Kemper os 5.1 Latency Test (+ Line 6 G10 test)

  • Yesterday I got a Line 6 G10 wireless system so I spent some time to test it. Apart the sound quality (that is very good to me) I started to "feel" the latency.
    Because ears are affected by suggestion I decided to spend some spare time to measure some latencies in the chain.


    I splitted the signal: one direct in the Motu 828 mkIII and the other thru the G10 to the Motu and I recorded the two tracks in Cubase Pro 9.
    The Motu had all the onboard effects off and I used only the line inputs (I measured on the mic/hi z inputs an extra latency so I didn't use them).
    Because the software only measure the milliseconds, I calculated the fractions watching the distance to the next millisecond.
    I used the waveforms for reference.


    I started with the G10 and I measured about 2.8 ms




    Than I disconnetted the G10 and plugged in the KPA with os 5.1.
    I used the Main outs
    First I measured the KPA without effects and without Amp section. Just conversion and Sens controls. 2.6 ms



    Then KPA with only Amp section. 3.4 ms



    And finally full loaded with Booster Flanger 3 delays (a loop dly, a Crystal, and a legacy) and Reverb
    Incredibly the latency was only 3.7 ms. I couldn't believe it so I tried many times with various souces, I also rebooted all. but always the same result.



    Of course the soundcard has its own latency like all digital mixers so I'll prefer monitoring on stage from the monitor out to avoid summing latencies.
    I hope it can help someone :)
    Daniele

    Edited 3 times, last by Danides ().

  • Thanx for sharing.


    In other words:


    I am typically sitting 3 meters away from my FRFR speakers. The toaster will throw me back by 3 meters. If I use my G10 it beams me back another 3 meters. In total now 9 meters. Still a small stage to move around on... but bigger than I expected.

    Ne travaillez jamais.

  • Not really, Spinner - the Kemper and G10 will "throw you back" 3 feet each, not metres, so it's really only 2 metres. ;)

    Damn! You mean small stages are just big enough for me?! :D


    (Back to school: speed of sound in air: 330 m/s = 1 ms delay every 3 meters = 9 feet. So you are spot on! Shame on me.)

    Ne travaillez jamais.

  • Speed of sound is 343 meters per second at 20°Celsius. It means 1,27 meters for a latency of 3,7 ms.
    Other latencies like wireless should be added. IMO It's preferable to keep monitoring as close as possible or low latency IE

  • Yes, well done Danides!


    One thing 'though: Shouldn't you be measuring a "random" handful of Rigs, seeing as each is gonna have a slightly-different latency unless Constant Latency™ is switched on?

    I loaded 3 delays, rev, and other fx just to test how much latency would be in an extreme situation, it shoulb be slightly lower for a normal rig.
    You are right, I could measure more profiles but it is more time consuming that it might seem, at least at the beginning.
    Anyway I forgot to say that I was in Performance mode, I don't know if it makes any difference

  • The very highest possible latency should be 4.7 ms -- the fixed value used when the "constant latency" feature is active. I suppose this is a theoretical ceiling calculated by the Kemper team so that it should be enough to produce the output signal even when the most demanding effects are active in every slot etc. (possibly taking into account planned future additions). In any case, your experiment shows that even with a fairly heavy rig, latency is still over 20% lower than this maximum.

  • Yes, well done Danides!


    One thing 'though: Shouldn't you be measuring a "random" handful of Rigs, seeing as each is gonna have a slightly-different latency unless Constant Latency™ is switched on?

    Ok, because the setting is already done, now it is quick.
    Here's another rig, a clean profile with standard basic effects (wha, comp, boost, delay, rev) most effects were off but, because they are loaded in the slot, I'm pretty sure they affect the process.


    I measured 3,65 ms

  • By all practical means: yes.
    The cons: Cable quality is affecting sound quality a bit, especially when the cable is on the longer side.

    The "signal" in conductor travels at 50-99% of light speed. So assuming lowest factor, 150km cable would result in 1ms latency. Other thing is degradation of guitar signal in long cables. Not 150km long, off course.