My problem is that I remember more or less 2 specific CDXL videos that was available "bb.xl" and "enya.xl", which was a Blues Brother Movie Trailer and a music video from ENYA called "Caribbean Blue" on the CD I got back in early 1994. Since I have huge problems figuring out what happened to my Amiga CD32 after 1996/1997 or so (reference my very own 4-hour long video), its very hard to check the contents of my own CD. Thanks to the collectors community, people have preserved the following known TOSEC versions:
The Demo Disc: Commodore Amiga CD32 Demonstration Disc 1.1 r72 1993
TOSEC: Commodore/Amiga CD32/Multimedia/[IMG]/Demo Disc, The - Commodore Amiga CD32 Demonstration Disc 1.1 r72 (1993)(Commodore).img
The Demo Disc: Commodore Amiga CD32 Demonstration Disc 2.0 1994
TOSEC: Commodore/Amiga CD32/Multimedia/[IMG]/Demo Disc, The - Commodore Amiga CD32 Demonstration Disc 2.0 (1994)(Commodore).img
Now, this gives us at least some clues for the version and year released. But, I went much deeper and dug out the actual meta data of the CD itself when it was produced. Finding the ISO image is very easy online, but scans or photos of the actual CD is very limited. I have at least added the best version of them below.
Internally, these demo cd's by Commdore was nicknamed "CATS", and this was also revealed during the Amiga Workbench leak some years ago, where we could find "amiga os source code 3.1\os-source\aug.cats\". There seems to be reference online to "CD-ROM produced by Commodore-Amiga Inc. labelled CATS Developer CD V1.0 (CATSCD10) - This was probably distributed to registered Amiga developers circa 1991." .
CATS = Commodore-Amiga Technical Support. It was a support/engineering liaison group inside Commodore International.
So, the demo cd's started their life as developer / experiment / trade show / for early preview to magazine and other persons intitially, before it apparently made it into a final release that got mastered and delivered with the official Amiga CD32 packages. Problem is that I strongly believe that at least the first 5 or 10 thousands of Amiga CD32 produced had V1.0 of this demo cd. When, we will dissect actual data found, labels, serial numbers and on CD - it is very clear that something does not add up. V2.0 for instance could only be delivered with Amiga CD32 for the 2k or 3k last batch of machines ever produced - and I have got the proof logically and digitally to prove this point once and for all!
Evidence #1 - Physical unit's serial number / label and CD Metadata
Now, my own Amiga CD32 was manufactured in September 1993, as per my own serial number:
"851 39 93 513515 00 8 055591 - Manufactured September 1993".
If we decrypt this specifically: Week 39 in 1993 = between 27 September 1993 > 3 October 1993. We can only assume that "055591" is the actual physical count.
and them my re-bought Amiga CD32 from ebay in July 2024, had
"851 35 93 365158 00 1 017929 - Manufactured August 1993".
If we decrypt this specifically: Week 35 in 1993 = between 30 August 1993 > 5 September 1993. We can then also assume that "017929" is the actual physical count.
It was said that Commodore only produced about 80000 or 100000 pcs until April (March!*) 1994, where the Philippines factory (Automated Technology Inc. ("ATI"), 3/F, 5600 Osmeña Hwy, Makati, 1235 Metro Manila, Philippines" shut down and Commodore US went backrupt, thus production was stopped and rumors told that the Philippines goverment prohibited export of these to USA before the bill was paid.
*From cdvw103.guide about the Philippines factory:
CSM Speaks - The Compuserve Conference On May 1, 1994, Mike Levin of the Commodore Stockholder's Movement appeared on Compuserve to address the future of Commodore and the Amiga.
"(1-29,Mike Levin) The factory in the Philipines has not produced since March"
Now, keep these numbers in mind when we reveal the actual CD production date burned:
V1.1: "CATS DEMO 1.1 rev 72" - 24 Sep 1993 @ 16:01:55
V2.0: "CATS Demo Version 2.23b" - 23 Dec 1993 @ 11:02:56
Now, lets break down the estimated production start for my known 2 units in life:
Unit Manufactured 30 Aug 1993 > 5 Sep 1993 (Labeled with August 1993) - Week 35 = 017929 (my current physical CD32)
Unit Manufactured 27 Sep 1993 > 3 Oct 1993 (Labeled with September 1993) - Week 39 = 055591 (my lost and original CD32 bought 994).
Lets add a friend of mine's very early CD32 to the mix:
8513293365101002009385 = Week 32-93 = 9 Aug 1993 to 15 Aug 1993 - Serial 009385.
And lets add other EBAY's people's CD32's S/N to really mish-mash remix this up in the calculations below:
8513293365158001004996 = Week 32-93 = 9 Aug 1993 to 15 Aug 1993 - Serial 004996
8513493365158001011649 = Week 34-93 = 23 Aug 1993 to 29 Aug 1993 - Serial 011649
8513793365158001032064 = Week 37-93 = 13 Sep 1993 to 19 Sep 1993 - Serial 032064
8513893365158001040624 = Week 38-93 = 20 Sep 1993 to 26 Sep 1993 - Serial 040624
8513993365158001056195 = Week 39-93 = 27 Sep 1993 to 3 Oct 1993 - Serial 056195
8514393365158001086961 = Week 43-93 = 25 Oct 1993 to 31 Oct 1993 - Serial 086961
8514893365158001100972 = Week 48-93 = 29 Nov 1993 to 5 Dec 1993 - Serial 100972
8515093365158003103856 = Week 50-93 = 13 Dec 1993 to 19 Dec 1993 - Serial 103856
8510394365158003114156 = Week 03-94 = 17 Jan 1994 to 23 Jan 1994 - Serial 114156
8510494365158002119160 = Week 04-94 = 24 Jan 1994 to 30 Jan 1994 - Serial 119160
8510594365158003121791 = Week 05-94 = 31 Jan 1994 to 6 Feb 1994 - Serial 121791
Serial System makes sense
Now, lets do all of them listed on this page, and we can clearly see the production flow.8513293365158001004996 = Week 32-93 = 9 Aug 1993 to 15 Aug 1993 - Serial 004996
8513293365101002009385 = Week 32-93 = 9 Aug 1993 to 15 Aug 1993 - Serial 009385
8513493365158001011649 = Week 34-93 = 23 Aug 1993 to 29 Aug 1993 - Serial 011649
8513593365158001017929 = Week 35-93 = 30 Aug 1993 to 5 Sep 1993 - Serial 017929
8513793365158001032064 = Week 37-93 = 13 Sep 1993 to 19 Sep 1993 - Serial 032064
8513893365158001040624 = Week 38-93 = 20 Sep 1993 to 26 Sep 1993 - Serial 040624
8513993513515008055591 = Week 39-93 = 27 Sep 1993 to 3 Oct 1993 - Serial 055591
8513993365158001056195 = Week 39-93 = 27 Sep 1993 to 3 Oct 1993 - Serial 056195
8514393365158001086961 = Week 43-93 = 25 Oct 1993 to 31 Oct 1993 - Serial 086961
8514893365158001100972 = Week 48-93 = 29 Nov 1993 to 5 Dec 1993 - Serial 100972
8515093365158003103856 = Week 50-93 = 13 Dec 1993 to 19 Dec 1993 - Serial 103856
8510394365158003114156 = Week 03-94 = 17 Jan 1994 to 23 Jan 1994 - Serial 114156
8510494365158002119160 = Week 04-94 = 24 Jan 1994 to 30 Jan 1994 - Serial 119160
8510594365158003121791 = Week 05-94 = 31 Jan 1994 to 6 Feb 1994 - Serial 121791
Logic dictates then for a potentional V1.0 demo cd:
If Commodore manufactured 10000 pr. week, starting in Week 32, do we really think that they kept xxx units on stock until Week 38 and then slipped V1.1 demo disc in it - and then shipped them? Of course not!
We have proof that S/N 4996 was produced in Week 32, so there must have been a Demo CD V1.0 that lasted for a many weeks. How come nobody even in 2026 can find this item?
The CD32 Demo Disc v1.1 was mastered on September 24, 1993 (Week 38), at which point approximately 40,000-45,000 CD32 units had already been produced. This indicates that earlier production runs - potentially up to 30,000 units - must have shipped with an earlier disc version, likely corresponding to the elusive v1.0.
Shocker: Or did Commodore produce and sell 40-50k units WITHOUT a DemoCD?
- Commodore sometimes swapped discs at distributors in the aftermath?
- Unsold stock may have been retrofitted with 1.1?
It was said officially that Commodore rolled out 10000 pcs by the week (estimating 2 work shifts), so well, what about:
Production flow calculation - If we roll the counter forwards and estimate:
Week 32: ~9,000 (confirmed by early unit with serial 9385)
Week 33: ~11,000 (estimated)
Week 34: ~14,000 (estimated)
Week 35: ~18,000 (matches within my known unit)
Week 36: ~27,000 (estimated)
Week 37: ~36,000 (estimated)
Week 38: ~46,000 (estimated)
Week 39: ~55,600 (matches within my known unit)
Week 40: ~70,000 (estimated)
Week 41: ~78,000 (estimated)
Week 42: ~82,000 (estimated)
Week 43: ~86,000 (estimated)
Week 44: ~88,000 (estimated)
Week 45: ~91,000 (estimated)
Week 46: ~94,000 (estimated)
Week 47: ~97,000 (estimated)
Week 48: ~100,000 (estimated)
Week 49: ~102,000 (estimated)
Week 50: ~104,000 (estimated)
Week 51: ~107,000 (estimated)
Week 52: ~110,000 (estimated)
Week 03-94: ~114,000 (estimated)
Week 04-94: ~119,000 (estimated)
Week 05-94: ~122,000 (estimated)
Mass production must have started in ISO Week 32 or possibly late Week 31 of 1993.
That corresponds to:
Week 32: 9-15 August 1993
Week 33: 16-22 August 1993
Week 34: 23-29 August 1993
Week 35: 30 August - 5 September 1993
Week 36: 6-12 September 1993
Week 37: 13-19 September 1993
Week 38: 20-26 September 1993
Week 39: 27 September - 3 October 1993
Week 40: 4-10 October 1993
Week 41: 11-17 October 1993
Week 42: 18-24 October 1993
Week 43: 25-31 October 1993
Week 44: 1-7 November 1993
Week 45: 8-14 November 1993
Week 46: 15-21 November 1993
Week 47: 22-28 November 1993
Week 48: 29 November - 5 December 1993
Week 49: 6-12 December 1993
Week 50: 13-19 December 1993
Week 51: 20-26 December 1993
Week 52: 27 December 1993 - 2 January 1994
Week 03-94: 17-23 January 1994
Week 04-94: 24-30 January 1994
Week 05-94: 31 January - 6 February 1994.
Well, a strong evidence that numbers out there of "80000" or "100000" is completely bullshit. The last unit 0594 serial number was seen on actual photo on a recent EBAY link in March 2026. So, for all your Commodore freaks out there - the actual offical statement of "how many CD32's produced" are _OVER_ 122000 units!
Mix that with statements found also on internet that "Phil factory has not produced since March 1994", we can most likely add a couple of thousands more.
Although Commodore documentation lists CD32 production capacity at approximately 5,000 units per week, this figure reflects single-shift operation. Serial number analysis indicates that output during peak weeks approached 9,000-10,000 units per week, strongly suggesting that the same facility was operating in multiple shifts during the initial production ramp.
So 5K/Week @ 1 shift = Knight & Day 2 shifts. Those workers went on day and night for us and Commodore :-)
Evidence #2 - Norwegian Computer Magazine Article - October 1993
In AmigaForum magazine issue #6, released in October 1993, the main author of the magazine (Øyvind Skogvoll) pre-reviewed an Amiga CD32 unit with a 2-digit serial number received from Commodore Computers, Oslo in order to create an article.
It has several clues on the contents of the demo disc that he had access to, and in fact, my screenshots of the article below (4 pcs), truly reveals I found digital script traces and word to support this on Demo Disc v1.1 (which means Commodore did not clean properly and left clues for me to find. It truly confirms there MUST be an earlier Demo Disc around, either with the first 1000's of Amiga CD32 sold - or in packages and boxes of previous Commodore distributors, garage's or somebody's huge lot of aquired Commodore office related materials, discs and CD's - somewhere in the world. Or is it truly LOST MEDIA by now?
This CD must be found and preserved, nobody knows how many where produced, how many actually made it to the public, or if any of the early previews managed to make a backup - or even kept the disc's after they sent the lend Amiga CD32 back to the Commodore bosses?
Evidence #3 - amiga os source code 3.1 (leaked version from 2016)
Since I now have specific words and filenames I can trace, there are an "aug.cats" directory that contains further clues.
bb.xl / Blues Brother trailer traces:
amiga os source code 3.1\os-source\aug.cats\wayne\gameselect\scripts\trio_cdxl
amiga os source code 3.1\os-source\aug.cats\wayne\gameselect\scripts\loop_cdxl
amiga os source code 3.1\os-source\aug.cats\wayne\gameselect\scripts\duo_cdxl
amiga os source code 3.1\os-source\aug.cats\wayne\gameselect\scripts\bluesbrothers_cdxl
All of these files contain the following text/script:
Bin:cdgsxl Demos:xl/bb.xl BACK Demos:xl/bb.back X 79 Y 53 XLMODEID NOPOINTER XLSPEED 150 ENDDELAY 120 FIREABORT DEFMON
Jurassic, Enya, Jon Bon Jovi references inside:
amiga os source code 3.1\os-source\aug.cats\wayne\gameselect\scripts\gameselect.config
contains:
GAME Robocod=execute >NIL: Scripts:dostartup CD0:Robocod
GAME Pinball=execute >NIL: Scripts:Choose_Pinball
GAME D/Generation=execute >NIL: Scripts:dostartup CD0:DGen
GAME Sleepwalker=execute >NIL: Scripts:dostartup CD0:Sleep
GAME Zool (PAL only)=execute >NIL: Scripts:Dostartup CD0:Zool
GAME Whale's Voyage=execute >NIL: Scripts:Dostartup CD0:WCD
@CDXL Creation=execute >NIL: Scripts:Creation_CDXL
@CDXL Jurrasic Park=execute >NIL: Scripts:JPSmall_CDXL
@CDXL Action!=execute >NIL: Scripts:Movement_CDXL
@CDXL Blue/Blade=execute >NIL: Scripts:Duo_CDXL
;@CDXL Enya=execute >NIL: Scripts:Enya_CDXL
;@CDXL Blues Brothers=execute >NIL: Scripts:BluesBrothers_CDXL
;@CDXL Bladerunner=execute >NIL: Scripts:Bladerunner_CDXL
@~MPEG BON JOVI=execute >NIL: Scripts:BON_JOVI_MPEG
DEMO Flight=execute >NIL: Scripts:startflight
@DEMO Graphics=execute >NIL: Scripts:ShowPictures.script
@DEMO About Amiga CD=execute >NIL: Scripts:dostartup CD0:Sachs
I'm pretty sure I had once read an article about the Amiga CD32 launch that mentions "Jon Bon Jovi" at some point, so, even there we have evidence of an alternative version of the V1.0 demo cd as well!
Evidence #5 - Mandrill texture mapped chunky pixels demo
Since I had already found this demo on V1.1, called "gamedemo.ld", I started hunting a bit more for the binary. This is what I found about it:This file is based on a various compilation of "TMapDemo.lha", whuch are out in the wild. and coded by "Chris Green - Graphics Software Engineer - chrisg@commodore.COM".
"GameDemo - texture mapping and Gouraud shading Demo by Chris Green"
The date seems to be around 18 Feb 1993.
The "TMapDemo" works on Amiga 1200 and has some FPS counter as well.
Get it on Aminet
Conclusion - and what now?
I think I have enough evidence with facts and physical traces that there must be an earlier or even several earlier Demo Disc somewhere that pre-dates the known V1.1 today - and people needs to start digging after it - it's not an Mandela Effect yet, but we have to move fast!What now?
Well, I ask you to simply dig out your Amiga CD32 and check every cd you have and see if you find it. Chances are that the one you find is the only one left ON THIS PLANET - and we need to preserve it at once :-)
Contact me at:
post@stone-oakvalley-studios.com
Any updates to this mega-hunt will be posted here - if any.
Update #1: Hi-jacked myself into a similar request on EAB forums
Update #2: Entry posted on Games That Weren't website.
Historic mentions online:
#1: Interresting post found on EAB Forums It mentions: music videos off the Commodore CD32 demo CD
(They had a demo CD with 3 music videos on it: Bon Jovi's "Blaze of Glory", one by Enya, and Seal's "Killer")
#2: From within legacy Google Groups posts, I find additional proof of "Enya", "Blues Brothers" and "Seal: Killer" presentations.
Link 1
Link 2
Link 3 (German, but easy to Google translate.
Most telling is this statement:
You obviously only saw the beta demo CD. The final version now released by CATS only has some recreational videos from sports and such as CDXL demos. No well-known movies. Probably there were license problems.
So, basically, I believe we're are really looking for "Commodore Amiga CD32 Demonstration BETA Disc" or an extremely rare delivery of this on the very first batch of Amiga CD32?
Anyhow, the reality of such CD's existing are at this point EXTREMELY RARE to get hold of.
Pending solution: A CD version v43 exists - smoking gun!
See my two screenshots below.



















