As far as i know, CQM is not a fake OPL3. Nor is it fake FM, it is just a different form of modulated synthesis. Now i do prefer OPL vs CQM by far and i would even put ESS FM above CQM. In fact and as far as FM is concerned i even like OPP better than OPL2 for instance. Too bad that it was probably considered overkill by AdLib. OPL2 was simpler and cheaper.

elianda wrote:

What equipment do you use to measure the SNR ratio?

Ideally, some type of standalone audio analyzer, which seem to range in price between several hundred, and tens of thousands of dollars. Keyboard Magazine used an Audio Precision System One when they gathered statistics for a number of soundcards back in 1994. I haven't any guess as to what the Sound On Sound folks used.

What is the variation for SB16 cards of the same CTxxxx number?

This is the kind of information that would be great to have, but that no-one seemed to bother with delineating in their reviews.

The available information isn't particularly meaningful anyhow, unfortunately. On one hand, you've got the manufacturer-provided specifications, which were often based on best case figures provided in spec sheets, and on the other, you've got reviewers that tested, but did so using "default" installation configurations, where no effort was made to optimize mixer settings, or disable unnecessary inputs and (noisy) spatialization features.

To what extent is the origin of the noise the card itself (independent of the system) and what is coming from the systems PSU, bus etc. ?

I think it depends both on the card, and the system and slot placement in which it's installed. A noisy amplifier chip will obviously follow the card around, but in my experience, bus noise has only been particularly bad/noticeable in one or two compact systems that I own.

carlostex wrote:

As far as i know, CQM is not a fake OPL3. Nor is it fake FM, it is just a different form of modulated synthesis. Now i do prefer OPL vs CQM by far and i would even put ESS FM above CQM. In fact and as far as FM is concerned i even like OPP better than OPL2 for instance. Too bad that it was probably considered overkill by AdLib. OPL2 was simpler and cheaper.

CQM *is* fake OPL3 because it squeals loudly and becomes distorted when calling upon advanced OPL3 features. It was only designed to 'play nice' with the standard windows midi canvas. A proper OPL3 clone will not cause distortion that makes it impossible to listen to when calling OPL3 functions. If people honestly have never heard CQM distort to the point of making your ears bleed, I will gladly record plenty of examples.

It's a very crude implementation of firmware that was never proofed thoroughly like ESS 'ESFM' or Crystal FM Synthesis. On that note Crystal FM Synthesis suffered from similar CQM flaws on later chips where Crystal cut costs on their silicon (it can range from a broken noise generator to broken notes entirely; which is a shame since Crystal FM has one of the best noise generators in its CS4236B implementation).

Cloudschatze wrote:

I'd like to think that the majority of people here understand that neither Creative's FM implementation, nor anyone else's, is exactly the same as Yamaha's (similar to the fact that not every General MIDI implementation is a Sound Canvas), and aren't so hung-up about it as to dismiss an entire soundcard on that basis. I happen to appreciate actual OPL output quite a bit, as I'm sure many do, but have systems with OPL-bearing cards for that express purpose. For me, if the AWE64 Gold lacked FM entirely, it would do nothing to diminish or detract from its other virtues.

Of course there are, but this thread wasn't meant to be some sort of comparison of everything versus the AWE64 cards. The original question was whether there is anything special about the AWE64, and whether it would be worth picking up. Several objective and subjective points were made to this end, from actual owners and users.

Well my side of playing the devil's advocate is explaining exactly that-- there is absolutely nothing special about it, and it is definitely not worth picking up unless free. I suppose it would be fine for a digitizer card, but then you may as well just use a proper PCI card.
So the problem of the AWE64 is... it never excels on any type of frontier.

5u3 wrote:

brostenen wrote:

For instance. My CT-2910 is having no hanging note, and it has no line noise.
Others, having a CT-26XX, are experiencing no line noise, responsive bass, yet still hanging note.

Sorry, I see no reason why the CT2910 shouldn't be affected by the hanging note bug. It has one of the buggy DSP revisions and it doesn't come with the CT1747 bus interface.
When I tested a CT2910 (almost nine years ago 🤣), it had the bug.

Sorry... Nope... Not having the hanging note on mine.
Tried playing Doom 1, 2 and Ultimate Doom. Epic pinball. Dynablaster. Monkey Island. And a load of other games, using this card on a P-133 in MsDos6.22. Not a single trace of hanging note or noise such as components noise going straight into the amplifier on the card. You know... When harddrives are reading of writing and such.
Had it hooked up to a stereo hifi sort of thing to check out the card.

Can you tell me how exactly to make that hanging note? As I am unsuccessful in reproducing this error. 🙁

Try Hexen with -warp 02 switch.

raymangold wrote:

CQM *is* fake OPL3 because it squeals loudly and becomes distorted when calling upon advanced OPL3 features. It was only designed to 'play nice' with the standard windows midi canvas. A proper OPL3 clone will not cause distortion that makes it impossible to listen to when calling OPL3 functions. If people honestly have never heard CQM distort to the point of making your ears bleed, I will gladly record plenty of examples.

It's a very crude implementation of firmware that was never proofed thoroughly like ESS 'ESFM' or Crystal FM Synthesis. On that note Crystal FM Synthesis suffered from similar CQM flaws on later chips where Crystal cut costs on their silicon (it can range from a broken noise generator to broken notes entirely; which is a shame since Crystal FM has one of the best noise generators in its CS4236B implementation).

I never read any kind of technical CQM documentation, as far as i know CQM was only Creative Technology because they acquired E-mu. This doesn't sursprise me at all, Creative has had a history of of acquiring technolgy from somewhere else to apply on their products. I'm a little sceptic about E-mu creating CQM to be exclusively an OPL3 clone so i'll leave that to whoever knows the techncal details of CQM technology.

Anyway, with that said i'll stick with what i said from the beginning. If i discount the Emu8000 synth the AWE cards have nothing else unique to offer that makes it standout. Even for DOS if low price is a requirement a YMF-71x card is a better option. But everyone's opinion is different...

raymangold wrote:

CQM *is* fake OPL3 because...

CQM is an FM implementation, designed by Dave Rossum of E-MU, that achieves a degree of compatibility with Yamaha's implementation by mathematically approximating the standard OPL3 waveforms using quadratic splines. It sounds different, and obviously isn't 100% compatible with the OPL3. This is pretty common knowledge at this point; there isn't any imagined debate here.

CQM is not fake "FM" though, which is what you've previously stated. If you want to refer to it as "fake OPL3," that's your prerogative.

It's a very crude implementation of firmware...

You never explained your position on this the last time it came up. As part of the larger hardware implementation, the quadratic spline waveshaper circuit itself can be hardware or software/firmware based. All of the available CQM patent information, including the dedicated co-patent, describe use of the hardware implementation. You seem pretty convinced that Creative used a software/firmware implementation of this waveshaper circuit though, and I'd like to know why.

Well my side of playing the devil's advocate is explaining exactly that...

Do you generally play devil's advocate with products that you have no direct experience with? Most of what you've stated on the subject, especially the commentary on the general output quality, has been exceptionally unsubstantiated.

Cloudschatze wrote:

raymangold wrote:

CQM *is* fake OPL3 because...

CQM is an FM implementation, designed by Dave Rossum of E-MU, that achieves a degree of compatibility with Yamaha's implementation by mathematically approximating the standard OPL3 waveforms using quadratic splines. It sounds different, and obviously isn't 100% compatible with the OPL3. This is pretty common knowledge at this point; there isn't any imagined debate here.

CQM is not fake "FM" though, which is what you've previously stated. If you want to refer to it as "fake OPL3," that's your prerogative.

That's what tried to say too. People must understand that there are tons of FM synthesis implementations. Just because FM Operator type L became the standard for IBM PC's and compatibles does not mean other types of FM synthesis are not FM implementations. Whether we like it or not, CQM is an FM implementation. So is ESS FM or Crystal. In fact some games do support ESS FM specifically don't they?

For someone who wants true OPL3 FM don't get an AWE64. That said some people might like CQM FM better than OPL. I still don't think there's anything special about the AWE64 cards discounting EMU8000. I don't like CQM, so my AWE64 Gold has the SB16 part disabled and i use it exclusively for its EMU8000 synth. I have 4 diferent sound cards in my system and at this point i just dislike one of them having a single OPL3 instead dual OPL2.

The OP got one cheap, so that is that.

I really like the Reverb, Chorus and Stereo expand options on the AWE64. It brings a bit of new life in older games. I believe the Yamaha cards have similar effects but I haven't got much experience with them.

For the whole series of Maxi Sound 64 cards as well as the EWS64 series the DSP effects can also be applied to the codec. The Maxi Sound 64 cards even bring a tool with effect profiles for most of the games of that time called Maxi FX. This is not special for cards of this time.
(GUS PnP allows also for effects)

brostenen wrote:

Can you tell me how exactly to make that hanging note? As I am unsuccessful in reproducing this error. 🙁

If you haven't noticed it yet, don't worry. I only mentioned it because someone might read your comment about the CT2910 not having the bug, and later be disappointed to find out the opposite.

Here is some info about the bug. At this point we have a pretty good idea of which SB cards are affected.

5u3 wrote:

If you haven't noticed it yet, don't worry. I only mentioned it because someone might read your comment about the CT2910 not having the bug, and later be disappointed to find out the opposite.

Here is some info about the bug. At this point we have a pretty good idea of which SB cards are affected.

Thanks.... My mistake. I have allways been told that the hanging note, is that error you get when the opl chip keeps playing a continuing sound until a cold boot or re-initialization of the drivers.
I have never been told that it was a midi add on sort of bug. And therefore I have never searched for info my self.
Sorry again. My mistake.

brostenen wrote:

Thanks.... My mistake. I have allways been told that the hanging note, is that error you get when the opl chip keeps playing a continuing sound until a cold boot or re-initialization of the drivers.

I think that kind of problem is common when you use an older sound card with a faster CPU. The software doesn't wait long enough for the sound chip to process the commands so the output might sound garbled, or maybe only some notes might hang because the off-command is not processed.

idspispopd wrote:

I think that kind of problem is common when you use an older sound card with a faster CPU. The software doesn't wait long enough for the sound chip to process the commands so the output might sound garbled, or maybe only some notes might hang because the off-command is not processed.

Nope....
I had that problem on an Cyrix 80484slc2-40 and an Intel 80486dx2-66 with both a Sound Galaxy 16 and a Creative SB16 Value Edition. Back in 1993 to 1996. (unless you would count 486's as fast CPU's)

That combination should be period correct. Still some games may use timing loops that are not long enough for faster CPUs, so if you didn't experienced this with all games I'd accept this as an explanation.

carlostex wrote on 2015-01-10, 08:58:

elianda wrote:

Maybe its a different issue, because with EMM386 and a lot of memory it reports 'Can not extend server pages'.
I haven't heard of this problem yet. Maybe check your bank file or generate a fresh one from the patches.

Do you have success using MegaEM 3.10 with GF1 Ultrasound cards? It never works for me, reporting an error that it can't load the patch bank file. With MegaEM3.03B i don't have these issues so i reverted to that. Maybe MegaEM 3.10 is only for Interwave cards?

I got similar message with Mega-Em 3.11 on GF1 (GUS classic) card: "ERROR: Can't find patch bank file.", but 3.10 and 3.11 are also support other Ultrasound cards too. I have realized that Mega-Em 3.11 with GUS PnP read this patch bank informations form the ROM of the card. This Mega-Em I got from the PnP DOS installed directory so this is why there is no patch bank files with it, because GUS PnP card no need this. Likely your 3.10 also from a PnP set/environment. I have not found other 3.11 version, but I have found a full 3.10 version and this package has patch bank files (*.bnk):
https://web.archive.org/web/19970220155138/ht … lic/sound/Util/
With these files I have made a full Mega-Em 3.11 package and I have upload it to my page:
https://gona.mactar.hu/GRAVIS/MEGA311.ZIP
and VogonsDrivers:
http://www.vogonsdrivers.com/getfile.php?file … menustate=49,38