Amiga Sound Quality

[xArtx]

No electronic component, especially no analogue electronic component, is actually a mathematically ideal implementation of its nominal function.

Where did you find that? I found it for the first time this year.
It's not only not an ideal implementation, but also an unintentional implementation of other components
i.e.. a resistor is also a capacitor and an inductor, and might even make a good RF antenna.
It matters more as the frequency becomes higher.

Digital electronics, in most part, you can ignore that except for the supporting analogue
jargon making something work. Clock crystals and their capacitors oscillation for example.

I can't convey audio that I can hear over the net to show what I want to show you,
but it's actually about an Amiga emulation on a PC that lacks Amiga characteristic after the digital output at least.

[commodorejohn]

I'm sorry, I don't understand what you're trying to say?

[pandy71]

Natural imperfections (or due HW limitation or due physics law) sometimes can be desired, sometimes not - Amiga sound can be relatively easy modeled as schematics are well known (analog part) and digital part is known up to some point (with some educated guessing we can assume hat it is known fully). It is a bit difficult to emulate additional distortions related to way for example how OPAMP are biased (connection to VCC/2 even filtered can be source of noises, also PCB layout can be far from optimal).

[Loedown]

Originally Posted by mc6809e

And that to me is why the Amiga audio hardware shouldn't be called sample based.

With no on board sound generation hardware it can be nothing but sample based, the only difference being that you need to program the sound as a series of numbers versus using an ADC to do exactly the same thing.

[Loedown]

Originally Posted by xArtx

Where did you find that? I found it for the first time this year.
It's not only not an ideal implementation, but also an unintentional implementation of other components
i.e.. a resistor is also a capacitor and an inductor, and might even make a good RF antenna.
It matters more as the frequency becomes higher.

Digital electronics, in most part, you can ignore that except for the supporting analogue
jargon making something work. Clock crystals and their capacitors oscillation for example.

I can't convey audio that I can hear over the net to show what I want to show you,
but it's actually about an Amiga emulation on a PC that lacks Amiga characteristic after the digital output at least.

Even resistors change their value over a period of time and can go out of tolerance if given long enough. A musician told me a while back that some TR808 drum machines are highly desirable because they have something subtly different to most of the others. I myself have a JX-3P that when I serviced I didn't retune the oscillators because I am used to them and I like their minor detuning.

The sound on a PC may just be too clinical for your ears, every Amiga would sound slightly different because of the age of the capacitors and the voltage levels and the amount of ripple being let through just to give a few examples. Looking at a data sheet for any analogue chip shows the tolerances of certain batches ( deviations ) and when you have a synth that has say 30 op amps in it then you are adding up and subtracting all the gain imperfections throughout the signal path. That's not even taking all the other components into consideration or thermal conditions. But exactly why analogue synths are usually left on for 45 minutes to an hour prior to playing so that they stabilize before tuning.

[xArtx]

There are some fancy capacitors, that are not electrolytics or wax,
that did not change their value over time too much
(not the top polyester one)

[Loedown]

You need a hobby

What are they from anyway?

[Thorham]

Originally Posted by Loedown

With no on board sound generation hardware it can be nothing but sample based

Indeed. You fill a bunch of chipmem and tell the chipset to start playing from that location at a specified sample rate, and the hardware plays the sample. There's no waveform generator, no nothing. Pure samples and sample manipulation, that's it.

[mc6809e]

Originally Posted by Loedown

With no on board sound generation hardware it can be nothing but sample based, the only difference being that you need to program the sound as a series of numbers versus using an ADC to do exactly the same thing.

There's plenty of sound generation hardware. DACs, PWM volume control, mixing voices together to form a stereo pair, pitch control registers, FM and AM modulation options -- this is all part of the sound generation hardware.

The only thing not in hardware is the waveform itself -- and that's a good thing!

A sample is a measurement of a signal. If the waveform comes from measuring a signal and storing samples into a memory then that's sample-based synthesis. If that waveform is derived analytically, then it isn't a sample. It's another form of synthesis -- wavetable synthesis.

[Loedown]

Originally Posted by mc6809e

There's plenty of sound generation hardware. DACs, PWM volume control, mixing voices together to form a stereo pair, pitch control registers, FM and AM modulation options -- this is all part of the sound generation hardware.

The only thing not in hardware is the waveform itself -- and that's a good thing!

A sample is a measurement of a signal. If the waveform comes from measuring a signal and storing samples into a memory then that's sample-based synthesis. If that waveform is derived analytically, then it isn't a sample. It's another form of synthesis -- wavetable synthesis.

Wavetables have been discussed earlier in the thread and whilst it is seen as its own synthesis method it's still sample based. AM involves adjusting the value of the currently played sample value either up or down by amount x and is therefore computed, FM adjusts the speed register up or down by amount y and is therefore computed. PWM isn't used, volume control is just a value in a register which adjusts an internal amp to Paula I would suppose and mixing voices together would be done post DAC with a mixing amplifier. Not one single element of this is sound generation but sound manipulation, effects and nothing more. To have sound generation you need an oscillator and there aren't any.

A sample is made using an ADC to transfer values to ram over time. The only difference in the Amiga is that you are putting the values into ram, skipping the ADC but doing exactly the same thing. Sampling refers to this technique but the remains in ram is the sample itself.

[pandy71]

Originally Posted by Thorham

Indeed. You fill a bunch of chipmem and tell the chipset to start playing from that location at a specified sample rate, and the hardware plays the sample. There's no waveform generator, no nothing. Pure samples and sample manipulation, that's it.

I disagree - wavetable is a way how you use your sample hardware, not how you create samples - samples are usually or synthesized (mathematical modeling) or recorded (natural sources).
You can easily simulate for example granular synthesis by quick change between DMA start buffer and this have nothing to way how you fill your buffer.
As there is no difference between hardware sine implementation and software sine generation unless sine produced by both methods is exactly the same.

Originally Posted by Loedown

AM involves adjusting the value of the currently played sample value either up or down by amount x and is therefore computed, FM adjusts the speed register up or down by amount y and is therefore computed. PWM isn't used, volume control is just a value in a register which adjusts an internal amp to Paula I would suppose and mixing voices together would be done post DAC with a mixing amplifier. Not one single element of this is sound generation but sound manipulation, effects and nothing more. To have sound generation you need an oscillator and there aren't any.

Paula can perform AM modulation (complex by using one channel samples to modulate second channel), similar for FM, complex FM can be performed by Paula.
PWM is used to adjust level (it was confirmed by observing Paula output, also some block schematics shows partially presence of PWM and it is almost directly exposed to programmer as a bit 6 in each AUDxVOL to turn off PWM - for bits 0-5 set as 1's maximum audio level is 63/64 which is typical for PWM a 0 is no level at all).

Also HRM mentioned other methods for audio generation - CPU driven and as a PWM driven.

[Loedown]

Originally Posted by pandy71

I disagree - wavetable is a way how you use your sample hardware, not how you create samples - samples are usually or synthesized (mathematical modeling) or recorded (natural sources).
You can easily simulate for example granular synthesis by quick change between DMA start buffer and this have nothing to way how you fill your buffer.
As there is no difference between hardware sine implementation and software sine generation unless sine produced by both methods is exactly the same.

The wavetable contains chunks of data which are samples, how is it any different?

[pandy71]

http://harderstate.com/tutorials-f22...ed-t22348.html

wavetable doesn't describe way how sample bank content is created - wavetable describe way how sample banks are used - so with CPU/Copper Amiga can be simple sampler or wavetable or granular or related way synthesizer - it is up to software to use available HW.
wavetable can use both -synthetic and recorded samples to form complex sound wave.

[xArtx]

Originally Posted by Loedown

You need a hobby

What are they from anyway?

I need yet another hobby?
http://www.youtube.com/user/ArtyFart

I forget exactly which vintage amp/receiver, but it was a junker.
If the discreet parts are well in spec, I'm keeping them in case I ever get hold of
something so good, that it's not appropriate to use modern counterpart components in it.

[Loedown]

Originally Posted by pandy71

http://harderstate.com/tutorials-f22...ed-t22348.html

wavetable doesn't describe way how sample bank content is created - wavetable describe way how sample banks are used - so with CPU/Copper Amiga can be simple sampler or wavetable or granular or related way synthesizer - it is up to software to use available HW.
wavetable can use both -synthetic and recorded samples to form complex sound wave.

Again this is another technique but ultimately Paula plays nothing but samples and what you describe here is front end or add on which can be implemented on the Amiga. Clever programming can achieve many things but ultimately it's just a sample player.

[pandy71]

Originally Posted by Loedown

Again this is another technique but ultimately Paula plays nothing but samples and what you describe here is front end or add on which can be implemented on the Amiga. Clever programming can achieve many things but ultimately it's just a sample player.

once i must disagree - going this way we can say that CD disc is nothing else than sampler as sample buffer is played trough DAC.
Even analog modeling synth are using buffer to construct sample which is turned to electric current by some DAC.

Amiga have RAM as sample buffer and it can be freely manipulated thus it provide required amount of flexibility to be something else than plain sampler.
It is only limitation of programmer how to use simple hardware - yes, Amiga audio HW have no other HW algorithm implemented than AM and FM modulation and low pass filter that can be switched OFF/ON.
But still in my opinion it going over simple sampler definition (as typical simple sampler will not provide FM functionality).

btw
is 1 bit is enough?
[ Show youtube player ]
I think that Future Composer is not sampler focused tracker...
[ Show youtube player ]
[ Show youtube player ]

Simply most of Amiga music software was sampler focused as sample based techniques was not available so widely before Amiga.

btw
How to classify this way of sound generation:
sample due of HW or not? [ Show youtube player ]
4Kb seem that there is no space for audio samples... but HW is even more dumb than Amiga (SRC is performed mostly by CPU on PC world)

[Mrs Beanbag]

We're splitting hairs here, it's only a word game to argue about whether it's a "sample" or not if the data has been procedurally generated rather than recorded. Anyway what if it was synthesised on another computer and sampled onto the Amiga? I can see no point in this line of argument.

The Amiga's sound generation is purely digital up until the DAC. It is not possible to generate any analogue wave. That's all that matters. Of course, square waves are easy.

But I gather some emulators and mod players "interpolate" the samples, reducing aliasing and giving what is in theory a higher-quality sound, but losing something in the translation.

In addition, Paula doesn't run at the 44.1kHz that a modern sound card resamples everything to, so there is bound to be a difference. Perhaps this is noticeable?

[Amiga1992]

The matter is the Paula itself generates nothing. Be it a sampled or cpu-generated waveform, there's still NO SYNTH. There is no synthesizer inside the Amiga. Period. Call it what you want, but you can't call it a synth.

But DACs generate their characteristics and the Amiga has a characteristic one. You can't reproduce precisely in an emulator, but the emulator is pretty spot on.

If you wanna analyze it more in detail the VST instrument Chipsounds supports Amiga sound emulation and you can load samples and play them either in modern, interpolated way or using Amiga style lack of it. You'll notice the difference right away.

[Mrs Beanbag]

Originally Posted by Akira

There is no synthesizer inside the Amiga. Period.

Well there's the CPU

If you wanna analyze it more in detail the VST instrument Chipsounds supports Amiga sound emulation and you can load samples and play them either in modern, interpolated way or using Amiga style lack of it. You'll notice the difference right away.

Indeed but you can't play them other than resampled at 44.1kHz

[john1979]

Originally Posted by Mrs Beanbag

Well there's the CPU

Which is generating artificial samples for Paula to play. It's not a synth in the traditional, SID chip style sense.

Indeed but you can't play them other than resampled at 44.1kHz

A lot of sound cards resample all input to 48khz. The Audigy series was notorious for doing it so badly it was possible for a good pair of ears to blind test the difference. Many people made use of software resamplers like the one included with Foobar to get round this problem.

Other cards will accept input at many different rates without resampling, and some like you say require all input to be 44.1khz.