View Single Post
Old 13 January 2017, 15:12   #60
Pat the Cat

Join Date: Dec 2016
Location: Nottingham, UK
Posts: 481
Originally Posted by paul1981 View Post
Amigas have 4 channels each with independent programmable non-conformant sampling rates - it can't be 100% accurately recorded or emulated by ANY computer system or digital can only be approximated.

A healthy starting point in this project would have been to individually save the instruments and upscale them seperately, but that would obviously entail rewriting the song.

To be honest, my extent would be to put the Amiga audio filter on, record at 192 KHz 24 bit or above straight out of the Amiga phono jacks, then apply compression and eq to taste. Job done! I don't have 192 KHz capability, only 96 KHz 24-bit but that would be okay too.

Could add some echo I suppose if it needed it, but it would have to be decided amount and type for each song. Perhaps some stereo reduction too, perhaps not in some parts of the song though, or perhaps in most parts of the song and then full stereo in particular places (could have fun with echo here and things).
Hooray! Somebody who actually understands Amiga audio. I wasn't trying for best quality possible, I was trying to emulate a viable hardware sound enhancer realistically. Indeed, capturing each of the 4 channels individually, and then mixing them, would give better fidelity.

Here's the way it really works - you have two 8 bits channels on left, two on right. Each 8 bit channel also has an amplitude control (volume), 0 being off, 1 being 36.5 dB loss, and the others arranged on a logarithmic scale with regard to decibel level up to 64 (full volume). So it isn't QUITE 14 bits on each channel, and when you sum that together, you get a 15 bit number, on each of the left and right channels.

A 16 bit sampler should in theory cope, but that's assuming that it has the same mid point as Paula output DACs. Even a .1% difference, you start losing some information. But 44.1KHz isn't enough sometimes - some audio can outshoot that.

The Amiga has two ways of doing Audio - Paula and the CPU. Paula limits you to 28,867 samples per second playback on NTSC, a little higher on PAL (which is why a tiny minotiry of Amiga audio doesn't play on NTSC systems).

So it's a completely different ball game if the sound playback uses the CPU to time things, which was used on some demos. Suddenly you are not limited to 28.867 KHz and can even output faster than the 44.1KHz of CD, as far as sampling and playback rate go.

This is why you should not depend on 16 bit hardware sampling of Amiga audio. It cannot capture the entire dynamic range. It will give you 99% guaranteed, but if you want complete accuracy, then things are a lot trickier.

The original demo samples usually came from vinyl or tape. Often the sources were speeded up when sampling, either to save memory space for the sample, or to capture sound at a high sample rate and play it back.

Don't think all Amiga audio is limited to 28KHz 8 bit 4 channel. It wasn't and isn't. Paula was, but even then, the volume control of each channel adds more dynamic range to what you can play back.

If you don't believe the above, check out the "Audio Hardware" section of the Amiga Hardware Reference Manual (any edition should do).

Originally Posted by Thorham View Post
My A1200 sounds pretty good when playing back WAV files, so I don't know where you got that from
I would love to have uploaded some WAV equivalents for Amiga users to play back... but github limits to a 25mb file size. I guess I could put a zipped wav up, or lha maybe... nope, didn't work. Zipped version is even bigger than the original wav. Sorry.

Last edited by Pat the Cat; 14 January 2017 at 15:05.
Pat the Cat is offline  
Page generated in 0.03983 seconds with 10 queries