HPAVC/HPAVC CD-ROM.iso/DA2WAV1G.ZIP/CDDA.DOC

Text File | 1994-10-25 | 35KB | 747 lines

CDDA Version 1.0g - This is the fourth public release of CDDA.

**** NOTE ****
 Release 1.0g may very well be the last update of CDDA.  My free
time for working on this is just about gone.  Unless I hear about
a major bug which can be easily fixed, or someone sends me new
programming information for a drive, I don't expect there will be
any new versions soon.

CDDA - What is it?
------------------

In February 1993, I needed to get some audio samples into digital
form on my PC.  Toshiba had just released the 3401 CDROM drive
which allowed reading of DA frames across the SCSI bus.  Because
there were no programs available to do this, I had to write my
own.  I used it to get the samples, and then left it alone because
I had no idea anyone else wanted to do this.  Some months later
CDGRAB came out, but they wanted much too much money for their
program, so I cleaned up the program and released it as CDDA09A.ZIP.

Note:
Some time later I heard from a very irate person who told me I
couldn't use the name CDDAxxx.zip because he was already using
the name.  I've looked everywhere, used ARCHIE, and asked people
who seem to know every program on the net, and no one has seen
this guy's program.  However, in the interest of keeping peace on
the net, I have changed the distribution file name to DA2WAV1G.ZIP.
The program itself will still be called CDDA.

The first release fully supported the Toshiba 3401 CD-ROM drive.  I have
since added Sony, NEC, Chinon, Plextor, Panasonic, Teac, Hitachi and
MSCDEX only. Unfortunately I don't have a Sony/NEC or drivers that
allow the MSCDEX part to work.  I have given prerelease versions of
my program to people who have to some degree had success with other
drives.  I have no guarantees that this program will work with your drive.

Since the release of 1.0a I have heard from people who were having
trouble with the NEC 74-1/84-1 drives.  So, I went out and borrowed
one for a few days.  To my surprise the 84-1 didn't work.  The next
thing I did was get the data sheets for the drive sent to me from
the NEC faxback.  There is no mention of being able to read DA frames
on that data sheet.  Next, I reread the NEC programming manual.  It
says about the 74-1/84-1 drives "Read CD-DA is under consideration of
content in the SCSI support commnds".  This suggests that it may
or may not be supported.  Next I called NEC tech support.  They told
me that "reading digital audio through the SCSI port is NOT supported
on the 74-1/84-1 drives.  It is supported on the 3x and 4x drives."
So, for the time being, I will leave in the code to support the 74-1/84-1
drives.  If I don't hear from someone who gets it to work in the
next while, I will remove the support and mention it in this file. I have
also heard from someone who has a NEC 211 drive.  I have no idea what
this drive is other than it is a 2X drive.  This person has had good
success running CDDA.  I have also heard from someone with NEC 210 who
has not had success.  If anyone knows more about programming for NEC
drives please let me know.

Different ROM revisions cause the drives to do things differently.  One
version may work just fine and others might not.  It is very tough for
me to help with drives that don't work, when I don't have one of
those drives to work with.  If your drive revision is on the bad list,
I don't know what to tell you.  Perhaps you can make a really big
stink with the manufacturer and have them upgrade your rom.
Here is a list of known good and bad rom revisions.  I will add to this
list as I receive reports from the field.

good                  bad
----                  ---
Sony 561 rev 1.9a     Sony 561 rev 1.7x
                      Sony 561 rev 1.8f
Toshiba rev 0283      Toshiba rev 2732
                      Toshiba rev 3593
NEC 211 rev 1.0       NEC 84-1 rev 1.0

Because the Toshiba/Sony/etc. drives need proprietary SCSI commands
to read the audio sectors, I chose to talk to the drive through the ASPI
interface.  I hope that most people with SCSI drives have gone the
ASPI route.

What drives support DA?
-----------------------

I have heard from many different people who say such-and-such drive
will read DA frames.  The following is a list of which drives which
I understand have the ability to read DA frames: (thanks to
[e-mail] for starting this list)

Apple 300, 300i, 300e (Sony CDU-8003, CDU-8003A)
Apple CD300+
Chinon CDS-535
NEC CDR 200, 300, 400, 500, 600, 900 series
NEC CDR 84-1/74-1 (with special firmware) (yet to be verified)
Panasonic CR-562B/563B
Plextor (formerly TEXEL) DM-3028, DM-5028
Sony CDU-561, CDU-31A, CDU-33A, CDU-55S
Toshiba XM3301 (Silicon Graphics)
Toshiba XM3401, XM4101, XM3501
Teac 4x
Hitachi
Matshita (Panasonic) 503

You may want to run CDROMINF.EXE and have a look at the output.  If
there no mention of being able to read RAW frames, and/or the
RAW READ SIZE is not 2352, there is no chance that CDDA will work
using the MSCDEX interface.  Also, if it reports that your drive
supports reading RAW, that DOES NOT mean that CDDA will work.  It
all depends upon the person who wrote the driver for your drive.  If
he/she passes DA frames to MSCDEX, you're laughing.  Otherwise,
you're out of luck.

There are rumors that the Mitsumi LU055 drive will work using the
MSCDEX interface, but I have yet to hear specifically what version
drive, and what version driver works.

If you know or think you know of other drives, let me know and I will
do some investigating.

How do I use this program?
--------------------------

This is a command line utility which allows the user to specify the start
and end points of the data transfer and the output type.  The start and
end points may be entered in one of three modes,: LBA, MSF and track.
Logical Block Address mode is the number of the frame from the start of
the disc. Minute, Second, Frame mode specifies the time from the start
of the disc in actual time.  Track mode allows you to dump an entire track
(or song) to disk.  The LBA and MSF are related by the following formula:
    LBA = Minutes * 60 * 75 + Seconds * 75 + Frames - 150
The lead in track is usually 150 because there are usually 2 empty
seconds at the start of a CD, but it can also be any number +- 75 frames
from 150.  A commmon number other than 150 is often 182 or 183.
The Toshiba, NEC and MSCDEX programming manuals define the LBA equation
to be as above with -150 at the end.  The Sony manual is very vague,
and it could be interpreted as either -150 or as -(lead in track).
Since two of the three manuals say clearly -150, I have chosen
-150 for the Sony as well.  I managed to do a fair bit of testing on
a Sony 561 drive, and it appears that the -150 figure is correct.
If anyone has more precise information on the Sony please pass it along.

The /ID option allows you to manually set the SCSI ID of the CDROM
drive you want to use.  This is useful if my software incorrectly
guesses your drive's ID, or if you have more than one CDROM drive
in your system.

The /MSC option allows you to manually set the MSCDEX drive letter of
the CDROM drive you want to use.  This is useful if my software incorrectly
guesses your drive's ID, or if you have more than one CDROM drive
in your system.  This option does not force the software into
the /M option.

The /M option forces the software to use only MSCDEX commands.  This
may or may not work with your drive.  If you don't have a SCSI CDROM
drive with ASPI drivers, this is your only hope.  This option would
be used with the Panasonic CR-562/563, Sony CDU-31A/33A drives.

The /NJC option forces the software to not do any jitter correction
to the data read from the CD.  This means that the program will run
faster, but that the files created cannot be guaranteed to be 100%
perfect copies.

For the options /S /E /F which expect a following parameter, remember that
there needs to be a space between the letter and the parameter.  ie if you
enter /S01:10:10, it would come up as an error, but if you entered
/S 01:10:10, it would be correct.  For the MSF mode, there need to
be 2 digits for each of minutes, seconds and frames or it will come up as
an error.  When entering the filename with the /F, don't include an
extension, as the software automagically adds the extension for you.

At present there are only two file formats supported, WAVE and Binary.
There is also the option to dump the data in HEX format out to the
STDOUT device.

The binary format has the samples stored in the order Left LSB, Left MSB,
Right LSB, Right MSB.  The samples are 16 bit 44.1 KHz stereo.

I am not expecting to add options to output files in 8-bit or 22.05 KHz.
This is not as simple as just throwing away samples, as this causes
aliasing in the output files.  Correcting this is way beyond the
scope of this program.

I am not planning to add many more formats because there
are plenty of other programs out there that will do the conversions much
better than I can.  Besides, the whole purpose of this program was to
get the raw data out to the hard disk, not duplicate SOX.

On every CD there is a bit which defines if copying a particular track is
permitted or prohibited.  CDDA checks this bit and will not continue
with the dump to disk.  I have included an override option /O which forces
the user to explicitly specify and to knowingly copy a copy prohibited
song.  Have look at the file ROYALTY included in this distribution,
which was honourably pinched from the program CDGRAB.  It is a list
of most country's contacts for paying royalties.

One thing to remember is that this audio fills up the hard disk fast.  It
takes between 9 and 10 megabytes per minute of music.  Because of this I
put in a check to make sure that you will have enough space to put the
requested samples.  As well there is the /U option which will give you
an estimate of how much disk space will be used without actually dumping
the data to disk.

Examples
--------

CDDA /MSF /S 10:14:36 /E 13:55:11 /F outfile /W
will dump using MSF mode from 10:14:36 to 13:55:11 to a WAV file

CDDA /LBA /S 106232 /E 109443 /F outfile /B
will dump using MSF mode from 106232 to 109443 to a Binary file

CDDA /T 2 /F outfile /W /O
will dump the entire track 2 to a WAV file with the override mode on

CDDA /T 2 /U
will display the estimated disk space for all of track 2

CDDA /T 2 /H
will dump the entire track 2 in HEX to STDOUT

CDDA /T 2 /M /F outfile /W /O
will dump the entire track 2 using MSCDEX interface only

CDDA /T 2 /MSC G /ID 4 /F outfile /W /O
will dump the entire track 2 from MSCDEX drive G, SCSI ID 4 CDROM drive


What is this jitter business?
-----------------------------

The following is a post made over a year and a half ago which discusses
the technical reason for a CDROM drives difficulty in accurately
positioning itself on an audio CD.


--------------------------------------------------------------------------
Newsgroups: alt.cd-rom,aus.cdrom
From: [e-mail] (Adrie Koolen)
Subject: Re: Reading Audio CDs - Why is it so complicated?
Keywords: cd, cd-rom, cdrom, audio, sampling, naivety
Organization: Philips Consumer Electronics, Eindhoven, The Netherlands
Date: Fri, 19 Feb 1993 08:23:53 GMT
Lines: 63

In article <[e-mail]> [e-mail] (David Elliott) writes:
>In article <[e-mail]>, [e-mail] (Diemo Schwarz) writes:
>|> What I don't understand is:
>|> Why should it be so difficult to extract pure audio data from a CD?
>
>It isn't.  The problem is that the SCSI controllers in most CD-ROM
>drives simply don't support it.
>
>|> The medium is structured by tracks and indices, isn't it?
>
>Yes.

Note that the `tracks' on a CD aren't layed out like tracks on a hard disk.
The track and index are just numbers that are stored with each sector on
the CD (in the Q-subchannel packet). The start address of a track is stored
in the Table Of Contents on the CD, indices are not. Indices are normally
not used, except index 0 for the 2 seconds pause at the start of each
track.

The REAL structure of the CD medium is a large continuous spiral, starting
at the center of the CD (at 46mm diameter). Finding a specific sector is
not as trivial as it is on a hard disk. You'll have to employ a kind of
binary search algorithm. Finding the start position of a specific index
is even more difficult as you don't even know where it starts.

>|> And every CD player somehow manages to ship the data from the disc to
>|> the D/A converter, doesn't it?

But at the DAC, they don't know which sector the sample came from or what
the relative position the sample takes in a sector.

>|> So, what's the difference between reading data CDs and music CDs?
>
>The format of the data, to put it simply.

Well, most decoders used in audio CD players, output a stream of samples
and sub-channel data, mostly used for their time code. As the decoder has
to adjust the spindle speed, it uses a FIFO to store the data. If the
FIFO fills up, the spindle motor is slowed down, if it gets empty, the
motor is sped up. The sub-channel data normally doesn't pass through the
FIFO, but goes directly to a microcontroller. That way, the microcontroller
can't be sure that it knows the exact address of the data, coming out of
the FIFO.

To circumvent this problem, the address of a sector is also stored in the
data itself, together with a 12 bytes sync pattern. This way, one can
determine the starting of a sector and its address just by looking at the
data coming out of the FIFO.

Here's the real problem: audio sectors don't have this sync pattern, nor
the address of the sector, in the data area of a sector. Most CDROM drives
use chips that are coming from audio players. The first CDROM drives were
just modified audio CD players.

To read audio sectors on a CDROM drive, you'll need a special decoder or
you'll need to connect a standard decoder with some custom hardware to
generate pulses to indicate the start of a sector and to synchronize the
sub-channel packets with the real 2352 bytes of data.

I hope that this explains it a bit.

Adrie Koolen ([e-mail])
Philips Consumer Electronics, Eindhoven, the Netherlands
--------------------------------------------------------------------------


Because of this, most every CDROM drive that can read DA frames cannot
accurately return to the exact same location on the disc every time.
It will usually return within a few samples either way.  It took
a lot of time and thought to write the code to correct for the
"jitter" efficiently.  My first try brought the transfer rate
to a crawl.  I have been refining the process to the point that
it barely has any computing overhead, but still I have to read most
every frame twice.  I am still working on getting this to work
faster.  Some people have reported an 8x slowdown from 0.9a to
1.0a.  I have no explaination for that large a difference.
It is nowhere near that large on my machine, as I see only about 3x.
I do agree that it is still much too slow. I have made a number of
improvements in 1.0e and 1.0g that help with this slow down including making
the receive buffer as large as possible (64K) when using real mode drivers
like ASPI and MSCDEX.  I have also added the /NJC option for those who
cannot wait for their files to be created, or for those whose drives
are generating "Jitter greater than 5 frames" errors.  This error
means that the software has become lost on the CD because the drive
was unable to return to within 5 frames of its last postion.



What kind of help is there?
---------------------------

There is very limited help from the command line for CDDA.  You can get
the "USAGE" by just typing CDDA.e.g.:

C:>CDDA

Usage: CDDA /(mode) [/U,H,B,W,M] [O] /S <start> /E <end> /F filename
           [/ID <cd-rom scsi id>][/MSC <mscdex drive letter>].
modes are LBA, MSF, T
/MSF - times in minute, second, frame format (MM:SS:FF)
/LBA - times in Logical Block Address format (xxxxxx)
/T   - send whole track to file
/S   - Start time
/E   - End time
/U   - estimated disk usage required for data
/H   - hex dump of sectors to stdout
/B   - write to file in Binary format
/W   - write to file in WAV format
/O   - override copy protection bit
/M   - override ASPI interface and use MSCDEX
/NJC - turn off jitter correction
/MSC - override MSCDEX find first CD-ROM
/ID  - override ASPI find first CD-ROM
Binary mode extension CDA will automagically added to the filename
Wave format extension WAV will automagically added to the filename
e.g. CDDA /MSF /S 10:14:36 /E 13:55:11 /F outfile /W
e.g. CDDA /LBA /S 106232 /E 109443 /F outfile /B
e.g. CDDA /T 2 /F outfile /W /O


Copying and Registration (does he mean this costs $$$?)
-------------------------------------------------------

I am allowing full freedom to copy this program.
It IS a fully functional version.
It is NOT Crippleware!
There is NOT a pro version available for an unreasonable sum of money.
I'm NOT asking for $ to register this program.

What I am asking is, that if you use this program to let me know (via EMAIL,
snail mail, Carrier Pigeon etc.), AND if your karma so directs you to make
a small donation to a local charity.  So far since the first release in
late '93, only a couple people have shown they have good karma.  I guess
this makes a rather sad statement about society.

I would also like to try and keep all the files together and unchanged
in the archive, so if you are passing this around don't change the files.
In this distribution the files contained are:

CDDA.EXE - the program itself
CDDA.DOC - this document file
ROYALTY - the list of people to whom you should pay royalties
CDROMINF.EXE - the MSCDEX exerciser program
CDROMINF.DOC - the text file describing CDROMINF
MSCTEST.EXE - the test program for MSCDEX only access reading DA
TOSHTEST.EXE - the test program for TOSHIBA drives reading DA
TOSH_RST.EXE - the program to reset TOSHIBA drives after fatal errors
NECTEST.EXE - the test program for NEC drives reading DA
SONYTEST.EXE - the test program for SONY drives reading DA
CHINTEST.EXE - the test program for CHINON drives reading DA
PLEXTEST.EXE - the test program for PLEXTOR drives reading DA
SCSIPING.EXE - the program which uses ASPI to search for SCSI drives

If you are one of those Shareware houses, I don't want to see this program
available for $6.99 plus shipping and handling.  I don't like seeing
the average Joe getting burned for a 10 cent disk, 1 cent label and 2 cents
worth of labour to make the disk, especially when he can get 600+ Meg
of really good stuff from something like the Simtel CD for $25 (thanks to
Robert Bruce for starting the cheap archive business).


How to get hold of me
---------------------
If you want to mail me about problems or to tell me I'm going to rot in
HE double hockey sticks for helping to copy music, put your message in
a file and copy it to the NUL device. :>

If you want to make suggestions or want to send compliments you can contact
me at the address below.

If you encounter trouble with the program CDDA, try and run the included
test program most suited to your drive.  There are four test programs:
one for each of the Toshiba, Sony and NEC type drives, and one which
is a MSCDEX only version.  These programs are scaled down versions of
CDDA and use the command line to specify the drive letter and ID.
The programs TOSHTEST, SONYTEST and NEC test take a single command
line argument.  It is -n, where n is the SCSI id of the drive.
eg. c:>toshtest -3
This will run TOSHTEST on the SCSI drive id 3.

There is also the program SCSIPING.  It searchs for are reports all devices
connected to the SCSI controller.  It uses the ASPI interface.

Finally there is a new program called TOSH_RST.  It is used to reset
a Toshiba drive after CDDA has stopped with a fatal error.  It uses
the same command line argument as the TOSHTEST program.


If you would like assistance in solving problems, please include all
dumps from all the related programs.  You should be able to
run the programs like this and get the dumps in a text file:
CDDA /T 2 /W /O /F testfile >dumpfile.txt

I have real difficulty in tracking down problems without these
dumps.  In the past the majority of people asking for help just send
along a short note telling me my program doesn't work, and what
they remember the error messages saying.  This just doesn't cut it.
I have to have the dumps to help.

The bottom line - RTFM before writing me.  Also read the alt.cd-rom
FAQ.  many of your questions can be answered by reading.
EMAIL: [e-mail]

Snail Mail:
       Jim McLaughlin
       449 Viewmount Dr.
       Nepean, Ont.
       Canada
       K2E 7P1


Things on my wish list of new functions
---------------------------------------

I want to add support for other CDROM drives that can read DA sectors
over the SCSI, but I don't have access to these drives.  If someone could
supply me with the programming information, I could try and do up a new
version of CDDA, if I ever get some spare time.

Aren't there other programs which do the same thing?
----------------------------------------------------

Yes.  There are several other programs out there which do some or all
of what CDDA does.  In the spirit of the Internet, here are all (I think)
the other programs which read DA frames.

CDDA.ZIP - ftp.cdrom.com - contact: [e-mail] (my old version)
DA2WAV1A.ZIP - ftp.cdrom.com contact: [e-mail] (my new version)
CDGRAB32.ZIP - ftp.cdrom.com - contact: [e-mail]
CDAR010.ZIP - wuarchive.wustl.edu - contact: [e-mail]
NECTOWAV.ZIP - ftp.cdrom.com - contact: [e-mail]
CDINFO12.ZIP - archive.utoronto.edu - contact: [e-mail]
CDDAREAD.ZIP - archive.utoronto.edu - contact: [e-mail]
??????.??? - posted to alt.cd-rom - contact [e-mail]
CDDA2WAV.ZIP - ftp.cdrom.com - contact: [e-mail]
READCDA.ZIP - ftp.cdrom.com - contact: [e-mail]

Questions I have received since the release of 0.9a and 1.0a
------------------------------------------------------------

Why is the program so slow?
This program is so slow because of the following:
1. My poor programming skills.
2. Many of the CDROM drives only read DA frames in 1X mode even though
   they may transfer data at 2X or 3X.
3. The amount of computing to do jitter correction.
4. Reordering the bytes from the CD to the WAV file.
5. Having to read 16 frames for every 10 actually written to disk to
   allow for up to three frames of jitter.
6. When there is more than three frames of jitter, I slightly shift
   the aim of the search and reread the problem frames.
7. My Toshiba seems to have a majic number of 10 frames, that if I
   read more than that at a time it takes 2-3 times longer than
   reading in bunches of 10 or less.  I don't understand it.  Since I
   have to read ahead 4 frames and behind 4 frames every time I read
   a block from the drive to correct for jitter, I can't very well
   keep my reads under 11 frames.  Other drives may have similar
   thresholds.


What is ASPI?
ASPI stands for Advanced SCSI Programming Interface.  It was developed
by Adaptec, and has been adopted by many SCSI card manufacturers as
a standard SCSI programming interface.  If you want more information
on ASPI call Adaptec [phone].

When I use /T, the start and ends of the track are a couple of
seconds away from where I think they should be.  Why is this?
In version 0.9a, there was a deep rooted bug which caused this.  I
have fixed it (I hope), and this should no longer be a problem.

When I play back my samples, why are there pops and clicks?
These pops and clicks can come from two places:
1. version 0.9a did NOT handle the jitter problem, it is handled now.
2. your sound card can't handle 16-bit 44.1 KHz samples.  Try using SOX
   or a similar program to change the sample to 22.05 or 11.025 KHz
   samples and try again.

The program stops with the error xxx.  What is wrong?
The 0.9a release had very limited error reporting.  I thought that
if it worked on my machine it work on everybody's.  This version
has much more error reporting.  As I understand it, CDDA will
not necessarily work correctly if SMARTDRV is installed.  This has
been the source of a lot of trouble.  As well some people have been
having trouble using CDDA in a DOS window in Windows.  I think
there may be a problem when being used with a DPMI manager installed.
So, the bottom line is DON'T run CDDA under Windows.

When will you release an OS/2 version?
I have no interest in writing a version for OS/2. Period.  However,
There is a program from Adaptec called VASPI, which will allow
you to run CDDA in a DOS window.  I got the following email with
the how-to for OS/2 from Stefan Eichner ([e-mail]):

> 1. Your CONFIG.SYS has to include the follow lines:
> [...]
> BASEDEV=TMV1SCSI.ADD
> BASEDEV=OS2ASPI.DMD
> REM DEVICE=C:\OS2\OS2CDROM.DMD /Q
> E=C:\OS2\MDOS\VASPI.SYS
> [...]
> !If no OS2CDROM.DMD is loaded you can't read Data-CDs!
>
> 2. Add to the autoexec.bat the start-comand for MSCDEX2.EXE
> (e.g. C:\MSCDEX2.EXE /D:MSCD001 /L:H /m:10 /v)
>
> 3. Change the settings of the DOS-Session on the WPS:
> DOS_DEVICE=C:\TSLRCDR.SYS /d:mscd001 /p:3
>
> Now CDDA works great!!!
VASPIBET.ZIP is available from ftp.cdrom.com

When will you release a Windows version?
I have no interest in writing a Windows version.  I don't have the time,
and I don't think Windows is a particularly good OS (if you can call
it an OS).  I have had a couple of offers to help, but I have been
too busy with the rest of my life to work on it.  Don't hold your
breath.

When will you release a UNIX version?
Never.  Although I run LINUX on my machine, I have no interest in
rewriting CDDA.

When will you be writing a version for my CDROM drive?
Getting the programming information from CDROM drive manufacturers is
like pulling teeth.  I've been hunting for this stuff for almost two
years, and have only managed to get the programming manuals for
Toshiba, Chinon and Sony.  If you have a drive and it does support DA frames,
I need the programming manual.  I also need about 5 or 6 extra hours
in each day.

I have a Toshiba drive.  Why doesn't your program work?
It appears that different ROM revisions cause the drives to work
differently.  For instance, my Toshiba 3401 is revision 0283.  I have
heard from people who have ROMs older and newer that can't use
my program.  I have also heard that Toshiba has changed the programming
interface on newer drives, but it is not reflected in their SCSI
programming manual.  I just don't know what is wrong.

Why doesn't your program work with my ASPI driver?
Several vendors supply ASPI drivers for their SCSI cards.  However,
all drivers are not created equal.  I have heard from people who
have had trouble with just about every SCSI card on the planet.  I
only have Adaptec's EZSCSI to do my testing.  I'll try and do my
best on case by case basis, but I really don't have any spare time.

CDROMINF works on my machine, so why doesn't CDDA?
CDROMINF only uses MSCDEX calls, and CDDA also uses ASPI calls.  It
is these ASPI calls that do the meat of the program.  If ASPI isn't
on your machine, or doesn't work correctly, then CDDA will never work.
Your only other choice is to try the /M option and use only
MSCDEX commands.  This is unlikely to help any SCSI drive owners.
Before writing me and asking for your drive to be supported, do a
little legwork first.  Call the manufacturer and ask for the programming
information.  Second, run all the test programs included with this
release.  These programs don't check for the drive type before trying
to read DA frames.  If any of these pass, let me know and I can
add your drive based on the programming information of an existing
drive.  It has happened before, so it might happen again.


Why is there tons of zeros at the beginning and end of my track?
Most every audio track has some silence at the beginning and
end.  This silence is actually digital silence which is all
zeros for the samples.

Can I use CDDA to read from DAT drives?
As far as I know there are only a couple of DAT drives out there
that allow reading of digital audio through SCSI.  In theory, I could
add that feature to CDDA, but since I don't have one of those drive,
and I don't have the programming manuals either, I seems very unlikely
that this will ever happen.

Sometimes CDDA doesn't work when I first put in a CD. Why?
I don't know.  Some people have reported that CDDA works better after
a new CD is put in the drive if you use some other utility to play and
then stop the CD.  This seems to force some sanity into some drivers.

Can you change CDDA to read the song titles from my audio CD?
There are no song titles recorded on an audio CD.  The following
is a post which explains this better than I can.


----------------------------------------------------------------------------
From: [e-mail] (Dave Platt)
Newsgroups: alt.cd-rom
Subject: Re: Ascii Track data on Audio CDs
Keywords: Audio CD Ascii Track
Date: 23 Feb 93 19:16:03 GMT
Organization: New Technologies Group, Inc.  Palo Alto CA
Lines: 41

>A friend told me that encoded on Audio CDs are the track titles, CD title,
>credits and even a bitmap image of the cover art.  Can anyone confirm or
>deny this?  If it is true, how does a programmer get to this information?

In general, this is _not_ true.

The disc table-of-contents has information about the total disc length,
number of tracks, and the starting time of each track.  It has no room
for human-readable text and there is no provision in the standards for
such, as far as I've ever been able to find out.

The audio tracks have "subcode" information associated with each frame
(75 frames per second).  The P subcode bit is used for primitive
track-signalling flags, and is rarely interpreted by modern CD players.
The Q subcode contains some useful information: control (type,
preemphasis, and copy-protect fields), address information, index
numbers, and sometimes the disc catalog/barcode number or the ISRC code
(serial number, country, year, owner ID) of the recording.  Still, no
room for text or artwork.

The R,S,T,U,V,W bits in the subcode can be used in a variety of ways.
The most popular de facto standard is for CD+G graphics.  These provide
a sort of slow-scan video (288x192 pixels, max of 4096 colors using an
encoded color lookup table).  The CD+G graphics can include almost
anything... lyrics, still photos from concert footage, and copies of the
cover artwork are all quite possible and have been done.

Only a relatively small percentage of audio CDs have CD+G graphics.

Some CD players have a "subcode out" jack which can be fed to an
external decoder.

Some CD-ROM drives allow the subcode data to be read over the SCSI (or
other) communication bus while the drive is playing audio CDs through
its built-in DACs.  Some [fewer] drives allow both the audio data and
the subcode to be retrieved over the bus and manipulated by the host
computer.  Details vary;  see your drive reference manual.
--
Dave Platt                                                VOICE: [phone]
              Domain: [e-mail]      UUCP: ...netcomsv!ntg!dplatt
 USNAIL: New Technologies Group Inc. 2470 Embarcardero Way, Palo Alto CA 94303

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

Changes in version 1.0g from version 1.0e
(version 1.0f was not formally released)

- ISRC code display added
- option NJC added
- Hitachi, Teac, Matshita drive types added
- bug in MSCTEST concerning empty buffers fixed

Changes in version 1.0e from version 1.0a
(versions 1.0b to 1.0d were not formally released)

- UPC code display added for NEC drives
- Toshiba mode select to return the drive to normal operation after
  reading DA frames had a conflict with EZSCSI drivers - this has
  been fixed
- length of WAV file did not always match length contained in header -
  this has been fixed
- added support for the Chinon CDS-535 drive (since I don't have one
  of these drives, this is untested)
- added support for the Plextor 3028/5028 drives (since I don't have one
  of these drives, this is untested)
- there was a bug which on some machines would allocate memory wrong
  when using the /M option, and would fill the file with zeros - this
  has been fixed
- program TOSH_RST was added for users of the Toshiba drives who
  have CDDA exit with a major error.  It resets the drive back to
  normal operation
- there was a bug on some machines that would allow the user to run
  CDDA fine, but the respective test program would fail.  It appeared
  mostly on the SONYTEST and NECTEST programs - this has been fixed
- during running with NEC drives extra debug messages were printed -
  this has been fixed
- the jitter correction routine has been rewritten - CDDA now runs
  about twice as fast as 1.0a did on my machine
- a problem with one version of the Sony rom has a peculiar failure
  which may be corrected by a rewrite of the sector search routine
- a problem with copy permitted discs sometimes required /O in order
  to dump to disk
- when the last track of a disc or the last LBA or the last MSF was
  selected to be dumped, some drives actually only allow reading
  of the second to last frame, and not the last frame.  since there
  is virtually no sound on the last frame of the disc, I have forced
  the program to stop at the next to last frame

Known bugs:
- someone has reported the program failing on a very large drive
  with very large amounts of disk space available.  The specific
  example was a 250M drive with 235M free.  It would appear that
  there is a bug in the Borland dfree routines.  I will be looking
  into the problem, except for the fact that my drive is only
  100M, so I can't duplicate it.
- a couple of people have reported that the WAV files created are
  not compatible.  I can't say much except that in one case the
  file size written in the WAV header was incorrect, and I don't know
  why.  I have however found that really large WAV files (>10M) often
  confuse some players.  I have seen a file work correctly on a couple
  of players, and not on others.  My guess is that some players out
  there are not interpreting WAV files correctly, or are not expecting
  very large files.

Wish list: (which may or may not get implemented)
- option to play selected range before saving to disk
- have start and end times include offset from beginning of track
- more CD-ROM drives supported