Amiga Sound capabilites.

[ATARI7800fan]

After listening to some Amiga game music I have come to the conclusion that it had some of the uplifting music or music that just makes you feel good in a way. Like this song for instance. Do you think that music back then had more soul or creativity. Does modern game music seems to lack.

 

[Shephda]

Computers in general had soul back then.

I remember our local Computer Show had hundreds of vendors, and all the Computer SIGS were there showing off their hardware.

 

You had demos, hardware hacks, programming, and so on. You had TI-99's, C64's, Amiga's, Atari's, and everyone was passionate about their computer. Go to a computer show now and it's all crap imported junk from vendors who really don't know what they sell or care.

 

The music in the demos were awesome, not only in composition, but how well they matched the demo itself.

 

I really miss those days.

[+wood_jl]

Indeed.

 

What I'm curious about, is how the Amiga sound compared to the Commodore 64.

 

Obviously, the Amiga sound was better than the Atari ST.

Obviously, better than the Atari 8-bit computer.

 

Obviously, it has 1 more voice than the C64, and the "stereo" where it piped different channels to either speaker. But that's not what I mean.

 

I mean, could the sound chip in the Amiga make more sophisticated sounds than the SID? I think the SID is pretty good. How does Amiga compare?

[nonner242]

Indeed.

 

What I'm curious about, is how the Amiga sound compared to the Commodore 64.

 

Obviously, the Amiga sound was better than the Atari ST.

Obviously, better than the Atari 8-bit computer.

 

Obviously, it has 1 more voice than the C64, and the "stereo" where it piped different channels to either speaker. But that's not what I mean.

 

I mean, could the sound chip in the Amiga make more sophisticated sounds than the SID? I think the SID is pretty good. How does Amiga compare?

 

 

Yes! I know the Amiga can blow away the C64 sound wise.

Just afew of my all time favs.

 

Shadow of the beast

 

Agony

 

Project X

 

Leander

 

GODS

 

SWIV

Edited May 6, 2011 by nonner242

[NightSprinter]

Well, from what I remember from various episodes of "The Computer Chronicles" the Paula sound chip was just a Digital-to-Analog-Converter. It used direct memory access (DMA), could play 8-bit PCM samples (which were generally sampled at 22KHz), and two of the four channels were mixed to left or right speaker for stereo. I mean it's possible to emulate the SID, just have a look at trackers like AHX and HivelyTracker. I'll admit that even SID sounds might need the CPU to create first.

 

If any of this is wrong, do correct me.

[AtariNerd]

Yep, that's basically it. It was mainly a D/A convertor that was software driven. There were soundbanks of instruments you could get, digitized samples, or with other software you could create your own samples. You called up the samples like any other soundtracker program, pulling the samples off your storage as necessary . Very much more memory consuming than a chiptune. It sounded great, of course, but at some expense.

Edited May 7, 2011 by AtariNerd

[JamesD]

Yep, that's basically it. It was mainly a D/A convertor that was software driven. There were soundbanks of instruments you could get, digitized samples, or with other software you could create your own samples. You called up the samples like any other soundtracker program, pulling the samples off your storage as necessary . Very much more memory consuming than a chiptune. It sounded great, of course, but at some expense.

DMA driven and software driven aren't quite the same.

Software driven would be like the Tandy CoCo where every byte out the D/A converter is sent by the CPU.

The Amiga used DMA where Paula would automatically send the bytes to the D/A converter and could signal the CPU when a transfer completed. Man I wish that idea had been used sooner.

[roland p]

Disclaimer: I'm not a music expert.

 

According to this thread, the amiga could do more than playing back samples only.

 

It sounds like most amiga music consists of samples only and not much filtering (ADSR, ring-modulator, cut-off, resonance etc. the knobs on a synthesizer).

 

It seems like on a c64 you can have it all, real synthesizer like sounds and sampling and filtering:

Skip forward to 1:46 to hear what I mean with sampling and filtering. It sounds like samples with a variable cut-off filter.

Edited May 7, 2011 by roland p

[Lendorien]

Honestly, I think the music designers int he older games had to work a little harder to get good sound combination out of the limitations of the equipment. Modern stuff is basically as capable as anything on radio.

[Petran79]

If your ears were used to minimalistic chiptune music and then you switched to an amiga or something similar it was a huge difference. Same with the SNES. If you had an expensive wavetable soundcard things were even better.

By the time games used cd audio quality music there was nothing exciting anymore regarding technology.

You can even hire an orchestra to compose the music for you. Interest shifted to the actual music quality.

Regarding music games reached perfection.

 

But despite this some Amiga tunes are eerie and haunting even though no game company would use those except in an uptodate remake or remix

[Chilly Willy]

Yes, that old tracker music was really awesome! You can still get a lot of that off sites like the ModArchive. Here's a link to one of my favorite tracker musicians from the Amiga, Scorpik.

 

http://amp.dascene.net/detail.php?detail=modules&view=6477

[NightSprinter]

I'm just curious as to how much CPU and memory that AHX and HivelyTracker would use to generate SID waveforms.

[Seob]

Indeed.

 

What I'm curious about, is how the Amiga sound compared to the Commodore 64.

 

Obviously, the Amiga sound was better than the Atari ST.

Obviously, better than the Atari 8-bit computer.

 

Obviously, it has 1 more voice than the C64, and the "stereo" where it piped different channels to either speaker. But that's not what I mean.

 

I mean, could the sound chip in the Amiga make more sophisticated sounds than the SID? I think the SID is pretty good. How does Amiga compare?

If i recall correct the atari ST was loved by the music industry for it's midi sound. So it not that obvious that the amiga has better sound.

The amiga on the other end found his way into the video editing business.

[Shephda]

Indeed.

 

What I'm curious about, is how the Amiga sound compared to the Commodore 64.

 

Obviously, the Amiga sound was better than the Atari ST.

Obviously, better than the Atari 8-bit computer.

 

Obviously, it has 1 more voice than the C64, and the "stereo" where it piped different channels to either speaker. But that's not what I mean.

 

I mean, could the sound chip in the Amiga make more sophisticated sounds than the SID? I think the SID is pretty good. How does Amiga compare?

If i recall correct the atari ST was loved by the music industry for it's midi sound. So it not that obvious that the amiga has better sound.

The amiga on the other end found his way into the video editing business.

She. Amiga is a female, Amigo is a male.

Edited May 8, 2011 by Shephda

[roland p]

If i recall correct the atari ST was loved by the music industry for it's midi sound. So it not that obvious that the amiga has better sound.

midi itself doesn't have a sound.

[NightSprinter]

I think he meant the MIDI capabilities, since the ST was a preferred system in music studios in its time. Back on topic:

 

I think another thing that made the sound capabilities of the Amiga so great was how easily one could add a sampler to their machine. Was quite a must for the tracker/demo scene.

[Seob]

Remember getting to friends of mine with a Amiga and action replay cart, to rip sound from games. Then transfer them to the pc and make our own music with those samples with modedit and a covox parallel "soundcard" . Still remember the faces on those amiga guys when we dragged in a pc with a tuner, inside the computer club, hooking it all up and played music on our pc.

[NightSprinter]

Dude, I have a Covox compatible parallel port DAC in storage (Disney Sound Source, any Wolfenstein 3D players heard of this one?). If I ever get my 486 DOS machine out, I ought to try it again.

[Seob]

offtopic we build those covox converters ourself. Our latest generation was even made with a circuitboard.

[ATARI7800fan]

Computers in general had soul back then.

I remember our local Computer Show had hundreds of vendors, and all the Computer SIGS were there showing off their hardware.

 

You had demos, hardware hacks, programming, and so on. You had TI-99's, C64's, Amiga's, Atari's, and everyone was passionate about their computer. Go to a computer show now and it's all crap imported junk from vendors who really don't know what they sell or care.

 

The music in the demos were awesome, not only in composition, but how well they matched the demo itself.

 

I really miss those days.

 

exactly, it seem like this days companies are to worried about making money and not taking chances with things, the mystery, the experimentation has vanished.

[kool kitty89]

Indeed.

 

What I'm curious about, is how the Amiga sound compared to the Commodore 64.

 

Obviously, the Amiga sound was better than the Atari ST.

Obviously, better than the Atari 8-bit computer.

 

Obviously, it has 1 more voice than the C64, and the "stereo" where it piped different channels to either speaker. But that's not what I mean.

 

I mean, could the sound chip in the Amiga make more sophisticated sounds than the SID? I think the SID is pretty good. How does Amiga compare?

That's a tough comparison . . . albeit comparing DMA sound (be it the Amiga or other set-ups with more or less capabilities -ie added things like hardware looping or less like limited/fixed sample rates meaning software not scaling is needed, etc), and comparing that with any realtime synth chip in general under the same conditions.

 

The Amiga can't even match POKEY, AY/YM PSG, or an SN76489 without using more memory . . . you technically could use very short looped samples, but that's where the problems come in. The Amiga requires the CPU to loop samples, and that's generally very little overhead for medium/long samples, but now we're talking extremely short (if not single cycle) looped waveforms . . . that's going to eat up a lot of CPU time. (so to do simple chip sounds on the Amiga, you'd want to use longer samples and waste more RAM because of that)

The TG-16's sound chip would be a prime example of what you wouldn't want to do. (it's sample based, but with hardware looping and only using very small 32 word long samples -which would be a ton of overhead on the Amiga- though another system with hardware looping -like the SNES or Ricoh PCM chip in the FM Towns or Sega CD- would be fine at simulating that -albeit with lower output sample rate -only 32 kHz- in both of those cases and degradation of the samples due to compression on the SNES -granted, that output sample rate only matters if you have really good quality amps in the system, and I'm not sure the PCE/TG-16 used such that the analog output would be beyond 32 kHz)

 

The more complex the sound generation capabilities of a given synth chip, the more memory you'd need to store all those different samples. (that would include samples of sounds with various amounts of filtering applied -you can't do analog synth style filtering in realtime on the amiga, so it would need to be pre-sampled) Same case for other forms of synth with many possible outputs, some virtually indefinite in flexibility.

You also have a lot more memory to "waste" with samples in the Amiga compared to some other computers/game systems (especially since game systems used ROM and not cheap disk media), so in a real-world context, being more wasteful isn't so much of a disadvantage. (it was more of an issue on the SNES where you needed samples for everything -vs a synth+sample system where a good programmer/developer could be far more selective with use of sampled sounds -and those that were used could be of considerably higher quality -that's especially significant since a ton of SNES games used FM or FM-like sounds that could have been matched or approximated -or beaten by a large margin due to the far higher clarity/quality of ~50 kHz realtime FM synth- compared to samples)

 

Not all small samples have to be short, mind you. You could use some relatively small samples that just stuck to a rather low sample rate output range on the Amiga (or were just a low sample rate in general for the SNES/other full sample playback systems, or software scaled MOD/sample systems). You'll get rather muffled and/or aliased sound (mainly muffling with heavy filtering and interpolation), and that can sound pretty good for some stylized music (it's one option used to conserve memory on the SNES . . . and also part of why a lot of music/sounds are rather muffled on that system) and is used on a number of amiga games. (limiting the amount of different instruments used also greatly cuts memory use)

For extreme examples, you should check out some of the 8k MOD tracks on battleofthebits. http://battleofthebits.org/arena/Tag/8k/ (that's 8 kB, not 8 kHz . . . just 8 kB for the entire MOD file) There's also some 12k, 24k, and 48k categories for some not quite as extreme examples. http://battleofthebits.org/browser/EntryByFormat/mod12k/ (12k is still pretty damn small though )

 

 

 

Trying to copy other sound chips will tend to give weak results on the Amiga (and some other sample based systems), but it's the overall flexibility and "easy" nature of use that makes it attractive. (don't need any knowledge of how to get good synth sounds produced -analog, FM, set, etc- but just need to find a sound you want to use and record it -plus cut it down to fit in a reasonable amount of memory and use optimal conversion methods for such)

 

There's also the possibility of software mixing to go beyond the base 4 channels on the Amiga, though doing that in most games with a 7 MHz Amiga would need to be conservative. (generally, added mixing is only done for fixed pitch/playback rate samples for SFX/percussion/etc and often uses the interleaving/multiplexing method -playing one sample after another at a higher output rate, like 2 14 kHz channels interleaved with 28 kHz output, or 3 9 kHz channels, 4 7 kHz, etc)

If you have the resource to do realtime mixing/scaling, there's also the option of using ~14-bit stereo by combining channels. (pairing the 2 left and 2 right channels with 1 set to minimum volume and the other at max to add together to approximately 14-bit of amplitude range -it could have been 16-bit if the volume steps had been wide enough; the PCE/TG-16 can do 10-bit PCM that way since the volume range is wide enough to allow 2 5-bit channels to add to 10 bits almost perfectly) In that case, you'd more or less be working with sound output similar to an SB-16 or Adlib Gold. (not technically as good as SB-16, but close enough in most respects -especially given the shoddy quality output of some SB-16 compatibles- there's also a bit more overhead from the need to use a look-up table to mix to the paired channels iirc, but that's not a huge issue . . . that and you'd be limited to ~28 kHz output max unless you were using PAULA in HD floppy mode)

 

 

 

 

Well, from what I remember from various episodes of "The Computer Chronicles" the Paula sound chip was just a Digital-to-Analog-Converter. It used direct memory access (DMA), could play 8-bit PCM samples (which were generally sampled at 22KHz), and two of the four channels were mixed to left or right speaker for stereo. I mean it's possible to emulate the SID, just have a look at trackers like AHX and HivelyTracker. I'll admit that even SID sounds might need the CPU to create first.

Emulating SID is not practical in realtime, but using a lot of samples could be combined with an emulation program to effectively emulate the SID in certain circumstances. (that's what the CoCo SID emulation/simulation examples do iirc) I've seen claims that AHX is realtime synth, but I'm pretty sure that's also sample based. (realtime wave generation would waste a lot of CPU time -very short samples are pretty bad too as mentioned above)

 

Also, 22 kHz was not "normal" but rather then high-end example for games. The majority of Amiga games used ~8 kHz samples iirc, though for instruments that actually played notes, that would vary. (sample rate was used to control pitch -that's why the Amiga's high sample rate resolution was so important; limited/fixed rates on multiple channels aren't much better than software mixing onto a single channel for that reason -the Falcon's DMA sound is also too low res to be of much use for hardware sampling like the Amiga)

 

 

Yep, that's basically it. It was mainly a D/A convertor that was software driven. There were soundbanks of instruments you could get, digitized samples, or with other software you could create your own samples. You called up the samples like any other soundtracker program, pulling the samples off your storage as necessary . Very much more memory consuming than a chiptune. It sounded great, of course, but at some expense.

DMA driven and software driven aren't quite the same.

Software driven would be like the Tandy CoCo where every byte out the D/A converter is sent by the CPU.

The Amiga used DMA where Paula would automatically send the bytes to the D/A converter and could signal the CPU when a transfer completed. Man I wish that idea had been used sooner.

Yes, DMA driven sound is quite useful . . . especially with a fair number of hardware channels and high resolution sample rate settings. The mac had DMA sound too and I'm sure the idea was around for a while before that too. (Mac actually had 22 kHz 8-bit stereo from day 1 iirc, but it was wired as mono with the 2nd channel dedicated to floppy drive square wave generation -it was also just a simple circular buffer in memory that got scanned out at a fixed rate, so not nearly as flexible as the Amiga or even some other simpler DMA set-ups)

Actually, I think the Amiga may be the only example of DMA sound hardware on a major platform where the sample rate resolution was high enough to use as hardware sampling (for pitch).

 

Pretty much all other DMA sound systems were fixed or limited sample rate output. (Mac was 22 Khz tied to the sync rate -albeit often treated as 11 kHz by only writing to every other byte iirc; Tandy DAC in late model T-1000s was fixed 48 kHz 8-bit iirc, Covox sound master was fixed rate too I think, Sound Blaster 1.x was 23 kHz 8-bit, 2.0 was 44 kHz 8-bit -presumably with 23 kHz support at the very least, pro was 22/44 kHz max, SB16 was 44 kHz 16-bit stereo, STe was 6.25/12.5/25/50 kHz, Falcon was limited to divisions of ~50 kHz, so pretty low-res too)

Still, even fixed/limited rate playback (even on 1 channel) is a lot better than software PCM playback . . . well unless you've got a bank of separate DACs and some high-res interval timers along with a very interrupt friendly CPU (like 650x). (the TG-16 almost has that, a game would just need to include a bank of high-res interval timers on-card and you could have 6 5-bit channel sample synth, or fewer if you did 10 bit channels or a mix of the 2, granted for 6 channels at ~22 kHz max -lower pitched down from there- you'd use close to 100% of the CPU time for a good playback engine . . . limit that to 3-4 channels at ~11 kHz, and you could probably do it in-game -or a mix of different rate channels- . . . as it is with the PCE, you only have a 7 kHz timer and 15.7 kHz scanline counter to generate timed interrupts, so the best option would probably be pairing 2 channels for 1 10 bit channel at 15.7 kHz and mixing/scaling in software -perhaps we'd have seen some of that if the TG-16 had gotten popular in the US or Europe )

 

Of course, that's a pretty specific example, and anything aside from a 650x or 680x CPU would quickly become unattractive for interrupt driven sound. (the C64 and A8 are examples where interrupt driven sound is somewhat practical relative to the overall CPU resource available -you have 3 POKEY timers and the CIA timers in the C64 to work with, and 4 4-bit channels to modulate for both -including the C64's hack channel- and 3 channels with lowpass filtering on the C64 . . . but a bit more CPu resource to work with on the A8 -considerably more with video disabled) For games, neither one is probably going to go beyond a single sample channel and probabl mostly for low rate percussion/FX samples. (maybe a limited variable rate synth channel . . . probably not sampling above ~7 kHz, probably more like 4-6 kHz max in most cases)

 

The CoCo III also falls into that category to some extent . . . albeit it's got limited resolution for timed interrupts and onyl 1 DAC to mix to anyway, so you're not going to get variable rate based note sampling. (you've got a lot of homebrew games doing sound that way though -including a multi-channel sample engine used in that Contra project)

 

 

 

Disclaimer: I'm not a music expert.

 

According to this thread, the amiga could do more than playing back samples only.

 

It sounds like most amiga music consists of samples only and not much filtering (ADSR, ring-modulator, cut-off, resonance etc. the knobs on a synthesizer).

 

It seems like on a c64 you can have it all, real synthesizer like sounds and sampling and filtering:

http://www.youtube.com/watch?v=Y6mXLxUZvzg Skip forward to 1:46 to hear what I mean with sampling and filtering. It sounds like samples with a variable cut-off filter.

No, C64 you can have very limited 4-bit sample playback with heavy CPu resource use. (albeit it's at least somewhat realistic to do in-game as the 6502 is very fast at interrupts, so a limited interrupt sample driver is somewhat practical -much less than what you see in the demos though . . . the A8 can play full 4 channel MOD too, but it's CPU intensive and you wouldn't see it in-game -perhaps 1 sample channel at a fairly low rate) Interrupt support is important as it greatly eases multitasking. (cycle timed code is difficult to use outside of demos as such . . . a big reason why you see some decent 4 channel software MOD players on the ST, but pretty much never more than 1 low fixed rate sample channel in-game)

http://yerzmyey.i-demo.pl/ (4 channel software MOD via POKEY)

 

Hell, the ZX spectrum can do some decent 3 channel sample playback stuff too (via the AY8910), for demos situations. Check out yerzmyey's stuff on youtube.

 

 

Honestly, I think the music designers int he older games had to work a little harder to get good sound combination out of the limitations of the equipment. Modern stuff is basically as capable as anything on radio.

Sort of, but a lot of newer (and new) games (talking the last 10 years up to today) still use realtime sound engines (pretty much all sample based -be it MIDI, or a tracker/MOD player of sorts), so there's still a fair amount of work required to get really good results from that.

It's interesting that you saw a massive percentage of early CD based games using redbook audio (with a few exceptions), but then a fairly heavy shift to realtime sound engines again by the middle of the 5th generation (PSX/Saturn) and other cases of streaming lower quality PCM or compressed audio (a fair amount of streaming ADPCM on the PSX and Saturn), the latter obviously for space reasons, but the former (realtime sample/synth music) could also be done to allow on the fly loading from CD.

A pretty big chunk of PS2, GC, DC, and some Xbox games used realtime music engines too. (you see a mix of that on current gen consoles . . . as RAM increases, pre-loading compressed audio files also becomes more and more practical -I'd imagine even some 6th gen games used MP3/OGG loaded into RAM -a 4 minute song at 64 kbps would be under 2 MB)

 

 

 

 

 

If your ears were used to minimalistic chiptune music and then you switched to an amiga or something similar it was a huge difference. Same with the SNES. If you had an expensive wavetable soundcard things were even better.

By the time games used cd audio quality music there was nothing exciting anymore regarding technology.

You can even hire an orchestra to compose the music for you. Interest shifted to the actual music quality.

Regarding music games reached perfection.

Yeah, except you'd be surprised how many games on CD/DVD/etc based systems don't stream music at all, so there still was/is a fair amount of "synth" music used in newer games too. (be it midi, tracker, etc, etc)

You saw an initial burst of heavy red-book audio use on early CD based consoles (very little on PC games interestingly, a lot of MIDI/synth stuff used exclusively, or non CD-DA audio files loaded to the HDD -perhaps a combination of space and the fact that a number of early PC CD-ROM drives may not have had analog sound mixed in -I remember we didn't have our CD drive set up for direct CD-DA playback . . . no music in BUG or Garfield ), but then a major shift to "synth" sound engines in many "32-bit" generation games and up to this generation. (albeit, as onboard RAM increases, compressed sound files become more attractive too -streaming compressed sound is obviously an option on a number of systems too, but also restricts on the fly loading from disc, if that's required)

 

That, and there's still a limit to how dynamic a pre-recorded score can be . . . you can make a really complex realtime music engine that had a ton of dynamic response features (changing with action on-screen). As memory/compression constraints get more flexible, you can have a lot more recoded snippets of music to allow a really dynamic engine that way too. (to the point where a realtime engine would have to be increasingly feature-rich to really have an advantage)

Back in the early 90s, with CD-DA, you had a huge advantage for realtime synth engines, plus a total lack of seek-time for dynamic changes (even if you did try that with CD, there's going to be some slight lag/gap in timing due to that). Of course, there's some cases where CD tracks didn't even try at all . . . that's the main thing wrong with the win9x remakes of X-Wing and Tie Fighter . . . sure you lose the classic shaded polygon look (for better or worse), but the main problem (regardless of liking the update in general) is the removal of the awesome iMuse midi of the DOS games. (original and highly dynamic tracks done in realtime -supporting a variety of platforms including soundblaster/adlib FM, moderate use of the SB-16's OPL3 features in 4-op mode, pretty damn good general midi support, roland MT-32, waveblaster, etc, etc -I forget if it has ultrasound support- )

They replaced that with one long looping compilation of various scores from the original trilogy. (totally un-fitting by comparison) X-Wing vs Tie Fighter did the same thing (the engine the remakes are based on) . . . X-Wing Alliance finally got it right with very nice dynamic use of the theatrical scores.

 

 

 

 

 

If i recall correct the atari ST was loved by the music industry for it's midi sound. So it not that obvious that the amiga has better sound.

The amiga on the other end found his way into the video editing business.

Yes, the ST's built-in MIDI interface gave it a major boost for support in the professional music sector. (a few games also supported MIDI music, namely Sierra games using the high-end MT-32 -as with their PC versions) It would have been interesting if some 3rd party (or Atari) had relased a low-end MIDI module for the consumer/game market. (then again, a simple cart based module would probably have been cheaper and much more flexible -direct interface to a sound chip rather than MIDI interface -still need external sound output in either case . . . or a piggyback plug to the monitor port sort of like Sega did for the FM module on the Mk.III)

 

 

 

 

I think he meant the MIDI capabilities, since the ST was a preferred system in music studios in its time. Back on topic:

 

I think another thing that made the sound capabilities of the Amiga so great was how easily one could add a sampler to their machine. Was quite a must for the tracker/demo scene.

Weren't there plug-in A/C modules for digital sampling/recoding (and playback) on the ST? (be it simple software managed parallel interfaced DACs, or actual hardware assisted modules)

 

 

 

Remember getting to friends of mine with a Amiga and action replay cart, to rip sound from games. Then transfer them to the pc and make our own music with those samples with modedit and a covox parallel "soundcard" . Still remember the faces on those amiga guys when we dragged in a pc with a tuner, inside the computer club, hooking it all up and played music on our pc.

And, of course, all the MOD players and trackers used with DMA sound of the Sound Blaster cards (a shame that no games seemed to make use of FM as well as PCM for music), prior to the Gravis Ultrasound, waveblaster, etc. (and in parallel with those for a number of early/mid 90s games)

Edited May 15, 2011 by kool kitty89

[roland p]

I once made a converter so I could use one of those 'wavetable' cards on my amiga. They accept midi signals so it was pretty easy to do from the serial port.

 

 

Edit: duke nukem on a yamaha wavetable card sounds really cool!:

Edited May 15, 2011 by roland p

[ATARI7800fan]

I once made a converter so I could use one of those 'wavetable' cards on my amiga. They accept midi signals so it was pretty easy to do from the serial port.

 

 

Edit: duke nukem on a yamaha wavetable card sounds really cool!:

nicely done.

[NightSprinter]

Kitty, I do believe there were a few. I've seen some on eBay now and then, but it seems more were made for the Amiga series (as some Amiga music packages even offered the LPT DACs as part of a bundle). I think on the ST line (before the STE), sample playback without those plug-on modules had to be done by CPU through the YM2149F chip.

[kool kitty89]

Kitty, I do believe there were a few. I've seen some on eBay now and then, but it seems more were made for the Amiga series (as some Amiga music packages even offered the LPT DACs as part of a bundle). I think on the ST line (before the STE), sample playback without those plug-on modules had to be done by CPU through the YM2149F chip.

Yes, though CPU grunt would be the route for a "Covox" type DAC set up too, and a possible software managed ADC for recording samples. (there's no way you could record without added hardware though)

 

The YM2149 route can be pretty good, but it's up to optimized conversion tables for using the 3 channels as a linear DAC (often aiming at 8-bit output, but higher res than that was practical too; I thin 9 or 10 bit could be managed OK without the steps getting too coarse -it still probably wouldn't sound as good as a true 8-bit DAC, but mixing to 9-10 bits might sound better than approximating 8-bits on the YM -I started a thread a while back asking if anyone knew more details on the actual logarithmic step sizes used for YM volume, but didn't get any responses). Yerzmyey mentioned that better trackers/MOD players can manage above 16 kHz with 4 channel MOD for a static screens and around 13 kHz in some demos with moving graphics iirc. (simple 3 channel sample drivers using each YM channel as a pseudo 4-bit DAC should be possible at significantly higher sample rates or with far less CPU overhead -that's how some fairly impressive sample trackers are managed on the Spectrum- but there seems to have been little support for that method on the ST -it's used more in game for just 1 channel of sampled sound, a fair amount of games used that to allow sampled drums/SFX without cutting out the other 2 sound channels -YERZMYEY also mentioned that the only tracker available on the ST using that 3 channel method sounds poorer than the Spectrum tracker -might be an issue with the YM vs AY chips, but more likely due to better optimization on the software end)

 

Here's some better examples of MOD done on a vanilla ST:

http://ym-digital.i-demo.pl/music.html

Edited May 16, 2011 by kool kitty89