Which Amiga computer technically has the best audio quality?

[jaycee1980]

And the A4000? This has R331 = 750, C1 = 4700pF. The values for R334, R335 and E303 dont directly translate because there is a summing amplifier after the standard Amiga audio circuit, which combines input from a "CD/DSP audio" header on the board.

However, we get ~44.1KHz with filter off, ~2.6KHz with it on. Im not sure what the signal amplitude is but it wouldnt surprise me if its' around 1V p-p which is a common standard for digital audio.

Note that this higher rolloff frequency with the filter off does not mean that Paula can actually reproduce those frequencies - the Nyquist principle still stands. You simply get more aliasing.

[8bitbubsy]

I remember A4000 (both Commodore A4000D and Escom A4000T) having a low cutoff, close to an A500. Haven't tested A4000T by Commodore.
Using your values, we can calculate the nominal (1-pole RC) cutoff with this formula: cutoff = 1 / (2pi * R * C).

1 / (2pi * 750 * 4.7e-8) = 4515.03Hz. Close enough! A500 = 4420.97Hz.

Remember though, that this is for the first low-pass filter stage, not the "LED" filter you can switch on and off. Also because of the non-steep (6dB/oct) filter slope, it won't cut off everything above 4.5kHz. It allows quite a bit above that, but the treble details are still crippled.

EDIT: Fixed some misinformation about what models I tested.

[Photon]

Originally Posted by 8bitbubsy

The proper "fix" is to match the RC values with that of the A1200.

As further input, musicians who regularly contribute to Demoscene music competitions seem "end up with and prefer" the A1200 to hear what they're doing.

If we are looking for which values to use for non-A1200 machines or certain revisions, I think we could do worse than use the correct A1200 values.

Up from that could be an improvement by an audio technician, but I don't think it should be from a theoretical calculation only. I think someone used to listening for artifacts, i.e. knowing what to listen for and be able to provide tests (such as a technically skilled musician) should also be involved.

If we choose to mod our Amiga to the ultimate filter and not the A1200 filter, this would mean musicians might want to target the subtly different platform for 4-channel music. (Compare specifying SID version in C64 music compos.) I.e. to hear it as intended, you'd need the HW mod, and emulators would need a change to support it.

I think I won't mod any of my Amigas for this, however if I find the values off for my revision, I would probably use the correct A1200/A4000 values instead of the very slightly different ones for another model.

So this thread is good for awareness at the very least.

[ShK]

Was it so that the audio circuit in the Amiga 1200, which handles alternating current (AC) signals, ideally should use bipolar (non-polarized) capacitors. However, in the factory, polarized capacitors were installed?

[8bitbubsy]

Originally Posted by ShK

Was it so that the audio circuit in the Amiga 1200, which handles alternating current (AC) signals, ideally should use bipolar (non-polarized) capacitors. However, in the factory, polarized capacitors were installed?

AFAIK the raw L/R signal coming from the Paula chip are positive voltages only (DC), and the last high-pass filter stage (those electrolytic caps you are referring to) will do the job of centering the signal (DC->AC). This means that the caps are operating within normal use, and shouldn't suffer from it. I could very well be wrong here, so please correct me if that's the case. I just think they leak/die even faster than the other caps because they are under constant voltage swings. Also, all those SMD electrolytics caps on the A1200 mobos are crap in general, something was not right with those SMD caps from that period.

@Photon: Yeah. I just wrote "fix" in quotes because I don't look at it as a fix to modify the A600 sound, I find nothing wrong with it. It's as designed. I was replying to h0ffman, really.

[jaycee1980]

Originally Posted by ShK

Was it so that the audio circuit in the Amiga 1200, which handles alternating current (AC) signals, ideally should use bipolar (non-polarized) capacitors. However, in the factory, polarized capacitors were installed?

No. I'm not sure who started that one, but they misunderstood how DC blocking capacitors work. There is always a positive DC offset on the output of the opamp, therefore bipolar capacitors are not required.

[no9]

How do all those hardware filter mods compare to more common methods of balancing Amiga sound, such as using a treble boost on an amplifier or more precise equalizing? I know that boosting higher tones will result in some increase in noise levels, but is the difference significant? Is there a noticeable phase shift or something? Has anyone made a comparison of that?

[jaycee1980]

Originally Posted by no9

How do all those hardware filter mods compare to more common methods of balancing Amiga sound, such as using a treble boost on an amplifier or more precise equalizing? I know that boosting higher tones will result in some increase in noise levels, but is the difference significant? Is there a noticeable phase shift or something? Has anyone made a comparison of that?

Restoring the signal by boosting things with an EQ also amplifies the noise coming from the machine. Better to take care of it at the source.

[ShK]

Originally Posted by 8bitbubsy

AFAIK the raw L/R signal coming from the Paula chip are positive voltages only (DC), and the last high-pass filter stage (those electrolytic caps you are referring to) will do the job of centering the signal (DC->AC). This means that the caps are operating within normal use, and shouldn't suffer from it.

Originally Posted by jaycee1980

No. I'm not sure who started that one, but they misunderstood how DC blocking capacitors work. There is always a positive DC offset on the output of the opamp, therefore bipolar capacitors are not required.

But is it so that Paula chip outputs a DC-biased signal and the nature of the high-pass filter stage in the audio circuitry means that the capacitors do handle AC signals post-filtering? This exposure to AC signals can lead to issues with polarized capacitors, such as leakage or failure, over time.

[ Show youtube player ]

[ Show youtube player ]

[ShK]

Not sure how accurate is silk screen's specification of "Electrolytic 22µF-NP" for C324 and C334 in Amiga PCB Explorer. The "NP" designation stands for "Non-Polarized," indicating that these capacitors should be non-polarized electrolytic capacitors.



https://www.amigapcb.org/

[no9]

Originally Posted by jaycee1980

Restoring the signal by boosting things with an EQ also amplifies the noise coming from the machine. Better to take care of it at the source.

I asked how much it increases and if anyone had compared it.