[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/vr/ - Retro Games

View post   
File: 137 KB, 688x480, YM2414_02.jpg [View same] [iqdb] [saucenao] [google]
1941149 No.1941149 [Reply] [Original]

You people know well about the issues getting in the way of authentic graphics reproduction on modern emulators, now something about sound:

Actual recordings of basic waveforms from an old sound synthesis chip, rather than what they would be in theory:

http://www.tinyloops.com/doc/yamaha_tq5/waveforms.html

It's not a chip that's known for being used in consoles or AFAIK even in many arcade machines, but it is a yamaha chip in a line related to all the other ones used.

>> No.1941152
File: 11 KB, 300x248, wave-2.gif [View same] [iqdb] [saucenao] [google]
1941152

This one's the actual triangle wave.

>> No.1941153
File: 10 KB, 300x249, wave-3.gif [View same] [iqdb] [saucenao] [google]
1941153

A half sine, perhaps just like the one adlib chips produced.

>> No.1941212
File: 11 KB, 300x248, wave-5.gif [View same] [iqdb] [saucenao] [google]
1941212

An alternating sine, a soundblaster 16 waveform.

>> No.1941214
File: 12 KB, 300x249, wave-7.gif [View same] [iqdb] [saucenao] [google]
1941214

>>1941212
A camel sine, another SB16 waveform.

>> No.1941247

And what do emulators produce?

>> No.1941330
File: 7 KB, 576x744, OPL_waveforms.png [View same] [iqdb] [saucenao] [google]
1941330

>>1941247
Emulators would produce sounds with these as components instead of anything like that.

>> No.1941345

>>1941149
Visual comparisons of waveforms are very bad for comparing audio. Post spectrograms instead so we can easily see how big the audible differences are.

>> No.1941361

>>1941247
>>1941330
Doesn't really matter right now though, we haven't seen any of the measured waveforms for any of the OPL of OPN chips.

>> No.1941380

>>1941330
What emulator is that from?

>> No.1941452

>>1941380
It's not from an emulator. It's the idealistic waveform that all sorts of emulators would use, particularly if people haven't inspected the actual real world behaviour of the FM synth chips and factored any such observations into the emulation.

>> No.1941462

How exactly is this going to work? The SNES alone has 6 different revisions not coutning the mini. There's no way to get truly authentic sound from emulation. Honestly it doesn't matter because emulation is so close you can't really tell the difference unless you're some audiophile tard that likes to listen to the worms chew in the piano while listening to Beethoven.

>> No.1941509

>>1941345
This. just looking at a waveform tells you very little about how it actually sounds, unless you're autistic or something— which a good deal of /vr/ is... so nevermind, I guess.

>> No.1941535

>>1941509
It's not really a case of how *it* sounds, it's a component waveform, so each of those affects how everything sounds, except for sounds synthesized with just pure sine waves.

>> No.1941545

>>1941462
What, you think sony redesigned the sound chip for each of those? That would be kind of silly, refining a technology when they were just getting close in order to dig for information to usurp nintendo's position soon.

>> No.1941546

>>1941535
Exactly. So what's the point of posting a waveform when a spectrogram will actually show you the overtones present in relation to the fundamental, giving the reader a much better idea of how the final heterodyne wave will sound?

>> No.1941552

>>1941545
(refining a technology for a competitor)

>> No.1941569

>>1941545
They don;t use the same chips. Has nothing to do with refining.

>> No.1941570

>>1941546
That would be a whole lot of spectrograms to post, adlibs, soundblaster pros, 16s, awe32s, awe64s, value soundblasters and clones/compatibles...

And a lot of that would be limitations of the DACs and other components on the card.

This explains a principle that would drive such a discrepancy in a simple form.

>> No.1941572

>>1941462
>>1941545
The SNES doesn't use FM synthesis though, it has a wavetable or sample based sound system.

>> No.1941582

>>1941149
This thread reminds me of all the accuracy issues surrounding Commodore 64 sound emulation (The SID chip: MOS-6581/MOS-8580). I didn't think that Yamaha FM emulation was possibly inaccurate still, but it makes sense.

>> No.1941590

Really, sound is a big issue for me with emulation. I always go out of my way to play a game on the hardware, because the sound itself is a lot warmer, and real. I don't think we'll ever get to that level of accuracy on an emulator, but if we did it would probably be from Byuu. Someone should ask him what he thinks on accurate sound emulation, since he's big on accuracy in everything else.

>> No.1941593
File: 190 KB, 300x400, in to the trash it goes.png [View same] [iqdb] [saucenao] [google]
1941593

>>1941452
>I haven't inspected the actual real world behaviour of an emulator

>> No.1941595

>>1941149
>You people know well about the issues getting in the way of authentic graphics reproduction on modern emulators
Yeah: there aren't any.

>> No.1941602

>>1941582
>accuracy issues surrounding Commodore 64 sound emulation
Most of that is people not appreciating how poor the manufacturing consistency was with the analog components. Even the real hardware doesn't sound like the real hardware.

>> No.1941606

>>1941595
But there are.

http://arstechnica.com/gaming/2011/08/accuracy-takes-power-one-mans-3ghz-quest-to-build-a-perfect-snes-emulator/

>> No.1941607

>>1941590
>the sound itself is a lot warmer
Audiophile for "distorted". We can measure and simulate that. Good SID chip emulators do exactly that.

>real
Audiophile for "I'm an magical thinking idiot".