In playing Sierra games I found that generally the MT-32 uses a 44000 Hz frequency..

- Alistair

Still working into the weee hours of the morning and trying a few alternatives. This may not be that bad after all. Its going pretty quick to get these tuned right. Yeah, I may ultimately mix at 44100hz... Probably won't until its a standalone driver though due to processor requirements.

Just a matter of oppinion, but...

I would recommend mixing everything at 44100 or 48000 to START with, and work down from there... That way you can test ALL of the sounds themselves to make sure they're correct, and then work your way down.

(Course, I run the AdLib and SB emulation of VDMSound at 44100, so I may just be spoiled ;->)

Tried your suggestion to see if it would help with the quality of my lowpass filter. Unfortunately, the filter didn't work any better. Worse yet, as expected, 44100 was the burden that broke the camel's back. I can't develop at 44100hz if I want to continue to use my 1333Mhz Celeron laptop to do the development. Good suggestion though. As planned, however, when I make it into a stand-alone driver (and optimize the hell out of everything with MMX) I'll make 44.1Khz the default for ya.

Actually, it seems the MT-32 's default tuning is 44200 Hz.

- Alistair

I'm curious, Spikey, what makes you think that?

I'm probably mistaken, because I know about MIDI, not the hardware stuff.

But I remember that when I hold the master volume and sound group buttons together, it says 'Master Tune: 442.0Hz'.

Don't mind me, I'm way outta my league here. 😀

- Alistair

Here's some trivia you then:
Tuning, of course, is the frequency where A4 is on the keyboard. From here, since tuning is logarithmic, it can calculate the frequencies of all the other keys on the keyboard. This is not the same as the sampling rate. 44100hz is 44100hz because it is exactly double the highest frequency humans can hear. According to Shannon's Theorem and formally the Nyquist Theorem, there exists a Nyquist frequency, that is, the highest frequency you can sample is exactly half the sampling rate. It is then no coincidence, that if most humans can only hear frequencies up to 22050hz, then sampling at 44100hz means you'll cover all the frequency range of normal human hearing. Likewise higher sampling frequencies (like 88200hz and 48000hz) are actually several of the following:
1) Used as advertisizing gimmics by companies like Dolby, Creative Labs and DTS to convince people their audio is of a better quality. (When really only about .1% of people can notice any real difference.)
2) Within the range of your pet's hearing
3) More or less only useful to retain precision when mixing master recordings.

It is, primarilly, the third.

And, for the record, I CAN hear the difference between 44100 and 48000 😜 (Then again, I can also see a HUGE difference between 100Hz and 120Hz)