Kemper Amps Forum

Thanks DHodgson, that helps explain it very clearly. OK so we are talking about the same effect.

The issue is that this is as far as I can tell a natural byproduct of compression and clipping (i.e. what makes up guitar amp distortion) rather than "digital aliasing", or at least of compression with a very short attack and decay, I actually get the same effect with my real amp (you can hear it in the demo clip I posted a couple of posts back).

What's happening is that with either clipping or a very short compression attack (and to extent release) you start turning your nice smooth sine wave into a nasty jagged square wave. This results in many more frequencies, harmonics based on the compression (how square the wave becomes basically) which sound like background notes at different frequencies rising and lowering together or against the main note. It's the same stuff as the "numerical noise" thread where I showed the spectrograph of what happens when you compress a signal, i.e. you get the same sounds or "artifacts" and it doesn't matter what the sampling rate is.

I'm not near my KPA right at the moment so I can't do this, but I wonder if anyone could re-amp that sine wave through their real pedals/amps for comparison (would need to be set to similar distortion levels as the patch in question).

I'm not sure how you can avoid except to artificially reduce compression attack and clipping, which would result in a smoother sound but a loss of the real distortion character of the amps and pedals in question. And to my ears that's much more fake and plastic or synthetic sounding, one of the horrible "tells" of modelers that kept me away from ever seeing them as a viable alternative before the KPA.

Real amps are gnarly, we've all got cheap VST's and old Pod's and BOSS units to do the plastic fusion guitar sound.

Quoted from "MadH"

Do you think that the more pronounced aliasing effect in the pedal block could be due to the pedals block running at a different internal sampling rate than the amp block? I think the amp block tube saturation simulation runs at 700 khz but maybe the pedal block do it at 44.1 khz. Maybe what I am saying does not make sense at all but that could explain as well why the fluctuation on the amp block is much faster and less noticeable compared to the one produced by the pedal block.

Spot on, MadH, I was thinking the same thing. Some or all of the Amp block is probably running at 16x oversampling, while the rest of the FX chain probably runs 44.1 kHz. Nothing wrong with this, it's a good call though the distortion pedals (and to a much lesser extent, any compressors) will suffer a bit.

Quoted from "Per"

Thanks DHodgson, that helps explain it very clearly. OK so we are talking about the same effect.

The issue is that this is as far as I can tell a natural byproduct of compression and clipping (i.e. what makes up guitar amp distortion) rather than "digital aliasing", or at least of compression with a very short attack and decay, I actually get the same effect with my real amp (you can hear it in the demo clip I posted a couple of posts back).

What's happening is that with either clipping or a very short compression attack (and to extent release) you start turning your nice smooth sine wave into a nasty jagged square wave. This results in many more frequencies, harmonics based on the compression (how square the wave becomes basically) which sound like background notes at different frequencies rising and lowering together or against the main note. It's the same stuff as the "numerical noise" thread where I showed the spectrograph of what happens when you compress a signal, i.e. you get the same sounds or "artifacts" and it doesn't matter what the sampling rate is.

I'm not near my KPA right at the moment so I can't do this, but I wonder if anyone could re-amp that sine wave through their real pedals/amps for comparison (would need to be set to similar distortion levels as the patch in question).

I'm not sure how you can avoid except to artificially reduce compression attack and clipping, which would result in a smoother sound but a loss of the real distortion character of the amps and pedals in question. And to my ears that's much more fake and plastic or synthetic sounding, one of the horrible "tells" of modelers that kept me away from ever seeing them as a viable alternative before the KPA.

Real amps are gnarly, we've all got cheap VST's and old Pod's and BOSS units to do the plastic fusion guitar sound.

Per,

It's true that compressors (and automation, too) introduce non-linear behavior that may add some new harmonics and/or DC bias, it's only when implemented in the digital realm that you open the door to the same types of aliasing problems you see with digital clippers - just weaker. Heck, old-skool guys still use analog compressors to add a bit of grit to bass by setting their attack and release times faster than the bass fundamental, effectively turning their compressor into a waveshaper! The difference though is that this type of intentional waveform deformation follows the rules of harmonic (and to some extent, intermodulation), distortion which makes sense in a musical, harmonic-series kinda way.

But more to the point, what we're talking about here - the stuff that sounds like ring modulation artifacts is indeed digital aliasing distortion. The dead giveaway, besides the obvious unmusicality of it, is that you can hear foldover; as the fundamental is bent up or down, the pitch bears a nonharmonic (and often opposite) wraparound relationship to the fundamental that behaves like the wagon wheel effect. The KPA's distortion components are fine, to the extent that they're operating in-band (which they mostly are) - but it's the additional alien weirdnesses that we're talking about, and these really have no counterpart in the analog world.

-djh

dhodgson, would you mind sending your findings and soundclips to support?
This is all very interesting.
Thanks for your audio clips btw!

Quoted from "schneidas"

dhodgson, would you mind sending your findings and soundclips to support?
This is all very interesting.
Thanks for your audio clips btw!

Glad to be of help. I can pass the thread along, but I'm pretty sure the Kemper engineers are already aware of the key stuff here and are grinding their teeth for having attention called to it. I know I would be.

-djh

In case anyone cares at all I had this type of aliasing which I thought was SPDIF problems with my rig.. turns out it was actually the Kemper has some corrupted internal memory settings. After a memory wipe (formatting the flash memory) and reinstalling the factory rigs and my personal favorite rigs, my kemper sounds fantastic and more importantly I can't detect the aliasing anymore on any notes.. I don't understand how this can be but it most certainly is. Maybe it points someone in the right direction.

Quoted from "mikeb"

In case anyone cares at all I had this type of aliasing which I thought was SPDIF problems with my rig.. turns out it was actually the Kemper has some corrupted internal memory settings. After a memory wipe (formatting the flash memory) and reinstalling the factory rigs and my personal favorite rigs, my kemper sounds fantastic and more importantly I can't detect the aliasing anymore on any notes.. I don't understand how this can be but it most certainly is. Maybe it points someone in the right direction.

Would it be possible to have someone else (dhodgson maybe) who has experienced the same antialiasing issue to confirm Mike's findings?

Radley should try this as well, noise and aliasing are his main problems...

I'd need a link to the v1.04 firmware to test mikeb's method. Mike?

-djh

Quoted from "dhodgson"

I'd need a link to the v1.04 firmware to test mikeb's method. Mike?

-djh

You can find older FW-versions on KPA websites Support section (starting from 1.0.0)

Choose "older operating system versions" from the category dropdown menu.

Here's 1.04

http://dl.dropbox.com/u/55111784/kaos.bin