Theta_Frost wrote:

there doesn't seem to be much of a point to coding a .SPC tracker given XM and MOD trackers

I had the same standpoint before I actually get into SNES development. Now I think that a SPC tracker would actually make sense. The main reason is to allow you to hear exactly what you get, because difference between what you hear in a XM tracker and in resulting SPC is major (not only in sound quality, but also in envelope timings, volume levels etc). Other reasons are that you simply can't make good use of the echo feature while composing, samples are filtered and looped in a special way that could affect to the timbre greatly (and you have to resample them to a very special frequency 33488 hz to make them loopable), you probably would like to use hardware ADSR, which is difficult to control using a converter, also control on the memory usage is difficult, especially if you going to use echo.

SNES does have unique tone, that's true, because it is not just samples, it is special kind of samples.

Stevens wrote:

.... Just going to throw this out there - what exactly would the use be for the MIDI port on the SNES Emulator SE (original device used to test/develop SNES games)?....

Answer? I mean if the SNES is a sample driven machine wouldn't MIDI be out of the question/useless without something to generate sound?

Screamforme99 wrote:

I mean if the SNES is a sample driven machine wouldn't MIDI be out of the question/useless without something to generate sound?

MIDI is just triggers and messages. You can use MIDI messages to trigger samples and change parameters and banks and all that; it wouldn't be much different to using the stock GM instruments (which are also just samples with loop points), in this regard.

Last edited by Victory Road (March 19, 2012 4:22 am)

Nobody discounts it, rather, you can obtain same or very similar effects without a SNES in sight, since there is no synthesis going on.   What 'tone' do you mean?
I surely associate certain sounds with the SNES, but that would be all.
Of course, for those actually wanting to deploy music on the hardware, tools are needed to hear what you are gonna get. But doing it just to obtain a certain aesthetic is what is usually 'disregarded'

Last edited by akira^8GB (March 19, 2012 10:41 am)

You could probably make a really interesting SNES chippy synth using short chip like sound waveforms (possibly even generated) and taking advantage of the SNES DSP's inbuilt realtime echo, noise generator, ADSR and FIR filter abilities btw *grin* - frankly it would be a beast :O)

Last edited by ne7 (March 19, 2012 11:55 am)

Yeah if the delay can go fast enough perhaps get some Karplus-Strong happening! Sounds fun...

ne7 wrote:

You could probably make a really interesting SNES chippy synth using short chip like sound waveforms (possibly even generated) and taking advantage of the SNES DSP's inbuilt realtime echo, noise generator, ADSR and FIR filter abilities btw *grin* - frankly it would be a beast yikes)

Now ThAT is something very interesting.

akira^8GB wrote:

Nobody discounts it, rather, you can obtain same or very similar effects without a SNES in sight, since there is no synthesis going on.   What 'tone' do you mean?

I'm assuming this was directed at me?

I'm not going to pretend I know too much about the technical side of this particular console. You could make this argument about any chip synth though - it's mostly just square waves, nothing that a laptop or a dedicated synth can't handle, but there are subtle reasons that many of us still work with original hardware having to do with waveform glitches, tone of output hardware, forced low fidelity due to technical limitations. The SNES may represent the transition out of waveform generation, but it still has those qualities as a sampler and to me it seems worth pursuing.

Also, this other thread about pursuing SGB's ability to draw on the SNES' hardware: http://chipmusic.org/forums/post/95099/

kineticturtle wrote:

You could make this argument about any chip synth though

But this is exactly the "problem", the SNES has no synthesis chip at all.

In fact the SNES do have some synthesis capabilities, although I doubt they were used often (to my knowledge there are few rare cases). It capable to generate noise (one generator for all channels) and modulate pitch of one channel with output of other channel, and few channels can be stacked this way - a kind of FM synthesis.

SNES's own tone is not in the synthesis capabilities, though - it is in the (relatively unique) set of limitations imposed on the sample playing. Combination of these limitations create distinctive sound, you can distinguish a SNES music from an Amiga MOD, a XM module, or a sample-based music from low-spec systems (C64, ZX etc - 4-bit sound) just by listening, without much trouble. Of course, it is rather easy to imitate any of these kinds of sampled music with modern sound software.

Last edited by Shiru (March 19, 2012 1:00 pm)

If I didn't tell you which one is which and if there was no artifaction on sample looping on the SNES version, the two tunes are exactly the same.

yeah sample looping would be easier with a proper tracker of some kind.   I spent ages back in the day getting those on the right boundaries.

Last edited by 4mat (March 19, 2012 1:49 pm)

So the only difference is the looping on the snes is sloppy?

Looping has resolution of 16 samples; samples are 16-bit but lossy compressed and decompressed on-fly - means specific distortions, forward-only looping, interpolation is applied, pitch range is limited. Plus hardware ADSR, panning/volume control with 8-bit resolution, and echo capabilities with configurable filter.

as far as i remember the wind in the intro to final fantasy 3 (6) was generated by the noise generator + filter as an example smile

http://www.youtube.com/watch?feature=pl … oA0#t=154s

I think the SNES sound or "tone" can largely be attributed to the sort of faux-wavetable synth way of handling samples. The memory would severly limit your sound budget, so you needed to be very economical with your samples. Most sounds that aren't overly complex, such as pianos would have a short transient and then a very short loop point, with volume envelopes taking care of the decay. Same thing for cymbals and snares with slightly longer loops,

Japanese SNES composers did use a lot of the same sounds. In particular the Miroslav strings, and a bassline I'm pretty sure is sampled from a DX7. I've also suspected they got the samples from some common library. And it almost seems to me like Nintendo provided tools for third parties in Japan but not for western developers.

I toyed around a bit with faking my own SNES samples last year. Although I have no insight on how the built-in echo/reverb worked so I just put a primitive reverb on the master channel.
http://tindeck.com/listen/tfbu