Running Kemper at 96k

  • higher sample rates can be very relevant for you. And audibly superior.

    Yes, but here's the little flaw in my example with a sports playing field. :D

    In sports, most attention will be on what's going on near the goal, right at the end of the playing field.

    Our ears though are way more interested in the "midfield" action.

    So it's typically not as dramatically "bad" and often not even audible in a negative way in a full music mix.

    It's very easy to make these terrible effects audible by using e.g. pure sine sweeps through a distortion (saturation) stage.

    But these are "lab conditions", not real world music.


    Bottomline:

    People with trained ears can hear the difference when they can compare.

    Is it even possible to produce good music in a digital environment without heavy oversampling in each plugin used? Yes, of course. Just like it was possible to create good music back in the day of vinyl when there were (are) physical limitations to e.g. bass response and bass transients.

  • why then has the lowly vinyl record been resurgent with listeners?

    I honestly think vinyl is more about nostalgia and / or being hip and trendy than how it sounds.


    There will always be those who long for the "good old days," when men were real men, women were real women and small, furry creatures from Alpha Centauri were real small, furry creatures from Alpha Centauri. On the other end of the spectrum, each new generation seeks to define its own identity in part by distancing itself from whatever came before. If dad listens to CDs, then vinyl is most assuredly cooler. The same is true with hipsters, the definition of which is pretty much, "do the opposite of what everyone else is doing."


    The only thing vinyl does have going for it, and in a big way, is the album experience. I do miss the real estate for artwork, liner notes and other visual aspects of the 33 1/3 world. With CDs it became so tiny as to be unreadable, and in the modern era of streaming (does anyone even buy music anymore?), it's gone entirely. I guess you could browse the band's website on your phone while listening to your Spotify playlist, but it's just not the same.


    Mostly, though, listening to vinyl makes you cool.

    Kemper remote -> Powered toaster -> Yamaha DXR-10

  • Mostly, though, listening to vinyl makes you cool.

    Too many people I personally know and respect in music enjoy vinyl to believe that.


    Some certainly do it to be trendy. But like with most things, there is validity in how it (re)started and why it continues.

    “Without music, life would be a mistake.” - Friedrich Nietzsche

  • Too many people I personally know and respect in music enjoy vinyl to believe that.


    Some certainly do it to be trendy. But like with most things, there is validity in how it (re)started and why it continues.

    And of course, no disrespect intended. Just my perspective on it.

    Kemper remote -> Powered toaster -> Yamaha DXR-10

  • I do my cruical recording on harddisks not larger than 1 GB.

    Every bit has more space on these less dense harddisks.

    The result is a more airy and relaxed track. You should try it. Difference is obvious.


    PS: Seagate harddisks sound the best IMHO. But I wlll do more tests.

    NOS drives from the 90's have the best tone. ?

  • And of course, no disrespect intended. Just my perspective on it.

    None taken. But with respect, you may concede the idea that a (still growing) industry exists on the basis of a pop-culture trend a bit thin?


    Fidget spinners these aren’t and pop-culture loves new and shiny.

    “Without music, life would be a mistake.” - Friedrich Nietzsche

    Edited once, last by Ruefus ().

  • None taken. But with respect, you may concede the idea that a (still growing) industry exists on the basis of a pop-culture trend a bit thin?


    Fidget spinners these aren’t and pop-culture loves new and shiny.

    I could certainly be wrong (wouldn't be the first time), but beyond trendiness and the visual experience I honestly don't see any value to vinyl.


    What some call "warmth" (including our favorite monkeys) could also be called "muffling of upper frequencies." Limited dynamic range. Rumble, crackles, pops, hiss and other physical noise from a needle digging into plastic. From an audio perspective, it strikes me as a decidedly inferior medium. I grew up with vinyl, and I'm absolutely delighted I don't have to deal with it anymore. A drummer stole all my albums back in the day, so I don't even have a bunch of them to be nostalgic about.


    That said, when it comes to art of any kind, "good" is extremely subjective. The very qualities that I think make vinyl terrible from an audio perspective can also be the exact things that others love. So, in the spirit of the album era I'll quote a popular t-shirt of that period. "If it feels good, do it." :)


    There's one other thing that I find very positive about albums. They encourage people to actually buy music. Anything that helps feed musicians is a good thing.

    Kemper remote -> Powered toaster -> Yamaha DXR-10

  • Well, vinyl sounds great in frequencies that actually matter

    As all of you probably for sure aware, this argument has been going on for many years now ad nauseum . It's just kinda techy and fun to debate about all the bits, samples and things, but as far as a Kemper goes, and guitar sounds, I don't know about you, but I don't want much 10-20K< in my guitar sound. I have my Kemper rolled off about 8K and it sounds great there through my rigs and recordings. For guitar sounds, below that are the frequencies that matter. And for that and music, 16 bit 44K does great. Great enough to sell millions and millions of CD's.

      I'm not talking about recording birds or whales, but guitar. I actually love the challenge of seeing how excellent I can get a 16/44 recording to sound.


    Like I said before, your techniques and skill will be a much larger factor in how good it sounds much like guitar players sound is in their fingers not amp.

  • All of this! The bass player in our group has a wall of vinyl and always wants us to sit around and listen to some some after rehearsals and I always make excuses to bolt. Then one night I stayed and listened (he had all these uppity tube amps driving everything) and having listened to only wav and MP3 for quite a while,I was blown away at the harmonic content and feel that it had. It was warm , juicy and punchy, with the sweetest high end.

    I believe I can recognize the sonic advancements of LP's vs CD's but to 90% of the people that would be buying a product, wouldn't have a clue. The song would matter more.

  • One way to do that is to enlargen the entire venue dramatically (aka 96 or 192kHz) and only take care of the remaining unwanted effects at the very end of the digital processing chain .... or .... make sure that each individual "saturation stage" plays by the rules and stays within the agreed limits. The latter requires time and calculation power which results in added latency.

    This added latency doesn't matter in post production but can be a serious issue in live production where latency must be kept to a minimum. :)

    We obviously agree that capturing recordings does not require higher sample rates, but sometimes processing does.


    Let me add some hard facts:

    - Running higher sample rates does not necessarily result in higher or lower latency, as some believe.

    - Consuming more calculation power does not result in higher latency.

    - Higher sampling rates uses more calculation power (and storage)

    - Consuming too much calculation power (due to higher sample rate or whatever reasons) on DAWs can result in drop outs.

    - You can increase the block size of your DAW to consume less calculation power, and trade it for a higher latency.



    Certain DSP processes require higher sampling rates, especially distortion. What is not necessarily known to the layman is that doubling or quadrupling the sample rate is only a small step, and will often not solve an aliasing problem. It's only a drop in the ocean. So much more is needed for many distortions. Then there is the majority of processes that do not require lifting the sample rates, such as linear, but expensive reverbs.


    In the Profiler we run our distortions for the Amp and pedals at 16x or 32x sample rate. That is 704 khz or even 1.4 Mhz!

    And there is some situation where we use even lower sample rates than 44.1 kHz, for technical reasons.

    Of course we do all this locally, concentrated on the task that requires this specific sample rate.

    If we didn't, then we would kill our processor and memory, but not gain any sound quality.


    So if you play the Profiler with the Amp and distortions, the sample rate will be lifted up and down many times.

    However, this up and down does not add any latency. This technique is known since decades.


    I wonder why there is serious plugins that do not utilize or offer oversampling where it's needed, or add latency when oversampled and make you correctly conclude that oversampling the whole project is the best solution.

  • We obviously agree that capturing recordings does not require higher sample rates, but sometimes processing does.

    Yes, we agree on that.


    But we don't agree on that. ;)

    First of all we need to understand the difference between system roundtrip latency (defined by audio interface driver design and audio interface buffer size) and channel/plugin latency (caused by plugin processing).


    Examples:

    1. We run a Cubase project with a single stereo audio track at 192kHz and buffer size 512

    Now we can use a Fabfilter Pro-Q 3 equalizer in zero latency mode. We don't have to use heavy oversampling and linear phase filtering. Channel/Plugin latency = 0ms


    2. Now we run a Cubase project with a single stereo audio track at 44.1kHz and buffer size 512

    Here we want to make sure we use linear phase mode at its max setting, just for the sake of an extreme example.

    Booom, we now have a channel/plugin latency of almost 1400ms ... still running fine on the above buffer size setting but obviously compensated by Cubase (all channels get delayed by 1400ms). And I still have quite some processing power left before I would need to increase the buffer size.


    3. As I wrote before, in post production it doesn't matter so much but it's something you'll immediately notice just by the delay when you hit the space bar to play the project and have to wait 1.4 seconds until the playback actually starts, haha.


    Result:

    System roundtrip latency is one thing, channel/plugin latency is another. In a 192kHz project you can run multiple stages of saturation, dynamics, EQ etc. e.g. in zero latency mode and don't care too much about the "mess" it creates beyond 22kHz. Only at the very end of the project when you create the audio deliverables (e.g. 44.1kHz), you should properly low pass the material (with a good filter), obviously.

    In live production you really don't want to deal with latency issues. In (PAL) broadcast for example, 40ms already equals 1 frame and I can easily see how lip sync is off. It's a very costly endeavour to compensate for such a delay (or more) by delaying all video sources by the same amount.

  • This was the point of my resumee: "I wonder why there is serious plugins that do not utilize or offer oversampling where it's needed, or add latency when oversampled and make you correctly conclude that oversampling the whole project is the best solution."


    Oversampling with minimum phase filters is absolutely latency free. And they have no real drawback, especially for this purpose. This is how we do it.


    I can see no reason to use linear phase filters for oversampling in plugins and not give the option to use minimum phase filters for true realtime use.

  • This is how we do it.

    Christoph, can you kindly explain me then why your Profiler (when I feed it a looping sine sweep) chirps like a cage full of grown up exotic birds once I engage any slot with at least a hint of saturation? I mean, seriously, the Kemper Profiler's modules aren't ANY better than e.g. any Fabfilter plugin in its simplest zero latency mode. To be honest with you, to me it sounds as if there's absolutely no filtering going on. Might that be to avoid extra latency? hint hint, cough cough ;)


    And just for the record: Running the Profiler in 96kHz mode doesn't change a thing ... which doesn't surprise me at all because the output sample rate doesn't dictate the internal processing.

  • Christoph, can you kindly explain me then why your Profiler (when I feed it a looping sine sweep) chirps like a cage full of grown up exotic birds once I engage any slot with at least a hint of saturation? I mean, seriously, the Kemper Profiler's modules aren't ANY better than e.g. any Fabfilter plugin in its simplest zero latency mode. To be honest with you, to me it sounds as if there's absolutely no filtering going on. Might that be to avoid extra latency? hint hint, cough cough ;)


    And just for the record: Running the Profiler in 96kHz mode doesn't change a thing ... which doesn't surprise me at all because the output sample rate doesn't dictate the internal processing.

    Sure.

    Aliasing can only be suppressed by a certain degree.

    It depends in the frequencies you feed in, the amount of distortion that you apply and the factor of oversampling.

    Aliasing can only be eliminated by an infinite oversampling.

    Be aware that typical guitar distortions produce much more saturation than typical Fabfilter settings.


    The zero latency / linear phase settings should not have any positive or negative impact to the aliasing result.

    It's the same with regular parametric eq's that can be set to either linear phase or minimum phase.

    The frequency response is the same, but the phase shift is different.


    For oversampling filters this all happens in the 20000 Hz region. For minimum phase filters the phase distortion is very small, due to the high frequencies. But the latency for the linear phase setting should also be small, not more than a handful of samples.

  • Aliasing can only be suppressed by a certain degree.

    It depends in the frequencies you feed in, the amount of distortion that you apply and the factor of oversampling.

    Aliasing can only be eliminated by an infinite oversampling.

    Be aware that typical guitar distortions produce much more saturation than typical Fabfilter settings.

    Ok, that's what you say ... and what definitely isn't quite right.

    • Fire up your Profiler.
    • Now send a looping sine sweep through your Profiler (for example a range from 5kHz - 22kHz in 5 seconds). Ideally via S/PDIF to rule out all potential AD/DA effects.
    • Switch off ALL slots, including the stack!
    • You'll hear the proper sine sweep with no issues.
    • But now, in one slot of your choice, add a Studio EQ and boost 6.4kHz by 5dB or more (just example figures)
    • Sit back, grab a glass of good red wine and enjoy the birds chirping.
    • Now this being out of our way, switch off the Studio EQ and load an amp (and cab if you want) with Gain at 2.0. Still pretty low gain, right? But more than enough to introduce you to the next family of chirping birds.
    • Last but not least, just for the fun of it and with half a bottle of expensive red wine in our head ... let's load BM - Magmaton clean from the Kemper Factory Content. Keep the Sine Sweep going and experience an instant night in the Amazon jungle. :)

    All this being said (and I can make a demo video if necessary) ... if I run Cubase in 44.1kHz and use e.g. Fabfilter Pro-Q3 or for more saturation e.g. the Fabfilter Saturn 2, it's not a problem to create chirping birds. But even the simplest option to increase filter quality will dramatically reduce this wildlife ... up to the point where it's pretty much completely gone with highest quality settings.

    In other words, the Profiler is just as good regarding aliasing as the cheapest, simplest plugin out there.

    Feel free to try it with amp block and all other slots disabled and only one slot with Kemper Drive, Kemper Fuzz or any other "distortion" module of your choice. All of them chirp like crazy.


    And hey ... you probably can't do anything about it because everything you could do (in the current Profiler) would take valuable time (increased latency). Why? Because you can't run the entire internal processing at 192kHz (let alone the 704 khz or even 1.4 Mhz you mentioned) and then properly filter at the end of the chain(s).

    Well, I don't accuse you of lying but I certainly know that there's something fundamentally wrong about this. And I'd rather not read rants about "bad" plugin developers as long as the Profiler doesn't do any better in this particular regard. ;)

  • I would like to see this demoed in a video. Interesting.

  • We are now leaving the latency topic by a bit.


    Your aliasing test is valid, and you can make clips of it.

    I did the test with a sine sweep, as you instructed. I only got a tiny bit of aliasing.


    When you play guitar, you have a hard time to produce audible aliasing, as widely known and shown by numerous of clips on the internet.
    How do we solve this discrepancy?

  • (when I feed it a looping sine sweep)

    Why would you do this? It's made to play guitar with, not electronics class experiments. Plug in and play it. It's awesome. I don't get all the energy spent trying to prove how awful a KPA is when I can produce a list of top paid pros making their living with them and killing it onstage with them.

    I can pick apart a 68 plexi or a 58 les paul too, but their flaws never stopped people with talent from making good music with them.

  • Why would you do this?

    Because I can! And on a more serious note ... the topic here is higher sample rates and why and when they CAN make sense. If you can't contribute, why answer?

    Plug in and play it. It's awesome.

    Don't worry about me, I do this a lot and since a much longer time than you ;) But still I am an audio engineer, not a guitar player with limited technical knowledge ... and if I see false claims, I will speak out (and try to explain). Thanks