5.5.0.13341 Public Beta discussion

    The introduction of the different sampling rates is a very important feature for professional studios...
    Many projects in the video/movie world are still at 48KHz and some prefers to work at higher sampling rates (96 KHz) for better quality.
    It's how the converters sounds at different rates that makes the difference. Even top converters like Burl, Prism and Mytek, used in most mastering studios, do sound different just by changing the sampling rate.
    So it doesn't matter if the Kemper sounds better or not at higher sampling rates, it's the whole process that is faster and better.
    Having the chance to bypass une stage of D/A conversion from the Kemper regardless the sampling rate of the project is for sure very handy, I personally prefer the sound I get coming straight from the S/PDIF, not to mention the fact you have less latency while playing, 'cause you bypass two stages of conversion (D/A from the Kemper and A/D from your converters).


    Well done guys!!


    Roberto


    P.S. If the Kemper could work with an external clock source would be a dream, but I fear it's not possible....

    Transpose in the Rig-section is the biggest win for me this release. Saves a stomp slot (with four stomps in the pre-section it's often not enough).
    Would be great if the same can be done with a Noise Gate stomp (which works differently from the noise gate in the input section).

    last night I used the Transpose and wow, I couldn't detect any latency. It seemed different than when I tried it a couple of years ago, not sure if latency was updated in this most recent update or before, but this is very playable.

    Quote from Per

    It wont make a difference unless you have a bat's hearing. 22.1k is well above most peoples upper frequency threshold. You really wont hear a difference unless there's an actual bug in the Kemper. But don't take my word for it, try it.
    Moving outside of the Kemper into the whole 96k argument then higher frequency rates actually often result in poorer quality due to clock jitter being introduced, not to mention when it comes to recording there's the issue of larger sound files. The frequency you need to get rid of things like aliasing artifacts from cheap plugins needs to be about 10x higher i.e. 415k, so with 96k you're only taxing your system more (you can test this using something you know aliases badly in your DAW and trying different sample rates). Once you get audio from an external source into your computer it's going to have a harder time too. It doesn't reduce latency because the bus and CPU on your computer isn't magically faster when you're throwing more data at it, in fact it could potentially create additional bottlenecks and you will likely have to increase your sample buffer to compensate, if you don't have to when you run at 96k then you should drop your buffer at 44.1 instead, if it's already at the lowest your driver will allow then it makes some sense to increase the sample rate while keeping the same buffer size as that will be a way to reduce, but otherwise it's not doing anything.

    Perhaps you can explain to me why if I run at 44.1KHz, with 128 Samples, I get a latency value of 13.1ms.


    Whereas, if I run at 96KHz, 128 Samples, I get a latency value of 7.29ms. Is Ableton just making this up to fool us?

    Quote from MetalMauce

    Would be great if the same can be done with a Noise Gate stomp (which works differently from the noise gate in the input section).

    So true. I don't use the noise gate in the input section anymore, because of that stupid "wah sound" when using your volume pot.
    The noise gate stomp is so much more sensitive to your playing. I would suggest to replace the noise gate (not the noise gate stomp) to another place in the signal path, maybe right before the ampblock, with the same parameters as the noise gate stomp.

    Quote from JSB

    Perhaps you can explain to me why if I run at 44.1KHz, with 128 Samples, I get a latency value of 13.1ms.
    Whereas, if I run at 96KHz, 128 Samples, I get a latency value of 7.29ms. Is Ableton just making this up to fool us?

    It’s because it’s samples per second. So if you set your samplerate higher then the number of samples run through the soundcard each seconds will be higher. Thus making the latency shorter. And your cpu work harder.

    And in the end, the love you take is equal to the love you make.

    Quote from JSB

    Perhaps you can explain to me why if I run at 44.1KHz, with 128 Samples, I get a latency value of 13.1ms.


    Whereas, if I run at 96KHz, 128 Samples, I get a latency value of 7.29ms. Is Ableton just making this up to fool us?

    44.1kHz means 44100 samples per second ... 128 samples buffer results in ... 128 / 44100 = 0.0029s = 2.9ms (in one direction) ... net buffer IO latency = 2 x 2.9ms = 5.8ms
    96kHz means 96000 samples per second ... 128 samples buffer results in ... 128 / 96000 = 0.00133s = 1.33ms (in one direction) ... net buffer IO latency = 2 x 1.33ms = 2.67ms


    The net calculated buffer latency above is smaller than the actual (total) system latency because you would need to add AD and DA and driver latencies. But it should show you how sample rate and buffer size depend on each other and why latency gets smaller with increasing sample rate. :)


    The problem is, that it typically won't help much because the load on the CPU increases the smaller the buffer (latency) is set. In other words, you get pretty much the same latency and CPU load results if you stay 44.1kHz but take the buffer size down to e.g. 48 samples.

    And don't forget that the audio interface also have buffers, and they usually (depends on the manufacturer/model) increase with the sample rate often resulting in +/- 0 difference in perceived latency.

    Quote from michaelmellner

    it should be that if transpose is use in Rig setting, it'll affects every rig in the kpa...until now it could only be used as per single rig
    EDIT: reading Ingolf statement I might have misunderstood how it works...sorry

    Confirm I was wrong: the new Transpose feature is a per rig feature. Sorry...

    "...why being satisfied with an amp, as great as it can be, while you can have them all?" michael mellner


    "Rock in Ecclesia" - new album on iTunes or Google music

    Quote from Dean_R

    How big of a difference between the 44.1 and 48k for the guitar tone might be something that most will adjust to very quickly, but many of the Synths and specifically software synths will sound much better at 48k. I'm sure many will argue that if programmed properly synths will sound the same at 48 and 44.1. but practically even reverbs sound better at 48K.


    So this is a great update to be able to use SPDIF at more than 44.1.

    Thanks everyone that you like our feature update!


    Now that we talk digital audio rates again, it‘s time for resolving some digital myths.


    If you have any device or software that sounds better at 48 than at 44.1, then it is very bad made.
    Especially if a reverb would improve its sound at 48 kHz, it would be a hell of bad code. Can you name one?

    Quote from scratch17


    The higher sample rate at the KPA's output will positively affect sound quality. The sound quality from a device's internal sample rates at super high frequencies only passes as far as the device's digital output. If you limit the sample rate to 44.1 kHz, you will not get the same sound quality as a 96 kHz output.

    Not true. There is no means for improving the sound quality for a digital device by outputting a higher sample rate on the digital output.

    Having listened carefully to KPA SPDIF to DAW at 44.1k in the past and now at 48k (all my projects are 48k) I don't hear a difference but since I find the SPDIF out sounds better I am grateful I can finally use it exclusively.
    I just spent the morning using KPA SPDIF to Cubase and had a blast.
    Thanks again.

    Will

    It‘s nice to have different sample rates but for the live use of my Kemper it would have been even better to have an effect like the Mimiq Doubler f.e.!
    It would be great if the Kemper guys could integrate an effect like that in future Updates.