about CD erasable and its impact.
CD erasable is different market.
with established base of CD rom - another medium - a different market. A back up point of view. Technology is not there. Won't be there for several years. Dana Parker did good job on standards, where they stand. The whole CD Erasable , where is market for it. Sticking to archiving. The whole legal market does not want erasable.
Because we like true archival we won’t see erasable.
Q: Why not as a replacement of the floppy?
Why Not?
NeXt had 260MB disc. You could take disc around. Erasable would fall right in there.
Solid hard case. $10 unit.
When standard hard disk is much more
Q: Standard, standard, standard
Quality quality quality
ISO between platforms, don't work on DOS, where are we going?
Will there be.
Incompatibility issue?
Platform independence?
Writer/media incompatibility - study done by JPL that SIGCAT alerted us all to, writers, media, make matrix, compared results of different media over years. Mike Martin. Jason worked with Mike, Real open question if problem or not. Those who write.
What will be done.
Groups going out to test. Doculabs. will do it again. 3x versus 4X. 3-4 of us at optical storage technology association - spearhead phase change - discussing same problems
Japanese contingent - signing discs have different effects. Method to allow writer to determine what kind of media - adjust write strategy.
So open issues there. As an industry, make it as easy as possible All agree.
Q: I've run into problems. Investigating. Not sure where going. There are problems especially with certain readers. Very infrequent. On chatter box of Internet.
Pete: question is also kind of leading to cross platform / Cross OS, but maybe file system problems. Disc issue is one problem. OS file system in CD-R ISO 9660 compatible across
Rockridge extension, POSIX files embedded. Apple has extensions - file type creator IDS all within ISO 9660 so if you have ISO Rockridge in Mac you would see icons - they look like they should.
With long file names. Truly a cross platform disk.. Those extensions don’t intrude on file system itself. Cray could still pull it up. OS/2 NT no problem. Underlying ISO structure is all maintained correctly.
Andy Young revising ISO 9660 spec to get all these things
Andy: Not doing these things on my own. As the project editor of ISO project. Things being addressed. See me.
Q: Look at industry growing so fast, trying to stay on top. Standards needed to be addressed. It's huge. As a result a number of issues being addressed. People involved, not that big. More as Japanese started pouring more into it, get involved. That's part of what SIGCAT is all about.
Add to what Royce is saying. small group - one half dozen mfgs producing media. 5 actual vendors producing the recorders. Small number of companies that are mfg. the product. Acceptance in marketplace is phenomenal. Dataquest over last couple years - we are about 3x best case projection. We do have quite a bit. Most getting involved at earlier stage - get compatibility between vendors. Avoid market fragmentation. Make sure everybody maintains compatibility across, neutral bodies raising compliance issues. How to take the tech forward in cohesive, reasonable manner.
When some think rapidly, others - media mfgs are disappointed - lack of a killer app. Is it going to backup. Historically we haven’t had that. onsey twosey format. Won't make Kodak plant burst at the seams.
Scanner data at Grocery store, data publisher. probably 1-1/2 to 2 years ago. Proctor and Gamble rep puts data on CDR - 2 500 discs per month. Like MCI, Sprint - maybe a killer app - mainframe app printer replacement.
What James has been describing is data on demand. Allow people to generate discs in almost real time - distribute in small quantities. Examples, besides IRI, hardware vendors who update apps. Applications, particularly in UNIX - 12 or 15 quarter inch tapes. A neat app - a menu, all the components on a CD, customer name. pushes a button, off it goes.
Are we there yet?
Yes. The killer app does this on a large scale.
Is everybody familiar with print on demand. One big hard drive, print 20 brochures as needed.
I think that's what killer app is going to be. Need to go further with writable jukeboxes. Software vendors around working on storage management.
Xerox docutech?
I like analogy between traditional offset - this organization has been about title productions
everything about printing has gone to printing on demand. More personalization. This is the killer app potentially.
COLD - computer output to laser disc - from database to optical disc. Tradtional optical jukeboxes.
componetnst to production CDR creation - printing and labeling the discs.
Royce: Don't use your pens. Only one that works.
Early years horror stories who used magic marker - all of a su
Sharpie is the only pen you can get away with. See anyone else using it - call a postal worker. ...
Sorry
One of
Doughnut labels. - full, quaRTER - HALF HAVE NOT WORKED WELL.
Those I would only use fo rmedium term. under 10 years. We don't know what will happen with glue and disc.
Fromagte - 11 seoncd transfer $1- 3,4000. If you do a lot of labeling from pagemaker, WP whatever 300 dpi work well
Couple folks still doing silk screen with higher numbers.
100 - 500 can be very good. One side has data - other has label. Next technology - two sided discs.
Any quesitons on printing?
Long term requirement 30 years.
Want to put data online. How does label hold up in jukebox.
Is jukebox with caddies? Non caddy? I would recommend putting label on the caddy.
Caddyless. Put information written on the inner hub. You won't touchi it. Full doughnut on the hub. This isdics 27. leave it in there.
Potential inside jukebox of heat, above room temperature?
Label can curl.
Print write on disc with printable media - avoids pealing. Barcoded media.
Marks discs correctly. Every disc has a serial number on it. Just use serial numbers.
If you aren't looking at them For anything that is long - no study on how long it lasts
Media is good for 84 years + - 25.
Wer are going to see 100 years plus from this stuff.
With labeling, don't count on it.
COLD - twice a year.
Ninetrack needs refresh.
What if new media comes along - writing to new media
120 terrabyt CDs - every thing on one single ROM
With printing infor -think of app - with pen good for 1 year
120 platter jukebox. None of those are labels. On the disc is the label. read the header electronically. No one ever opens juke box. Just line up jukeboxes, leave them.
3 vendors sell printabel media - Ricoh, ...
A number, 3M coming on line. A nubmer
goes back to a very TDK , Toyota Uki,
only like 4 vendors from OEM point of view. Others may private label.
From mfg
We'll estimate market.
As teh amount of recorders go out. Media consumed continutes to increase, more vendors entering the market.
or one thins. Rimage printer. Does not require printable media - will print directly on to blank media.
Q: Plastic sleeves? Jewel cases NON
No cases, studies. They stick. Univenture wanted to avoid. In indiana, pattented, claim to have addressed that issue with
STatic does not effect that. Vinyl you worry about static. Won't hurt the media.
Potential of market is so huge, taking over market. Problem will go away as fast as it arises.
Next hot one - high sppeed writing - over @X - Kodak's 6X writer.
What will 12X be all about?
Very fast. Next step will be 8X. Hopefully better drive mangement. Current problems.
8 - 10 people have 4X writers.
File fragmentation problems.
Writers deal with input stream. Writing faster is not really a reality.
Newer drives wil have larger buffers, stream it out difectly.
8X an d 10X will do a bit fo us. It is changes that are much less.
10X 6 minutes.
4X 11 minutes. Differnce is not greate
difference 1X to 2X 60 to 30 minutes, very drastic. Higher and higher speeds, difference is minimized. I'm not sure what that's going to do to eveyone's ambitions.
RAID would not be ISO 9660. Would not help much due to latency.
In speciic 6X series at single spped. duplication systems.
4 four times recording - 16 in an hour off the hard drive.
Disc to disc - premastering was done with Corel - your player has to be running twice as fast as ther recorder.
So at least the software wil allow you to do that.
CopyWriter 2X reader, 1X writer.
Huge companies, called 3M. Can't send out secte material.
Number of solution. 600 top secret. Don't want to take couple weeks - 16 per hour. Not too bad.
6X writers, pair. lights out 75 in 10 hour period. Kodak trnaporter. Two ganged together. Data/Ware.
There are anumber of those tings. Upwards of 100,000 to get what you need.
Suppose you copy 600 discs. How do you know they are good?
Basically if you are looking at a numbe of things. Use a facility.. If you dn't have an option.wait.
In terms of integrity, testing media - most will do bit verification. In terms of actually testing data - form 10 to 50,000 that do significatn error checking - slow. So, it doesn't meet your need at the moment.
When you look at writers on the market. In Kodak 6 speed, more diagonstics - read during write. pit verificatin, not exactly full verification, but helpful.
You could test that equiment is
Double verification, cut then independent reand and verify - true disc quality checkres. AT some point you need to say that you have the quality you need.
Will depend on specificc needs. Quality managemnt will set spec.
Process will be the same you use in Microfilm.
Q: back to idea of quality and the check . In bit for bit. different platform, it cant' read the data. Unrecoverable errors, tracking problems.
Hardware, softwrae.
This is an answer I'd like to hear. Where do you find, when you get it back. Today, where today are you finding the probjems. - hardware, software or media.
None of those three - how user used system.
Media tends to be good. Will continue to be lots of bad media. Unfortunate, but it wil occur.
Very rare.
More likely, people just getting involved, 100 people - how many coasters - toward front end of expericene.
We've been finding that recent runs of writers are much much better.
3rd genreation of recording equiment. Anyone who has old equipment knows. Looser toleratnces, media problmes.
Old days, more than a year ago. Hardware and media interface problems. Tryping to understand how to write in these enviromnet.
Everyone has gotten much, much much better. Issues were in configuration, system maintenance, DOS file fragmentation - peopl know about this now.
It takes a long time for the read heads to go up and pick up data. On small buffere system that amount of lag time underflos the channel.
90% end user. Buffer underrun due to wrong SCSI card. Once multisession, education, young industry. Good books, when buying or using . Here is my system. What do I need. What bus do i need. If I'm running an ISA - can I put my scnnaer on that too?
Tell what system you have. Help you walk through learning the system. More instatn gratification, less reading of manuals.
Reading about new Phillips dreive - incremetnatl writing mode.
Follow up to verification - people get confused about
physical
logical
E32 errors - doesn't meean andything
Logical - does it run, does it work.
100% verification - never done in the past.
Not that big a problem, getting better.
Expericen with Phillips writer, some won't read a video clip. Some won't read the media at all.
The players are all different.
Older players are problem. Bad laser. Low powered laser.
Writer is 522 - shouldn't have that problem.
WIth 521.
There is a problem with older drives reading CDR - mfg have higher reflectiveity. Trying to pass through an optical dye, not on pressed media. Silver discs are much brighrter than blue.
Older ones want 70% - expecting 80% reflect. CDR only delivers 70% it comes under the threshold of what old reader was tuned for.
Older than last year.
Table questio non multi-packet writing.
4 people know aobut this
ways to write
Disc it once methong - all done at same time - lead in , lead out
Track it once methond - leaving lead in lead out ready - fixating
Packet writing - has been CDR noise for 3 years - writes out cluster of data and a little extra info and where it will write to next - like a floppy on a CLD media Both good and bad.
If you are familair with 9660 directory is at the begining. If you are going to writer there several times, and its not rewriteable you need new file system. Then you have to make sure it will be backwards compatible with older discs.
13490 - file system designed to handle this stuff - a ways away.
SIGCAT 1998
How many pieces of media wil be sold in 1997, how many write3rs how many reaers
Royce:
Craig said - 1 year - 750000 to 1.5 million CDR in next 365 days.
Say just - another REAL SERIOUS installed base. Kick off
Writers
1997 - between 2 - 3 million
Media 30 - 80 million pieces.
Dave:
I would agree with media figure - 30 to 80 - I like a nice 50 million unit range.
As far as drive - about half of what Royce said. I don't think mfg capacity is there
35000 to 750000.
=
200000 I gave. More toward 360 000 We are component contrained.
Demand is there but compnents - units ahve been shrinking dramatically.
1988 2 big boxes. Now 1/2 height. There will be some constraints.
360 000
Media - 50 million swing is right. More towards 80 million.
=
I am going to 3.5 jmillion writers. Everyone wil need a new one.
65 million pieces.
Corel, peter hall
COREL CD CREATOR
Source from CD, WAv or Music Cache.
Choose tracks, clickadd now. Layout happens in backgound.
Want to create now?
When first installed, tests are run - part of CD Creatore. Check CD ROM - for audio discs. Check out hard disk for IBM think pad, many IDE drives, so we can't write 4x speed.
Check out recorder for you. Checks out scsi bus termination. Simulated recording, 20MB. Finds errors, explaines them. ONce intot actual audio. Move around.. Edit wav file, edit, bring it back in. Special things, like merge two tracks. While effecting - jewel case is being created for us. Contents is what is interesting. Authomatically generated for you. Track 4 is two tracks - make this Demo track . When changed here, chages ther. Reorders, etc.
Other things you wil ant to do. Spec paper color. So you do more defintions. And you can insert object s iwth OLE two - maybe you have Corel Draw. What that will do is add a copy of that. OLE2.
It will check registry to offer these to you.. Let's bring in graphics here.
Idea was not to reinvent the wheel. OK to resize, repostio for jewel case insert.
Creates a spine for you. Any windows printer. A pair of scissorss.
Yellow cue card - if they fall out of Wizard, tell you what to do. Use drag and drop. Put into wav file editor. If I do Wav file into audio editor, it will come up as well - written out as Red Book audio - so CD Creator does resampling for you to convert to Red Book audio.
One more thing. Take advante of CD Creator as an engine. OLE2 - write your own app, VB drives it. Here is an example PCD Creator - pictures come from PhotoCD, you pick pohots, make couple prints. Take out that disc. Get some from another disc.
PCD Creatore knows how to display, reorder, undre file menu creates CD in background. So you can write these VB assps with manyl
Even under Excell spreadshett version 5.
What is price?
You'll see it for 159, 169 mfg is 249 - price is really low. Drive the markte to put CD on
Multisession?
Yes.
In fact. You can do a CDPLus disc.
=
Dean
Moniker
used to be with Meridian Data
with sony electronic pub
Spira - unleashing CDR into mainstram.
look at CDR as a giant floppy disc
what you saw - lot to learn,
What moniker did was to look, not at publishign model. Most at highlevel technolgoist.
What about myself or your secratry that just want s to back up on CDr.
Data storage model Really only drive letter, drag and drop access to CDR. Take from any drie letter, including your server. Drag and drop - all premastereing takes place in abackground. No serious thinking to doo. Takes 300 K floppy and you are off and running..
No devloper or consultancy integration problem.
All indexing, problems solves can benefit from easy to use product like spira.
Takes advantage of CDR
50 million people using this. Psychology of CD RO is a winner.
You can create a mm applicatin. Take it and drop it on our drive letet and go hoem. The saseier it will happen, the better.
Multisesion is great. Archival is great. Mnay peolpe will go from publishing to just backing up their hard disc. Drag and drop and walk away. Manual is 24 pages if you need to look at it.
you can stop session, any ISO 9660 player. STart next session.
How much overhead. Orange Book has 15 MB. 40 seesions per disc. Size - smallest is 15 MB.
At 40 you got nothing in there. Save a drawing - make sure important drawing is not on magentic media
249 list
brochrues New company hope to distribute
Certainly no other product with our approach. User friendly.
Windows only. Windows level dreivers.
Works with new Rich
Sony 920s
Latest generation of multisession
We will have Windows 95 product.