Sound Card Comparison 2005!

[HondaSiR]

Very interesting, Tom. So now we know that there are at least three types MT-32s based on their "slight" sound differences. The 1987 MT-32... the 1990 ROM-PLAY MT-32 (probably has better S/N ratio than 1987 MT32)... the LAPC-1/CM-64/CM-32L... and the CM-500. I wonder what group the MT-100 belongs to?  :?

[Zemus]

The Audigy comes with a software synth that you can use in Windows, but is mainly used by the SB16 emulation driver when you play DOS games. It also comes with a synth that uses soundfont banks only. There should be 3 banks (1mb, 2mb and 4mb I believe) included with the card, but I think it's optional if you want to install them. I've barely used the soundfont banks and the SW synth, but I don't think they sound very good. The SW synth is particularly bad, almost on the level of the SW synth that comes with Windows.
The strength of the Audigy's synth is the fact that you can load it up with your own sounds, so they only included some rather small banks for those that occasionally play a MIDI file. MIDI enthusiasts that buy the card is probably going to load it with their own sounds anyway, so why bother with an expensive sample set that would only make the card itself more expensive?

[Laust]

I'm downloading the LAPC Ogg to see how it compares to playback on the CM-500 in Mode B.  There's DEFINITELY a difference when playing the MIDI through the MT-32 and CM-500.  Right off the bat I noticed, the music plays with very little reverb in the beginning of the song, on the 500.  The MT-32 seems to ignore the reverb settings and renders the whole song at the same reverb level.  And you're right -- there's many differences in the sounds -- and also, how many sounds play.  (I noticed flute harmonies in some sections on the 500 that weren't heard on the MT-32.)

One of the most strange differences in my opinion is the choir around mid way through the tune, which changes pitch when played on the MT-32, but not on the CM-32L.

Sure is a lot of SysEx in that MIDI file!  Curious about the constant barrage of Program controllers during most of the music.  (I only noticed because the MIDI input to the CM-500 comes from the MT-32's MIDI THRU port, so the MT-32's display shows each patch change.)

Yes, it's very messy. The music editor this was composed with is modelled after a classic Amiga tracker (notes entered by hand or by using the PC keyboard as a virtual piano, commands such as vibrato, etc. are hexadecimal numbers entered alongside the notes), and was probably written using only the LAPC-I manual and a lot of guesswork. It's as close to banging the hardware directly as you can come on a MIDI device. This probably explains the constant stream of controller changes and why the tunes almost always cause buffer overflows or checksum errors on the MT-32 if you don't add delays between SysExes.

Another curious thing to note is that the percussion is not on channel 10, but on channel four (I think, but it's one of the first eight anyway). On the MT-32 you can see it changing the instrument almost constantly (at intervals matching the percussion you hear) and in fact it ties up the MT-32's processor so much the volume control won't respond

And...if this is a LAPC-I song file, why does it have SysEx for the MT-32's display?  "BL & ASF - Inferno"  and "Title Tune" show up on the display while you're transmitting the SysEx.  Just for reference, I suppose.

That's only there because I had both my MT-32 and CM-32L hooked up and thought it was cute.

In case anyone's wondering, I recorded the tune using MIDI-OX, to log the MIDI to a human readable (well, more like easily machine parsable) text format. This is then processed with a couple of Perl scripts I wrote to insert the display sysex, put appropriate delays between the SysEx commands, and finally to fade the tune if appropriate. As far as I'm concerned this is easier than using Cakewalk or similar

Definitely some differences between CM-500 and LAPC playback, too.  Not much, but enough to notice.

Interesting! Care to record that?

[apeman]

The Audigy comes with a software synth that you can use in Windows, but is mainly used by the SB16 emulation driver when you play DOS games. It also comes with a synth that uses soundfont banks only. There should be 3 banks (1mb, 2mb and 4mb I believe) included with the card, but I think it's optional if you want to install them.

Minor nitpick It's 2, 4 and 8MB actually.

[moturimi1]

And, I noticed from moturimi1's MP3 recording of the original SC-55 Sound Module that the Hall1 SysEx delayed transmitting at the start of the CNI music...by about a second.  Whether this was caused by the device or by the playback software, I don't know.  I haven't noticed that with any other samples so far.  The Hall1 SysEx is placed about 1.2 seconds before the first note plays, which should be plenty of time for the software to send it.  It's only 11 extra bytes of data to send...  Interesting.

Maybe I made a mistake during recording. I think the SC-55 handles the SYSEX without any problems. I can rerecord the file if you want.

[Ari]

Yes, it's very messy. The music editor this was composed with is modelled after a classic Amiga tracker (notes entered by hand or by using the PC keyboard as a virtual piano, commands such as vibrato, etc. are hexadecimal numbers entered alongside the notes), and was probably written using only the LAPC-I manual and a lot of guesswork. It's as close to banging the hardware directly as you can come on a MIDI device. This probably explains the constant stream of controller changes and why the tunes almost always cause buffer overflows or checksum errors on the MT-32 if you don't add delays between SysExes.

Another curious thing to note is that the percussion is not on channel 10, but on channel four (I think, but it's one of the first eight anyway). On the MT-32 you can see it changing the instrument almost constantly (at intervals matching the percussion you hear) and in fact it ties up the MT-32's processor so much the volume control won't respond

I don't know if you remember, but I tried to record this one a while back too, and complained I kept getting hanging notes in that tune. I see you managed to avoid it completely. Did you do anything special? or did using MIDI-OX do the trick for that?

[Alistair]

Minor nitpick  It's 2, 4 and 8MB actually.

To use this point, re: the Audigy, it sounds not so bad if you use Creative's 8 MB Synth. Certainly not great, but better than that piece of crap 2 MB one.

- Alistair

[Tom]

Another curious thing to note is that the percussion is not on channel 10, but on channel four (I think, but it's one of the first eight anyway). On the MT-32 you can see it changing the instrument almost constantly (at intervals matching the percussion you hear) and in fact it ties up the MT-32's processor so much the volume control won't respond

Not uncommon for Ken Allen to do in Sierra soundtracks.  Just ask Ari.     But I know the reason Ken Allen uses a non-percussion channel for drums.  The Inferno probably used custom percussion sounds, unlike Allen.

Ari, did you find one of those MT-21's yet on Ebay?


moturimi1, no...not necessary to record it over.  It's fine, hardly noticeable.  I was just curious about it.

Sure, I'll record a CM-500 version of the 'Inferno' MIDI.  

I had to get to work so I didn't get a chance to experiment with the MIDI, but when I originally played it on the CM-500, I may have accidentily had the GS parts of the synth turned off...may have affected it.  I noticed the difference in some of the percussion sounds that may have relied on GS synth?  Anyway, I'll play around with it some more when I get time.

[Tom]

I had a chance to play around some more with the Inferno MIDI this morning and found that in default power-up Mode B (CM-32L mode) on the CM-500, it plays identical to the Ogg recording from the CM-32L.  I'm not yet totally familiar with the music, so there may be subtle differences I'm not yet catching, but for my money -- it plays the same on the CM-500.

[Ari]

Not uncommon for Ken Allen to do in Sierra soundtracks.  Just ask Ari.     But I know the reason Ken Allen uses a non-percussion channel for drums.  The Inferno probably used custom percussion sounds, unlike Allen.

Ari, did you find one of those MT-21's yet on Ebay?

Nope, still looking...  :wink:

Ken Allen really makes converting to GM quite exauhsting... grr...  :smt013:

[Alistair]

Ken Allen really makes converting to GM quite exauhsting... grr...

Cool emoticon.. and how so exactly? (Just curious to see what you're thinking.)

- Alistair

[Tom]

Probably the biggest hassle is with percussion tracks.  And I second that grrr!   :smt091

[Alistair]

You mean for example, the Jones In The Fast Lane percussion 'switching' between drum programs?
Or do you mean percussion on channels other than the drum part number 10?

- Alistair

[Ari]

Ken Allen uses Channel 7 as a percussion track occasionally, but not all the time.
So what I had to do is clone channel 7 and change it to channel 10 (since GM supports only one percussion channel - 10), remove the parts that aren't percussion from the cloned track,  and remove the parts that are from the original track.
In addition,  I had to clone channel 10 and remove everything but the reverse cymbal notes, then change it to an unused non-percussion channel such as channel 1, and change the patch to a reverse cymbal.
I had to do that because the only GM drumkit that has reverse cymbal is the Electronic drumkit, and I can't use that. The Standard or Orchestra kits don't have a reverse cymbal for some odd reason.

As I said - Major pain in the ass...  :roll:

[Alistair]

Which tracks, for example, are you referring to?

- Alistair

[Ari]

The SQ1 introduction, for starters...
you can go through the rest of the soundtrack yourself. I'm sure you'll find those tracks.

[Laust]

I don't know if you remember, but I tried to record this one a while back too, and complained I kept getting hanging notes in that tune. I see you managed to avoid it completely. Did you do anything special? or did using MIDI-OX do the trick for that?

I remember. I think you were using Cakewalk 3 at the time? In this case, CW3 just can't handle the flood of MIDI data and I think there are even some controller changes it simply ignores. MIDI-OX just records everything sent over the MIDI port, nothing more nothing less. In this case, a MIDI file generated from MIDI-OX' output won't load in Cakewalk 3 either (same problem with hanging notes), so the problem lies with Cakewalk itself. I think you've had more success with more recent versions of the program, anyway?

I wouldn't necessarily recommend recording MIDI via MIDI-OX to everyone, but it works for me. Keep in mind that I record the tunes directly from the music editor, so I can stop and start the tunes as I please. If I had to do extensive editing of the MIDI data, I probably wouldn't use MIDI-OX' output (or maybe I would, but it'd be a lot of work )

I had a chance to play around some more with the Inferno MIDI this morning and found that in default power-up Mode B (CM-32L mode) on the CM-500, it plays identical to the Ogg recording from the CM-32L. I'm not yet totally familiar with the music, so there may be subtle differences I'm not yet catching, but for my money -- it plays the same on the CM-500.

I would expect it to sound the same. I'm not surprised if the slightly dodgy mode on the CM-500 (Mode A? the one that also caused a few Sierra soundtracks to fail) would't work, but via the others I would expect it to play fine. Still, if you find any differences between LAPC and CM-500 playback, do let me know

[Alistair]

The SQ1 introduction, for starters...
you can go through the rest of the soundtrack yourself. I'm sure you'll find those tracks.

SQ1's intro percussion 'Deep Snare' should convert straight to 'Standard', shouldn't it?

- Alistair

[Ari]

Yes it does, but the MT-32's Reverse Cymbal translates to Chinese Cymbal on the GM standard drumkit.

[Alistair]

So.. you move the 'E4' note (I think E4, anyway) to a new channel fr 'Rev Cym' as you said before. Common enough phenomena with MT-> GM conversions, certainly not unique to Allen's compositions.

- Alistair