Latency of the KPA

  • There's been talk about the latency of the KPA - Although I won't have my Kemper til next week, I have noticed some pretty obvious latency on some of the YouTube demos where you can hear the acoustic attack from the electric guitar clearly ahead of the sound coming from the KPA.... Anyone else?

    All modelers known to man 8o

  • That's mainly because the mic of the camera is in front if the guitar and farer from the loudpeakers

    "Music is enough for a lifetime, but a lifetime is not enough for music" Serghei Rachmaninoff


  • There's been talk about the latency of the KPA - Although I won't have my Kemper til next week, I have noticed some pretty obvious latency on some of the YouTube demos where you can hear the acoustic attack from the electric guitar clearly ahead of the sound coming from the KPA.... Anyone else?


    I've had mine for 2 days, and the playing feels ok to me, sometimes it feels a bit strange but it could be the noise gate, I didn't measure it, but it feels about the same as the Axe II, imagine it can't be much more then 2-3 ms, but if the noise gate isn't set right for the guitar you're using, it feels like more latency.

  • I have the KPA from one week VERSION 1.0.4 BETA, I notice a very little latency in some stressing clean pattern with strong attack.


    Then I measure it with a test (to see if it was subjective or not), I put the Guitar with same line in into BOSS GT-10 MultiProcessor and contemporary in the KPA LINE IN and with the same AUDIO CARD (MBOX PRO) I registered both OUTPUT from GT-10 and KPA.


    So, KPA has 6 ms more latency than the BOSS GT-10. If you sum on it 1-2 ms to reach the KPA and other ms added from the sound card it can be noticeable.


    I believe it is a compromise that we have to accept for such real digital modeling, but anyway I hope that Mr. Kemper will work with his team to reduce as much as possible in some way this 6 ms latency more than other multieffect consolle.


    Maurizio70

  • Negligible. I came up through old school r'n'b and funk. Dry, tight, in the pocket. So far as I've noticed with anything like that there's nothing amiss. I think those with latencies need to look elsewhere for the culprit.


    60 thousand notes a second sweep playing might give a different result. :D


    If the noise gate is over zealous there'll be a noticeable ramping up of the volume which cuts off the attack and gives the impression of latency.

  • I believe there are few imprecisions in this kind of tests: when you increase the number of tracks you increase latency as well (from both, audiocard converters and DAW). The only way to make a precise measure would be to send a test signal through the unit and measure when it hits the output, without an extra audiocard with A/D converters, pc and DAW. IMHO this is not accurate

    "Music is enough for a lifetime, but a lifetime is not enough for music" Serghei Rachmaninoff


  • The only way to make a precise measure would be to send a test signal through the unit and measure when it hits the output, without an extra audiocard with A/D converters, pc and DAW.


    I agree. A simple poor man's way of measuring latency would be to a use a single sound source, preferably an extremely short attack transient, split the signal and send one signal to left side of a two channel audio I/O, the other through the KPA and into the right side. Record the sample.


    Assuming the audio interface has no difference between the L/R channels the difference would be measurable as the difference in time of the two audio tracks. This negates measuring any latency through a DAW, interface etc.


    It should be fairly easy to measure the KPA latency down to the sample this way. To determine the exact difference within the context of a particular sample rate flip one track 180 degrees out of phase and nudge it until 100% phase cancellation appears. The amount that track is nudged is the latency.


    From what I understand 1ms is approximately the same as sound traveling about the distance of 1 foot. So for example 5ms would be similar to standing 5 feet away from the sound source.


    Disclosure: I'm not an scientist of any kind. If there is a flaw in this let me know.

  • Connect a cable output to input of your sound card. Send a signal thru the cable and record the return.
    Save the record and open it with audacity. Zoom (ctrl+1) and check when begins the signal. The unity is 1 ms so you should see what is the latency of your daw + soundcard, this is the space at the beginning before the signal appears.


    Now connect the same cable to the output soundcard -> input of the KPA then the output of the KPA -> input soundcard. Send a signal and record the return.
    Save and check into audacity when the signal begins.
    Then substract time 1 from time 2 and you will get the latency of the KPA. IMO


  • Why would you measure latency with an Mbox? Those have inherent high latency as it is. I don't know why there are so many people talking about the latency....the original 2ms figure is correct. Try it with with either headphones or direct monitoring it will be proper. Conversion adds latency, buffer sizes add latency. You must remove those from the equation to have a proper reading.

  • Why would you measure latency with an Mbox? Those have inherent high latency as it is.


    if you record both units with the same soundcard, the soundcard's latency doesn't matter as long as you compare the relative difference... but i agree that 6ms is a bit too much adding it to the boss unit's latency.

  • Main page of the Kemper homepage.
    Log in. Click on your "logged in as" name - you're in your profile now.
    Click on the "register your amp" tab. You can also upgrade your membership status there...

  • Main page of the Kemper homepage.
    Log in. Click on your "logged in as" name - you're in your profile now.
    Click on the "register your amp" tab. You can also upgrade your membership status there...


    Thanks Schneidas
    I've found it already and registered my KPA :)
    Cheers,