96khz 24bit profiling

    Monotone's right in every way, I can understand why he's walked away from all the "insight errors" presented in the developing communication within this thread.
    Don't shoot the messenger, the Kemper's a great box, (in a world of trade off's), but it would be a much better box at 96k and marginally even better again at 192k.
    (obviously providing all the analog, digital and software engineering changes required where addressed).
    Main problem is present availability of content and the present projected price tag.
    Kemper are aware of all this and would be researching the next generation - of Professional digitising rate Profiler.
    Just hope I'm around.
    Snowy.


    Thanks! I've been meaning to watch that for some time. Haven't got the time to right now, but it's bookmarked.


    Ethan is a funny character. He's right about a lot of things, I think, but he's also maybe a bit simplistic in his approach sometimes. All in all, I think he's a good counterweight to what seems to be the majority of the internet :) at least, he makes people who a receptive to his ideas think a little bit more for themselves.

    Please, back up your claims it would "be a better box" at 96 kHz. I won't go into 192 kHz, because that just isn't true (read up on AD/DA converters physical limits and why 192 kHz actually degrades audio). I can't think of any reason other than convenience (for digital input without having to upscale at higher sample rates).


    In any case, the analogue outs are equally as convenient and I'd love to see a blind test comparing them with the spdif, with decent cables and inputs.



    What is it you like about the frequencies you are physically unable to hear, if you don't mind my asking?

    Snowy, you came (fell) out of the blue... sky... on us, mate. You say he's right in every way, and imply that we have our heads (ears?) in the sand?


    Well, you're not in it, but snow does fall onto sand. I recommend you buy a Kemper, if you don't have one already, when a 384kHz model is released, just to be on the safe side.


    Quote from sambrox

    Please, back up your claims it would "be a better box" at 96 kHz. I won't go into 192 kHz, because that just isn't true (read up on AD/DA converters physical limits and why 192 kHz actually degrades audio). I can't think of any reason other than convenience (for digital input without having to upscale at higher sample rates).


    In any case, the analogue outs are equally as convenient and I'd love to see a blind teat comparing them with the spdif, with decent cables and inputs.


    I'm perfectly happy with the analog output, Sam. That's all I'll be using for recording, just as most would do with conventionally-mic'd guitar amps.


    Most mic pres don't have digital interfacing at any rate (pun intended).

    Quote from snowy

    Don't shoot the messenger, the Kemper's a great box, (in a world of trade off's), but it would be a much better box at 96k and marginally even better again at 192k.


    Monotone, is that you? :D


    The Kemper can still be improved, but its sound quality has absolutely nothing to do with sample rate. There's a more=better fallacy going on here.

    Quote from Michael_dk


    I think this video explains everything :)
    It also seems I was wrong - you only need two points per WHOLE wave to reproduce a perfect sine wave.


    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.


    What a great video. Not only very informative (I will make so many people watch this) but also very well done from a didactic point of view, combining clarity and density of information. Not one obsolete word. A pleasure to watch. Thank you @Michael_dk.


    Quote from sambrox

    I'd love to see a blind teat.


    Who wouldn't!?

    Nyquist is only relevant to repetitive waveforms, that's not music.
    Lot of stuff on the Web is just plain wrong or misused, the DAC reference on aliasing was written some 20 years ago to justify a DAC companies position on sticking with a 96Kb DAC when everybody else in the World was moving to 192Kb.
    I listen to Redbook, (one of the reasons music sales plummeted around 20 years ago) and I try too avoid it.
    (The misrepresentation of Nyquist was used to justify Redbook digitising rates at that time).
    A lot of 96k x 24 which is much better providing it was recorded properly.
    Some 129K x 24, this can be stunning - if the recording techniques employed justified its use in that particular recording situation.
    (it's the only digital music that can make me cry, like some old vinyl masters I still have)
    A lot of my music is based around electric guitars and the difference between the above standards is still apparent, (providing everything in your listening chain is reproducing the sound at the required quality).


    Once you hear this stuff on the right equipment you can't go back.


    The Kemper runs at Redbook CD quality, but with 24 bits which gives the music more dynamic range, (dynamic range contributes to musical emotion, you can have dynamic range and compression going on at the same time in a recording of combined instruments).
    The cost of computational power alone required to build a Kemper running "properly" at 192 x 24 and performing all it's incredible magic is probably the show stopper at this time.


    Many electric guitarists I know that have been playing at high volumes for a long time now have damaged hearing, it comes with the territory.
    Just because some of us can't hear the difference doesn't mean it's not there.
    We need more "Monotone's" to fuel future Kemper improvements.
    Cheers Snowy.

    192 kHz at 24-bit depth isn't actually physically possible with today's technology (even the highest grade and most expensive converters max out at 20-21 bit and lowlier variants will actually degrade the audio). This is a fact.


    As for 92 kHz, if you're bouncing down to 44.1 anyway, you're probably going to again degrade the audio in a worse way than if you recorded at 44.1 in the first place. This, unfortunately, is also true. Dithering is always a trade off. I won't dismiss the possibility that there might be a pleasing difference when listening to a recording recorded at a sample rate greater than 44.1 through equipment that is able to reproduce frequencies above 20 kHz (there's a psycho-acoustic effect apparently, even though you aren't actually able to hear the extra information), but it would require an excellent, clean chain throughout (not least your monitors; if your tweeters can't go above 20 kHz, the extra info will tend to be smeared downwards in the spectrum, colouring the regions you can hear with unmusical harmonic distortion) and the jury is still out on whether it is just confirmation bias at work or not.

    Quote from sambrox


    Neil Young and his digital music player partners ought to watch it, too [facepalm]




    24/192 Music Downloads ...and why they make no sense
    http://people.xiph.org/~xiphmont/demo/neil-young.html


    If something is beyond human physical perception and we can't hear it in blindtests, then it simply doesn't matter, except for audiophiles and product marketing. xiph.org is great for debunking audio myths in a very pedagogical way.
    As sambrox mentioned earlier, human confirmation bias is very real and often takes over our decision making, we hear or see what we believe. It can be bypassed by blindtesting.
    link confirmation bias https://en.wikipedia.org/wiki/Confirmation_bias


    Some "audiofile" friends of mine were so excited about how "superior sounding" 24/192 albums are compared to ordinary cd's.
    It's beyond human hearing to tell the difference (they market a myth) so I setup an abx blindtest for them, comparing 24/192 to regular 16/44.1 cd, and mp3 320). they failed all the test and never could tell, but when they knew what they were listening to they said it was still superior. They refuse to accept being victims of confirmation bias and it makes them feel comfortable in their strong personal convictions.
    It's also funny reading on studio forums how some argue over which dithering sounds best, when it's beyond human hearing to even tell if dithering is on or off in a song.


    another good link
    Audio Myths & DAW Wars
    http://www.image-line.com/support/FLHelp/html/app_audio.htm


    As posted earlier, this one is just excellent.
    D/A and A/D | Digital Show and Tell (Monty Montgomery @ xiph.org)

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.


    Audio Myths Workshop

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

    Where to start.


    I won't indulge in a wordy lecture addressing all half understood theory's/opinions verbalised as facts.


    I have never had a client book the studio to record sine waves which are easy to digitise and have been for many many years.


    Snowy is right. A guitar signal is not a sine wave. There is alot more going on than just fundamental notes. This fact makes a huge impact on how well converters work if lazy coding has been used. I'd like to see the chap in the video run a complex sound wave from a instrument with multiple harmonics beyond the fundamental.


    To say odd-even order harmonics in guitar don't effect timbre is an excuse for lazyness. At any point in musical production. This is where compromise is made.


    Correct 44.1khz on paper is sufficient bandwidth for human hearing according to Nyquist (a term internet gurus seem to throw around without understanding it and its shortcomings in the real world) however, if the anti aliasing filters are cutting out frequencies because of lazy coding or hardware limitations etc etc (guitar note is complex and made up of added harmonics) you are NOT going to get full bandwidth at even at 44.1k. Sorry sambronx there is more to the story than just Nyquist theory and 'covering the bandwidth of human hearing range' at 44.1k'.


    Just because a digital converter running at 44.1khz can accurately replicate a sine wave at 20k does NOT mean it can replicate a complex sound wave digitally (say from a guitar) without distortion or aliasing issues. The solution? Cut off the offending frequencies.


    Raising the sample rate will assist the need for aliasing filters and help lazy coding amongst other things reducing latency. Nothing to do with increasing the hearing bandwidth which so many here seem obsessed with the only reason one would increase frequency range.


    I will stop here.

    On a side note, it has been proved that humans can perceive sounds below 20 and above 20k Hz. We don't use ears of course, but complex structures in our head.
    I've posted a link to such researches last year responding to a post by Eng. Kemper.


    For this reason, I believe we can actually benefit from digital recordings made above the Redbook standard.


    :)

    But you're still talking about frequencies beyond the range of actual hearing, when you say "cut off the offending frequencies". 44.1 can reproduce frequencies from 20 Hz to just shy of 22 kHz just fine, thank you (22,500 Hz in theory, although the bandwidth filter to combat aliasing of harmonics outside of the audible spectrum means it's slightly less). Ever heard a full mix master on CD? What could be more complex than a whole band playing? Or how about pink noise?


    Again, I won't outright dismiss that ultrasonic frequencies won't have some psycho-acoustic value, but unless you are listening through a professional studio playback chain to a master above 44.1 kHz, it doesn't matter one bit. Who are you recording and producing music for? Consumers won't have access to the facilities to hear what you deem as the extra important data, and even if they did have the facility to play back a master greater than 44.1, the probability of their system making it sound worse than a recording made at 44.1 is very high indeed. There will always be a market for audiophiles and people daft enough to spend thousands on speaker cable, but Joe Public listens through his bluetooth speaker or iBuds via a $10 DAC...

    Edited 4 times, last by sambrox ().

    Like 1

    'there will always be a market for audiophiles and people daft enough to spend thousands on speaker cable, but Joe Public listens through his bluetooth speaker or iBuds via a $10 DAC...'


    Exactly why this whole thread is pointless imo. Sorry but on anything anyone I know listens to music on, there would be no discernible difference

    Quote from viabcroce

    On a side note, it has been proved that humans can perceive sounds below 20 and above 20k Hz. We don't use ears of course, but complex structures in our head.
    I've posted a link to such researches last year responding to a post by Eng. Kemper.


    For this reason, I believe we can actually benefit from digital recordings made above the Redbook standard.


    :)

    Yes, but until the lowest common denominator gets raised a few levels, it'll all be for nothing. Once iBuds have died a death and streaming services/Internet bandwidth catch(es) up, then we can talk ;)

    Quote from viabcroce

    On a side note, it has been proved that humans can perceive sounds below 20 and above 20k Hz. We don't use ears of course, but complex structures in our head.
    I've posted a link to such researches last year responding to a post by Eng. Kemper.


    For this reason, I believe we can actually benefit from digital recordings made above the Redbook standard.


    :)


    Can you post a link to this?

    Higher sample rates are useful when dealing with audio processing to eliminate aliasing artifacts, which is why the Kemper does in fact operate at a much higher sample rate than its convertors, but for simple playback 44.1k is more than enough.


    A higher bitrate will not result In a less "hi-fi" sound (quite the contrary). But the use of a good room IR verb in your daw after the Kemper will, the space parameter can help a little.


    Decent monitoring and room treatment will help a great deal too, as well as doing some actual profiling in order to do A/B tests rather than trying to A/B with someone else's profile and an amp in the room (you can only compare the recorded result, or using the same amp cab as output for both the amp and the Kemper if you want accurate test).