Do all digital amps sound this way? What even is it? Mystery solving.

Discussion in 'Recording Studio' started by JohnIce, Nov 14, 2018.

  1. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    So, I've been hearing some annoying things in my Kemper for a while. It's a sort of phasing issue in the top end as the note decays. I've tried different guitars, different cables, different headphones, nothing helped.

    Then I tried my old Axe-Fx Std. Different, but similar. Tried demos of ReAxis, Spark, Logic amps, and Toneforge. Again, slight differences but all of them have this phasing, kind of bitcrushy bullshit as the note decays.

    Thinking it STILL might be a problem with cables or my interface, I got the bright idea to try using a software instrument instead, bypassing my interface and anything electrical in my apartment, working strictly in the box. So this is a virtual electric piano into Toneforge. Listen to the way the distortion kind of swooshes and disappears halfway along the decay of the note (at about half or 25% volume to the electric piano, it sounds a bit tucked into the background). The lower the velocity/amplitude of the note, the shorter the swoosh:
    https://soundcloud.com/jntzero/dist-test/s-NV5bP

    ^You might need headphones or decent volume to hear it, hopefully you can hear what I'm hearing.

    I notice that the more gain you use, the longer it takes for this swoosh to manifest itself, meaning that if you play a typical metal chug riff you're likely to not hear it. But if you play a dynamic, delicate solo with more of a crunch tone, it can sound like absolute garbage to me.

    So my questions are:
    a) What is this audio phenomenon called? I've heard of phase distortion (which is what I'd call the sound, but maybe it's something else).
    b) Is there ANYTHING one can do to get rid of it? Like I said, Kemper, Axe all the way down to Logic's generic Distortion plugin react the same way.
    c) Do ANALOG distortion pedals behave this way too, and I've just never noticed it?
     
  2. lewis

    lewis SS.org Regular

    Messages:
    5,370
    Likes Received:
    1,735
    Joined:
    Aug 16, 2011
    Location:
    Norfolk, UK
    Could also be present in tube amps too. Therefore, its in everything and is just what happens.? I notice more fizz in the kemper than the Axe II though for sure.

    lol

    I did give the clip a listen through monitors and didnt really notice anything going on that was going to give problems.
     
  3. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Try headphones, you should be able to hear it. It's not a "problem" so much as it sounds a bit unpleasant and artificial I think. It mainly bothers me when I'm playing unaccompanied by myself, which is every day though.

    I tried some Kontakt libraries where instruments were sampled with distortion already on them and I don't hear this type of decay. But if I turn on whatever distortion is built into Kontakt, it introduces it again. So I'm really leaning towards the idea that this is kind of a universal attribute of digital distortion, sort of like how stepping in filters is an unavoidable part of digital synths vs. analog ones.
     
  4. Drew

    Drew Forum MVP

    Messages:
    27,721
    Likes Received:
    3,084
    Joined:
    Aug 17, 2004
    Location:
    Somerville, MA
    I'll check this out tonight. :yesway:
     
    JohnIce likes this.
  5. tedtan

    tedtan SS.org Regular

    Messages:
    4,231
    Likes Received:
    705
    Joined:
    Dec 2, 2009
    Location:
    Never Neverland
    This is called quantization distortion and it happens as the number of bits drop below a certain point as the note decays.


    Yes - use analog gear. Unfortunately, analog has it's own noise floor, too, but it is much better in this specific regard IMO.


    No, this is a function of digital bit depth and is only noticeable when the bit depth decreases as the note fades out.
     
  6. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    That is incredibly helpful, thanks a lot! That makes a ton of sense too, I don't know much about bit depth or bit reduction but I do relate the sound of bitcrushers to what I'm hearing here. For now I'm just really happy to know what to google for, thanks to you! It's nice to find out something objective in the analog vs. digital discussion for a change.
     
  7. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Reading more about it, I tried the same test with and without sending white noise into the plugin. It took a lot of noise to get there but it did eventually make the quantization distortion inaudible. Cool, learned something about dithering in all of this :lol: I guess from now on if I want to play softly on the Kemper I'll have to ask my drummer to work the ride cymbal to dither me. https://soundcloud.com/jntzero/dist-test-dither/s-LjGEV
     
  8. AirForbes1

    AirForbes1 SS.org Regular

    Messages:
    161
    Likes Received:
    57
    Joined:
    Jan 13, 2015
    Location:
    Toronto, ON.
    You can hear the little phase thing even on laptop speakers.

    In the Helix, they call it squirrels I believe (if I'm equating the term with the correct phenomena). It sounds different to what you have. On the Helix it kind of sounds like tearing paper as the note decays. I've only heard it with Marshall type models and on the "edge of breakup" kind of settings.
     
  9. tedtan

    tedtan SS.org Regular

    Messages:
    4,231
    Likes Received:
    705
    Joined:
    Dec 2, 2009
    Location:
    Never Neverland
    I never thought about that approach, but it seems like it would work.
     
  10. Drew

    Drew Forum MVP

    Messages:
    27,721
    Likes Received:
    3,084
    Joined:
    Aug 17, 2004
    Location:
    Somerville, MA
    Ok, I didn't get to listen to this last night, but...

    ...shouldn't quantization distortion only be an issue if you're changing bitrates somewhere along the way? If so, shouldn't you be able to bypass this by just tracking your Kemper in the same bitrate as it's functioning...?
     
  11. Flappydoodle

    Flappydoodle SS.org Regular

    Messages:
    717
    Likes Received:
    532
    Joined:
    Feb 26, 2018
    If you increase your interface input gain, the note will sustain longer and you won't hear it?
     
  12. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Can't really find any info on the Kemper's bitrate, unless you mean bit depth which is 24 on both the KPA and Logic. These examples weren't done with the Kemper though, I used Logic's own instruments, which I assume would be optimised in every way with Logic itself.

    I also plug my Kemper into my Focusrite and directly out to my monitors most of the time, and while I can't find any bit depth toggles on my Focusrite I read its spec sheet to be 24 bit.

    The input gain is plenty high, I'm just playing very softly.
     
  13. Flappydoodle

    Flappydoodle SS.org Regular

    Messages:
    717
    Likes Received:
    532
    Joined:
    Feb 26, 2018
    I am not able to replicate your finding.

    Les Paul into Apogee Duet into MacBook Pro running Logic

    Four audio samples:
    1. DI
    2. High gain (Fortin Nameless, ML "best IR in the world)
    3. Low gain (Turned Nameless gain to 0)
    4. Logic "Steinway Grand Piano" instrument playing C in 3 octaves, and two discordant notes

    Accessible here: https://www.dropbox.com/sh/43k1kgpds7c2ieg/AAAZmS_WOxoSOD7_C1yEh2qCa?dl=0

    I turned the noise gate off in Fortin, as you can surely tell, otherwise the note is cut off

    I let several notes fade to nothingness and I can't hear those artefacts

    Everything in Logic is stock - 24 bit recording enabled

    Perhaps try my DI in your setup, or advise how I can replicate this for myself
     
  14. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Thank you so much for doing that! :yesway: It's hard to tell if I really hear it in your examples, I'm not sure I do. My Nameless demo just expired, that's a bummer.

    My Kemper isn't set up for reamping right now, but I did try your DI through Reaxis and Toneforge and they definitely had the artifacts going.

    For the sake of comparison, as we both have Logic, would you mind running your DI through Logic's amp sim on the "A-List Clean" preset, full gain and no reverb? When I did, something very interesting happened in the noise of your DI. At about 5 seconds in (on my bounce), there's a serious bitcrushing swoosh happening, among other things. https://soundcloud.com/jntzero/flappydoodle-di-test/s-48mRT It's quiet, so try headphones.

    I'm asking purely to understand what this phenomenon is a little better and rule out any problems in my setup, I'm normally not advocating for the use of Logic's amp sims ever :lol:
     
  15. Flappydoodle

    Flappydoodle SS.org Regular

    Messages:
    717
    Likes Received:
    532
    Joined:
    Feb 26, 2018
    Sure, I'll do it for you. Dumb question - how can I download the DI from Soundcloud? I can play it in the browser, but don't see an option to download.
     
  16. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Oh sorry, I meant your own DI, which is what I used for my last clip :)
     
  17. Drew

    Drew Forum MVP

    Messages:
    27,721
    Likes Received:
    3,084
    Joined:
    Aug 17, 2004
    Location:
    Somerville, MA
    Gotcha, I thought you were tracking the Kemper into Logic. Google suggests its 44.1khz, though, so if you were using the Kemper and tracking at something other than that (especuially at something not an even multiple of 44.1), that could potentially be a source of trouble.
     
  18. Descent

    Descent SS.org Regular

    Messages:
    823
    Likes Received:
    119
    Joined:
    May 7, 2014
    Location:
    Houston, TX
    That's why I still can't shake off my use of analog tube gear. Every time I hear those digital artifacts of a high gain tone fading out I run back to my Marshall, Mesa and Orange amps. One of the best and easiest ways to pick digital modeler in an audio test.
     
  19. JohnIce

    JohnIce Singlecoil Enthusiast

    Messages:
    5,117
    Likes Received:
    1,954
    Joined:
    Apr 4, 2009
    Location:
    Gothenburg, SWE
    Yeah I thought something like that could be the case, which is why I was surprised (and a little relieved) to see the same results using a virtual piano in the box. That said, I did actually switch my interface to 88.2 and though it may be wishful hearing, I might think the Kemper sounds a little bit smoother in the high end now. I'm using TRS cables rather than S/PDIF so maybe my interface at 44.1 is slightly out of sync with the Kemper at 44.1, and using 88.2 makes up for that? This is pure speculation on my end, I know very little about how this works and that theory could be dead wrong.

    I find it interesting though that after so much critical listening to digital guitar tone, I'm actually sometimes hearing what I think are digital artifacts in tube amp recordings. For example, this intro has all the warble and crackling in the high end that I associate with digital modelers, yet it is as far as I know a JMP.


    ^If you asked me now I would swear that was a Kemper, but that record came out in '97 :lol:
     
  20. Descent

    Descent SS.org Regular

    Messages:
    823
    Likes Received:
    119
    Joined:
    May 7, 2014
    Location:
    Houston, TX
    It could be badly dialed noise gates as well, or digital fx noise gates. This might be some gonkulator or bit destroyer or something like and a Fender single coil or out of phase pickup, that for main sound so they're destroying the tube sound somehow on purpose.

    The fact that they said they used a certain amp doesn't mean they actually did. Their producer might've used something else entirely, and their endorsement says they'd have to say used amp X.
     

Share This Page