OctaMed sound studio

[Karlos]

Actually, OctaMED SS is a bit of an Achilles heel for UAE. No amount of fiddling around with UAE's settings can get rid of the lag between pressing a key and hearing the sound; something no genuine amiga suffers from.

Of course, when it comes to basic playback of a completed track, it (understandably) performs very well, particularly on many-channel modules.

[Karlos]

Even with the Maestrix redirecting the audio through AHI?

Might be worth speaking to Toni Wilen about it to see if it's fixable or not.

Not tried proper WinUAE with host-native AHI drivers for some time, so I can't say for sure, but it always seemed like the issue was more to do with the host audio latency than anything. I guess that from the host perspective, UAE is just another application that presents itself as a stream source to be buffered and mixed into the final output. That will always add delay.

[Karlos]

We have spoken about this before, as I used WinUAE in the recording studio before, I did get the latency down below 10ms, and certainly low enough to be imperceptible to my feeble human ears

You're going to have to remind me how you achieved that. I think I am going senile. Also, I forget stuff.

Though I tried to use real hardware as much as possible for that "authentic" 8bit audio... Very few will ever appreciate the effort I went to

I do recall you sending me a sounndbite and my having to identify what I thought was from Paula

[Karlos]

Side note, I do plan to release a "PaulaAudio" AudioUnit for the Mac, that will crush to 8bits, EQ and introduce the quantisation errors (that Karlos and I worked out a few years ago) to any audio stream... Watch this space

I remember that. Good old non-linear Paula

The other effect you need to emulate properly to get that unique Amiga sound is the sample-rate aliasing, which might be slightly trickier to get right. I've always thought that ringing sound it adds to things played at low playback rates sounds a bit like what you'd get via an aural exciter.

[Karlos]

A simple "Fnaarrrrr!" would have sufficed!

http://en.wikipedia.org/wiki/Exciter_(effect)#Aural_Exciter.

[Karlos]

PC sound cards work by hardware mixing the sound to a fixed output rate (44KHz or whatever).  The Amiga was unusual in that it worked by speeding up and slowing down the individual samples.  This means the aliasing noise changes with the pitch the sample is played at and thus becomes part of the sound.

I think it's a bit different. IIRC, and I might be wrong about this, it isn't that it plays samples at different rates (ie, higher values for faster rates), but instead uses a sample period (physical duration between sample changes) to define the playback speed. This is a reciprocal quantity with respect to frequency - hence lower period values result in faster playback and vice versa. I'm a bit hazy on the details now but I seem to recall that these periods are expressed in a "ticks" of a clock that's derived from the system clock and runs at around 3.5MHz. This means that arbitrary playback rates aren't possible - the "tuning" of higher notes gradually becomes less precise than lower ones since the period value becomes smaller (thus having fewer bits of precision).

There have to be some interesting slew and phase effects resulting from this method of playback that are hard to reproduce in a conventional mixer.