Cloudschatze wrote:

One workaround is to play affected games from Windows, where software-based, circular buffering is employed for MIDI playback.

it was 98se where i experienced this first in duke3d, using an AWE64. the thread i've linked also mentions the issue occuring both under windows and DOS.

BinaryDemon wrote:

...I'm confused with your obsession with 16bits.

Beg your pardon? The explicit intent was to quantify 16-bit PCM support in DOS, which has commonly been stated/regarded as being almost non-existent, given the prior absence of any widespread research or testing. Thank-you for the derogatory commentary on the effort.

Like I see the spreadsheet contains a lot of 16bit 11hz mono games - for example wouldnt 8bit 22hz stereo be a preferable experience?

Yes, 8-bit @22kHz should be preferable to 16-bit @ 11kHz in most cases.

auron wrote:

Cloudschatze wrote:

One workaround is to play affected games from Windows, where software-based, circular buffering is employed for MIDI playback.

it was 98se where i experienced this first in duke3d, using an AWE64. the thread i've linked also mentions the issue occuring both under windows and DOS.

Similarly, there are accounts of the issues not being present when Duke Nuke 3D is run from Windows. This conversation should probably move to the thread you'd refered to, or a new one even. It's possible that we're dealing with a difference in the behavior of Windows' "generic" MPU driver, as opposed to that of Creative's own driver-set.

Cloudschatze wrote:

Thank-you for the derogatory commentary on the effort.

No insult intended, best of luck with your project.

While I really like what you've done here, I think you should add a column for source material.
For example, many westwood titles use 4bit custom ADPCM compression at various sampling rates in different titles (often 22kHz I think).
Some games actually have 16bit samples as was mentioned in the thread, others have very low quality samples (8bit at 11 or even 8kHz).

In the interest of completeness... 😀

tpowell.ca wrote:

While I really like what you've done here, I think you should add a column for source material.

It's already there. Click on the image of the spreadsheet in the initial post to access the full Google Docs version.

I think Stargunner can go in the list too, it has SB16 support and allows up to 45KHz mix rate.

Cloudschatze wrote:

It's already there. Click on the image of the spreadsheet in the initial post to access the full Google Docs version.

Well then; excellent work Cloudschatze. 😀

BinaryDemon wrote:

I'm not an audiophile at all, and only have a basic understanding of the terminology being used here but I'm confused with your obsession with 16bits. Like I see the spreadsheet contains a lot of 16bit 11hz mono games - for example wouldnt 8bit 22hz stereo be a preferable experience?

YES -- but that doesn't mean that 16-bit audio isn't noticeably better than 8-bit audio. A lot of people have poor genetics or have aged to the point where their sense of hearing isn't high fidelity enough to distinguish between them. There are, however, many people that can distinguish and highly prefer 16-bit audio because of how much clearer it sounds. The difference is pretty stark to me. I'll be the first to admit that anything beyond 16-bit is pretty silly, though. Except for mixing/mastering purposes, 24-bit audio is ludicrous.

It's very similar to any discussion in finance about compound interest formulae. The interest rate parameter is the most important factor to consider, but the number of compounding periods per year also matters. It's just that the law of diminishing returns kicks in far earlier with the number of compounding periods than the interest rate.

The reason for the obsession over 16-bit audio in the context of 1990s DOS games is that it was an awkward transitional period in a lot of ways. The Gravis UltraSound cards were *clearly* the best sound cards(at least for sampled audio), but due to various market reasons they weren't remotely as well supported as the Sound Blaster [compatible] cards. So if you really want to experience the best of the era, you have to cobble together a combination of cards to use with this or that game. To make matters worse, the authentic Creative Labs Sound Blaster 16 cards were fractured into a ridiculous number of models with various pros and cons, making it even more difficult to get this stuff right.

After putting in the effort to build up a proper collection of the best hardware of the era, it makes sense to figure out which games best supported 16-bit audio so as to get return on investment. =)

I think the point of this topic was simpler: the transition from 8-bit to 16-bit audio occurred more or less at the same time as the transition from DOS to Windows and floppy to CD-ROM.
So it is a valid question to wonder if any DOS games ever made use of 16-bit audio at all, and if so, which games were those, and to what extent did they make use of it?
We can answer that question now.

Lets not forget what a garbage fire Creative was in mid nineties. They started by releasing broken 16 bit product, and pretty much never fixed it. Already mentioned SB16 bugs might be why developers avoided 16 bit to cut down on support costs.

I posted this already: "InfoWorld Jan 1993 has 16ASP test where the card couldnt even play non compressed 44KHz sounds without severe distortions! prompting Creative to sent technicians just to conclude there is nothing do be done, blaming standard popular Adaptec bus mastering SCSI controller card". Add no Pro compatibility, cracks&pops, noise, VIBRA CQM and you have a good reason for skipping it.

https://books.google.pl/books?id=AzwEAAAAMBAJ … 6%20asp&f=false

rasz_pl wrote:

Lets not forget what a garbage fire Creative was in mid nineties. They started by releasing broken 16 bit product, and pretty much never fixed it. Already mentioned SB16 bugs might be why developers avoided 16 bit to cut down on support costs.

I posted this already: "InfoWorld Jan 1993 has 16ASP test where the card couldnt even play non compressed 44KHz sounds without severe distortions! prompting Creative to sent technicians just to conclude there is nothing do be done, blaming standard popular Adaptec bus mastering SCSI controller card". Add no Pro compatibility, cracks&pops, noise, VIBRA CQM and you have a good reason for skipping it.

https://books.google.pl/books?id=AzwEAAAAMBAJ … 6%20asp&f=false

I have several Sound Blaster Pro 2.0 cards, and quite a few different SB16 cards. I use both, so the point about having no Pro compatibility is valid, but has a solution. (Just own both!)

I have been able to get my hands on several CT2940s which have authentic Yamaha OPL chips on them instead of the CQM garbage, so that point of yours is out. The CT2940s have extremely high SNR and no cracks or pops, so that point is also out.

As for the 44KHz playback having distortion in it, I don't know. But your attitude of complete rejection of this line of hardware is due more to frustration and laziness than it is technical merit.

aquishix wrote:

of yours is out. The CT2940s have extremely high SNR and no cracks or pops, so that point is also out.

cracks/pops are due to firmware bug (Single-Cycle DMA, afair interrupt is send one sample too late).

aquishix wrote:

As for the 44KHz playback having distortion in it, I don't know.

me neither, apparently neither did Creative engineers according to InfoWorld, and thats in a paper being paid tons of money at the time by Creative (full page ads).