A crazy theory about analog/digital

[Tepeix]

This is a crazy theory with no proof at all !)
(Just subjective and naive feel, but is this possible ???)

Using overdrive and distortion as guitarist, i never feel like they add false note.
But now doing some experiment with digital overdrive, i see that there's a lot of intermodulation.
Even without aliasing, when playing more than one note, the notes will ring mod with each others.
Doing some false notes (or harmonics) with some intervals.

The same must occur in analog world, or an overdrive will only be an harmonic fixed pitch.

But i have the impression that in analog intermodulation will be less present and some harmonics
that occur with one note are more preserved when playing more than one.

Is it possible that in the analog world, electrons (or others particles)
will not add ass only one wave (sum off all frequency) but only a part
of them ?? Some particles following a path that is more attracted by one frequency ??

In others world each particle could follow all the frequency or only a part of them ?
Making a difficult limit in digital to do the same if it's true, as we have to extract all frequency,
do the process with each of them individually and with the sum, then adding all ?
And that's maybe why multiband overdrive is more used in digital ?

Héhé that's just a theory that come in my mind.
Maybe the better sound of analog overdrive in this case come from another behavior
that reduce the sound of ring mod intermodulation ?

[Spogg]

... the better sound of analog overdrive...

To judge the difference between analogue and digital overdrive you would surely have to have a supremely accurate model of an analogue circuit in DSP form, otherwise the two will never sound exactly the same. I think it’s too much of a generalisation to say analogue is better, or digital is better for that matter.

In my view, if you can make an exact DSP replica of the transfer function of an analogue process it should sound and behave exactly the same way, irrespective of the technology employed.
Of course, making an exact replica is the tricky bit!

[Tepeix]

I hope you are right in this
It's true that they are some simulation that come very close.
I could maybe never hear the difference.
Well, i never really try with recent device.. (best i have is the boss gt10)
I have tendency to thing that maybe we could feel it playing live, but not so much with a record.
Maybe the latency ?
I begin to have more doubt about my theory..

But there's also the psychological theory.
Knowing that the process is digital, could add subjectivity and make play a little differently for some ?

I don't try for now to replicate, i want more to have some soft overdrive to add flavor, sounding good but not specially like something that exist. But maybe i must understand better what happens in analog.

Héhé, maybe i find even more unprovable theory : In analog we are more merged in the process because the most subtle energy and frequency are flowing with no limit. The sound preserve is quantic part and could access directly to our soul. (where material become energy then virtual infinity... ...)
When using digital, playing live, the sample rate reduction and bit reduction doesn't change the audible sound but we could feel it and it could affect emotion...
That's a little paradoxal because the soul doesn't need to have a feedback sound but the player could like to have it.. ...

Well that's crazy..
But psychological or true sonic taste i always organize my pedalboard to mix all the digital process in parallel.
That could make some latency problem.. To avoid them i use more than one digital unit in series to increase the latency. (Maybe it doesn't fix all)

[Spogg]

I’m inclined to agree that cognitive bias comes into it to a great extent. The only real way to rule that out is to not know what you are using, so you are able to make a “blind” judgement. Then it would come down purely to the sound you actually like, rather than being influenced by knowing the process that leads to it.
Of course you need someone else to set up a system for comparison and the truly scientific method would be “double-blind” so the person who offers the alternatives also doesn’t know what you’d be listening to. The double-blind method eliminates unconscious clues your assistant may send you.

But it’s not just about the sound itself. I would say that creativity could be affected by something as innocent as the colour scheme or even what you paid for the effect!

In the end I think what matters is what you like and what inspires you.

[tulamide]

From a programmer's perspective, there's a simple, yet drastically influencing, difference. digital is precise, analog imprecise.
For example, when in digital you can often do something like
if x == 0
in analog you ALWAYS have to do something like
if x somewhere around 0 (as you can see, there isn't even a term for it in programming languages)

In essence, one could say that the faulty-ness of analog actually helps covering issues, that stand out clearly in precise realms like digital. Wether that's good or bad is in the mind of any individual.

[Tepeix]

Thanks for the precision about imprecision !

I think god and bad together.)
The paradox is i'm always fighting to get a "perfect clean" sound in analog.
But if the sound is really clear it lacks some strength (in my subjective hearing)..
So i like very subtle overdrive.
(I buy 5 danelectro transparent overdrive v2 and broke 4 of them.
With the gain very low i even thing that this pedal strangely clean the signals.)

Now that i go further in coding i understand more what happens when trying to get a "perfect clean" in analog.

Maybe analog could be best when we don't already have the simulation or cpu power.
But for sure in linear digital is better.

I imagine a sound cart where we could send some process to an analog bus for easy clipping

[Spogg]

Something else about lack of precision in the analogue world is that if you can get a super-accurate replica of an effect, it may not sound exactly the same when compared with another example of the same effect. This is down to component tolerances, temperature, variation in semiconductor behaviour and component ageing. Production runs may well use the components from the same batch of transistors and so on, but there can be variations between batches. In critical parts of the circuitry designers often had to make extensive use of trimmers to compensate for this. I have memories of this myself and the same circuit would end up with different settings for the trimmers to get the performance of the circuits to match. And of course over time the trimmers would need re-adjusting to return to the original. I suspect (without knowing for sure) that many analogue effects don’t use trimmers, so the variations between the same models could be more apparent.

There is no similar constraint in the digital world, as tulamide said, which means that a digital effect will always sound the same, assuming the ADC and DAC are of modern quality. But it may not sound exactly the same as every example of analogue hardware being emulated.

[Tepeix]

I even find someone that try to replicate this variation component in digital:
https://jatinchowdhury18.medium.com/bad ... fdbe4e980c

Could be interesting but, i prefer an accessible control (or maybe a random array that the user could reset).
Maybe the more interesting idea is to have slight difference in left and right channel.

[Spogg]

I even find someone that try to replicate this variation component in digital:
https://jatinchowdhury18.medium.com/bad ... fdbe4e980c

That’s an excellent and detailed article!

I seem to recall that tolerance coding is done by testing the components individually (or quality controlling a batch), then labelling them accordingly. As per his graph, a 1000 Ohm resistor with a 10% tolerance will never be less than 5% away from 1000 Ohms. So if a 1000 Ohm resistor measures 1019 Ohms it will be banded as 2% tolerance (and it’ll cost more as a result).

I haven’t thought about this stuff for years!

[Tepeix]

If you click on the name of the author they are a lot of interesting stuff.
(The site could make thing that we need some subscription but this way we could access it)
I get my inspiration from him to try some non linear filter.