Pre-purchase Kemper questions

  • Of course - thanks. So that means basically that your lowest buffer size will result in lower latency - whereas if you normally have a buffer size of 512 kB at 44.1 kHz, you'll get the same latency at a buffer size of 1024 kB when doing 88.2 kHz. But the effect on clicks and pops would be the same for both these cases, as I understand?


    Yes, this is correct. However, buffer size that concerns the I/O latency of an audio interface is measured in samples not kB.


    Quote

    If this is correct it follows that the "lower latency at higher sample rates" is only really a benefit if you are at the lowest buffer setting "already", correct?


    That is if your system and your interface drivers allow it, yes.

  • Will you be using XLR or TS for reamping? At what sample rate?


    As others have said, it doesn't matter 'cause the hookup's all-analogue.


    FWIMBW, I'll be sticking with 44.1kHz for the foreseeable future; when I upgrade to MOTU's AVB system the convertors will crap all over what I'm running now (MOTU 24I/Os), so even at 44.1 they'll sound pretty mega IMHO.


    My view on high sample rates (96kHz and above):
    When recording live instruments in high-quality rooms using high-end mics and pres, higher sample rates help capture accurately the subtle room reflections and nuances of the playing. In this sort of scenario, the extra CPU cycles and disk space are justified IMHO. The same goes for ultra-high-bandwidth sources such as percussive and plucked-and-bowed stringed instruments. Even flute comes to mind.


    Now, when recording, say, my Kemper using bass or guitar, my outboard synths or drum machine, or my vocals from an acoustically-dead booth (in the works), it's more a case of simply capturing the mono sources with sufficient bandwidth to be able to faithfully recreate them upon playback. No ambience whatsoever will be recorded; it'll be added artificially. These sources fall well within the standard 20Hz -> 20kHz bandwidth captured at 44.1kHz, and in the cases of bass and guitar (Kemper) their top end will be rolled off during mixing anyway.


    The only caveat to the latter (my) approach is that one would have to, in theory at least, ensure that not too many in-line processes are imposed on the individual tracks as each one introduces a degree of distortion (deviation from the original material). These days, most pro plug-ins do far less damage in this respect than those of latter years by way of oversampling and so on.


    There's just one more important concern many may have with lower-res recording IMHO:
    All those lower-quality mono tracks' being summed by the DAW. Again, IMHO, summing engines have improved, and the one I use (MOTU's DP engine), has been awesome for years. This, combined with the fact that the artificially-placed ambience and delay will bring much in the way of high-frequency detail into the picture (akin to, but not as authentic as that which one would capture using the higher-res outfit I described initially), means that, once again, IMHO, someone running a home setup sans fancy rooms should theoretically be able to mix it with the big boys. Not surpass them, but at least get somewhere close to the ballpark. Looking over the fence maybe.


    OK, OK... listening to the game on a portable radio from the carpark. :D

  • All of the above assumes that you're only recording through the Kemper, which is fine at 44.1KHz. But what if you want to add in some acoustic drums, or vocals? Some horns? Or a String Quartet? Etc. I want to be able to record EVERYTHING at 96KHz and mix at the same Sample Rate. It would be a completely different story if mixing Sample Rates wasn't a problem. But it is!


    Maybe one day we'll be able to do that. Until then, I'm recording the Kemper through its analog outputs.

  • Higher sample rates can be beneficial for SOME cases. This is NOT due to the sample rate itself, but the analogue circuitry before the actual conversion (e.g. the extremely steep high-pass filter).


    I believe it's more beneficial in more cases than you think. Even if you're working all In The Box, a higher Sample Rate gives your plugins more sample points to work with, which gives you better sound quality (i.e. Soft Synths). Sure, there are a lot of plugins that offer the option to up sample internally, but why not just do that from the source? Computing power is there and HDD space is cheap. And, like I mentioned earlier, if you record a lot of acoustic instruments on top of your electric guitars, then the higher sample rate comes in handy too. Not to mention its future prove.

  • I believe it's more beneficial in more cases than you think. Even if you're working all In The Box, a higher Sample Rate gives your plugins more sample points to work with, which gives you better sound quality (i.e. Soft Synths). Sure, there are a lot of plugins that offer the option to up sample internally, but why not just do that from the source? Computing power is there and HDD space is cheap. And, like I mentioned earlier, if you record a lot of acoustic instruments on top of your electric guitars, then the higher sample rate comes in handy too. Not to mention its future prove.



    Don't all DAWs upsample internally anyway for its internal processing?


    EDIT: I may be thinking of bit depth here. Actually, I'm pretty sure that's it.

  • All of the above assumes that you're only recording through the Kemper, which is fine at 44.1KHz.


    Exactly Jose; that was my point, although I included any dry less-than-20kHz-bandwidth source in the argument (e.g. dead-booth vocals). I should have added outboard drum machines and synths to the list too, which I'll do now.


    But what if you want to add in some acoustic drums, or vocals? Some horns? Or a String Quartet? Etc.


    I covered that first up. I suggested that any acoustic source that included a room-sound component or that sported ultra-high harmonics would benefit from the use of higher SRs.


    I want to be able to record EVERYTHING at 96KHz and mix at the same Sample Rate. It would be a completely different story if mixing Sample Rates wasn't a problem. But it is! Maybe one day we'll be able to do that. Until then, I'm recording the Kemper through its analog outputs.


    I use Digital Performer, and I'm pretty sure I can use source material of any sample rate I like; I can only guess that all tracks are up or down-sampled internally by the mixing engine. I'd have thought all DAWs could do this, but I can't speak for them.


    I've a feeling I must be misunderstanding you, mate.



    I believe it's more beneficial in more cases than you think. Even if you're working all In The Box, a higher Sample Rate gives your plugins more sample points to work with, which gives you better sound quality (i.e. Soft Synths). Sure, there are a lot of plugins that offer the option to up sample internally, but why not just do that from the source? Computing power is there and HDD space is cheap. And, like I mentioned earlier, if you record a lot of acoustic instruments on top of your electric guitars, then the higher sample rate comes in handy too. Not to mention its future prove.


    Agreed, but as I said, I see no advantage where dry sources such as synths, guitar and bass processors, drum machines and even vocals in dead booths such as the one I'll be building soon are concerned, Michael.


    Although I've mentioned "acoustic sources captured in good rooms with good-quality outboard" as being prime justifications (IMHO) for higher A/D conversion rates, I think it's fair to say that some dry sources can indeed benefit too. I can only guess what they might be, but in a pinch I'd say that instruments rich in odd-order harmonics (mainly percussive ones) and plucked and bowed stringed ones would be first on my list.


    Thankfully in my case no dry sources will be super-high bandwidth ones, with the dead-booth mono vocals' being the only truly acoustic one. In case anyone wondered, the reason I should be able to "get away" with just that one mono, dry source is... ahem... the Line6 acoustic, sitar, dobro, resonator, 12-string-and-what-have-you modelling. I can only hope that the use of HQ 'verb will carry off the illusion in the mix that said instruments are "real enough". They'll only be used intermittently, often as fairy dust anyway.


    To sum up:


    IMHO, for dry, less-than-20kHz-bandwidth sources, 44.1kHz at the A/D-convertor stage is absolutely fine.

  • I use Digital Performer, and I'm pretty sure I can use source material of any sample rate I like; I can only guess that all tracks are up or down-sampled internally by the mixing engine. I'd have thought all DAWs could do this, but I can't speak for them.


    All DAWs do that, but this is NOT done internally by the mixing engine. It's done BEFORE you import the files into the session.


    Quote

    I've a feeling I must be misunderstanding you, mate.


    Me too :P


    Quote

    Agreed, but as I said, I see no advantage where dry sources such as synths, guitar and bass processors, drum machines and even vocals in dead booths such as the one I'll be building soon are concerned, Michael.


    Ever heard of Soft Synths that up sample internally (some call it High Quality mode or Divine mode, as is the case with Uh-He's Diva)? Why do you think that is? It's basically the same principle as with other types of plugins that up sample internally.



    Agree.

  • All DAWs do that, but this is NOT done internally by the mixing engine. It's done BEFORE you import the files into the session.


    Aah... that makes sense; it'd add an unnecessary load to the real-time mix otherwise. Thanks mate.


    Me too :P


    So... what did you mean when you talked about not being able to mix sample rates, Jose?


    Ever heard of Soft Synths that up sample internally (some call it High Quality mode or Divine mode, as is the case with Uh-He's Diva)? Why do you think that is? It's basically the same principle as with other types of plugins that up sample internally.


    Yes I have, and I agree with the principal. Folks need to understand 'though that the sound of said VIs isn't improved one iota; the existing data is merely represented by twice as many samples, as you'd know. It's only done to provide the mixing engine with higher-resolution fodder, as you'd also know.


    This is why I may well run the mixing process (for the summing and FX-processing resolution) at 96k post recording; it all depends upon whether or not my AVB system, which'll have to be run at 44.1 or 48kHz due to the USB-imposed limitation for 64 channels (can't install thunderbolt on my 'puter), will allow my DAW to run at the higher resolution whilst the I/O, which'd obviously only be used for the 2-bus monitoring of the mix at this stage, is set at 44.1kHz. I'll have to wait and see.


    Agree.


    Glad you you, Jose. Take care, mate.

  • So... what did you mean when you talked about not being able to mix sample rates, Jose?


    I meant that, so far, it is impossible to have different sample rate files playing in the same project. How would the mixing engine keep up with several files playing at different speeds in order to keep them in sync? It would be a nightmare.


    Ever had a sample rate mismatch, where a higher sampling rate file is imported, without conversion, into a lower sample rate project? You end up with a file that sounds lower in pitch and plays at a slower speed, and the opposite is true if you invert the scenario. How fast or slow the file plays depends on the ratio between the audio file's sample rate and the project's.


    Yes I have, and I agree with the principal. Folks need to understand 'though that the sound of said VIs isn't improved one iota; the existing data is merely represented by twice as many samples, as you'd know. It's only done to provide the mixing engine with higher-resolution fodder, as you'd also know.


    But it DOES improve the sound quality. Wether you hear it, or care about it enough to invest in the higher processing load and disk space footprint it brings with it is another matter. That said, not all plugins are created equal, and some may not benefit from using higher sample rates like others. YMMV!


    Take care!

  • But it DOES improve the sound quality. Wether you hear it, or care about it enough to invest in the higher processing load and disk space footprint it brings with it is another matter. That said, not all plugins are created equal, and some may not benefit from using higher sample rates like others. YMMV!


    96KHz is the sweet spot for my DAW and interface. Really wish it was an option on the KPA.

    Kemper Powerhead w/remote & Kabinet
    Focusrite 18i8 (2nd Gen) - Windows 10 - Ableton Live - Yamaha HS-8's - DT770 80 ohms

  • All of the above assumes that you're only recording through the Kemper, which is fine at 44.1KHz. But what if you want to add in some acoustic drums, or vocals? Some horns? Or a String Quartet? Etc. I want to be able to record EVERYTHING at 96KHz and mix at the same Sample Rate. It would be a completely different story if mixing Sample Rates wasn't a problem. But it is!


    Maybe one day we'll be able to do that. Until then, I'm recording the Kemper through its analog outputs.


    I don't think we'll see the Kemper at anything else than 44.1KHz since that's what the vast majority of all studios use to record music at.
    It's the long standing studio standard most pros use.
    The music forums are filled with threads about 44 vs higher and blind tests show there's no benefit in going higher than 44.1KHz since producers can't tell what is what when listening. It's similar with analog vs digital recording, users can't tell the difference.
    I'm all for choice so if they add 48 for film and also higher rates it's better for those who want it, but I doubt they will since most won't use it or that it might take resource from other features, but who knows what they will do.

  • I meant that, so far, it is impossible to have different sample rate files playing in the same project. How would the mixing engine keep up with several files playing at different speeds in order to keep them in sync? It would be a nightmare.


    Ever had a sample rate mismatch, where a higher sampling rate file is imported, without conversion, into a lower sample rate project? You end up with a file that sounds lower in pitch and plays at a slower speed, and the opposite is true if you invert the scenario. How fast or slow the file plays depends on the ratio between the audio file's sample rate and the project's.


    Of course.


    As I said, I thought DAWs simply SR converted tracks on the fly if needed, but as you pointed out, it's upon actual import that this takes place if need be.


    But it DOES improve the sound quality. Wether you hear it, or care about it enough to invest in the higher processing load and disk space footprint it brings with it is another matter. That said, not all plugins are created equal, and some may not benefit from using higher sample rates like others. YMMV!


    Take care!


    I stand by my statement. It's the higher-resolution "fodder" the mixing engine's fed (by the VI that's now been up-sampled) that allows calculations to be made with more precision (provided the mixing engine's running at this higher rate) - useful for summing as well as FX processing, as you'd know. "Straight" playback of the VI sans any summing or processing 'though should sound identical both with and without subdivision of sample length.


    The existing data (the VI's samples) cannot be enhanced in any way through the process of oversampling; it's merely a preparation for processing "down the line". IOW, information cannot be added; all the information you're ever going to get is contained within the source samples.


    Another obvious way of putting it:
    Picture a basic 4-pixel picture - a black, a white, a black and another white square (top left, top right, bottom left and bottom right). Divide those four sectors each into however many segments you like and reconstruct the picture. Does it look any different?


    ... And another: Whether you use 8 lego blocks or 80 (imaginary!) much-smaller ones to make the same shape, you still have the same shape.


    Obviously, if my argument were true, it'd be an exercise in inefficiency to add redundant data to a digital representation of anything... and it is... unless it were to serve a purpose such as I suggested - that of servicing a higher-res engine like a mixing one in a DAW.

  • I think we are now going in circles here. Remember that some plugins over sample internally? My argument is, why not simply record at the higher sampling rate and avoid the internal over sampling process?


    Also, VI's don't over sample in order to create a higher quality version of a sample (as in an audio file used for manipulation). They do so in order to use better antialiasing filters. This in itself yields better sound quality in the audible range.

  • Before I bought my Kemper, I was using Amplitube, Guitar Rig and TH2, all PC based modelers. With this setup, latency is a killer, the processed signal can sound like it's behind what you are playing in real time, if the interface is set too slow.


    At 96KHz, I get a latency value of 7.90ms,with a 1ms buffer. 7.90ms is the time it takes my input signal to get back to the output side of my interface, that includes all VST plugin processing. It was after the fact, that I noticed how much better everything also sounded, at 96KHz.


    The lower the latency, the better performance and sound from my DAW and interface. If there was any possible way to get the KPA to run at 96KHz over S/PDIF, I would be thrilled.


    I'm getting great sound out of the KPA analog out, @96KHz, so I'm not really missing anything, it's just a convenience and re-amping feature that would be greatly appreciated.

    Kemper Powerhead w/remote & Kabinet
    Focusrite 18i8 (2nd Gen) - Windows 10 - Ableton Live - Yamaha HS-8's - DT770 80 ohms

  • I think we are now going in circles here. Remember that some plugins over sample internally? My argument is, why not simply record at the higher sampling rate and avoid the internal over sampling process?


    Also, VI's don't over sample in order to create a higher quality version of a sample (as in an audio file used for manipulation). They do so in order to use better antialiasing filters. This in itself yields better sound quality in the audible range.


    Yes, Jose, after I posted yesterday I realised something:


    Whilst the samples themselves will not sound any better, the manipulation rendered by the plug-in itself, such as filtering, amplitude and pitch modulation and so on will be less destructive in this case.


    I was looking at it from the point of view of the actual samples and not taking into account the fact that in most situations they're being manipulated by the VI involved.


    I'm with you on this, mate.