Others have complained about the latency, and it seems to increase the more you are transposing.
Also, others have said the problem feels worse with cleans rather than high gain.
Kemper cut the latency down via a firmware update a while back, so I think this is the lowest it'll go.
There is a limit in how much you can reduce the latency of this type of effect. Normal use is 1 or 2 semitones.
You get used to the slight delay live and I don't have a problem. I know the quality of sound is there.
I've never used a pitch shifter before but I have used Transpose to -12 to make guitar sound like bass for recording and have not noticed latency issues.
unfortunately latency is unavoidable as the software has to read at least a full waveform (usually two) in order to determine the reference pitch then the algorithm needs to convert it to the new pitch. I wonder if you might have better results playing an octave higher and transposing down two octaves as the initial pitch detection would be twice as fast. However, I don’t know if the downshifting algorithm would be slower and negate any sampling benefits.
One thing you might want to try is blending in some of the original signal to give you the transient attack. I have seen some posts where people claim other transpose pedals are faster than the Kemper but I don’t believe the results they are experiencing are physically possible due to the need to sample a reference pitch. I suspect that these other pedals are blending in some dry signal or some white noise to create a fast transient and convince our brains that they are hearing the transposed signal quicker than it is actually happening.
Don't know how the Kemper works but you don't need to detect the pitch in order to shift it down a fixed interval. Basically the transformation is the same no matter the input frequency.
Don't know how the Kemper works but you don't need to detect the pitch in order to shift it down a fixed interval. Basically the transformation is the same no matter the input frequency.
This has been discussed a few times on this forum and I don't think it is as simple as that. If it were that easy, other modellers would be able to do as well as Kemper or better. Even the dedicated Digitech Drop is no better.
Don't know how the Kemper works but you don't need to detect the pitch in order to shift it down a fixed interval. Basically the transformation is the same no matter the input frequency.
In theory that may be be true but in practice there are a great number of other factors which apparently dictate that the pitch is known. I have to admit most of the technical stuff is WAY over my head ? but I think I “sort of” understand the concepts at very basic level.
I don’t won’t to be one of those self proclaimed experts that crop up on forums and spread false information even though they believe it to be right. So I did a bit of reading this morning and found this research paper which highlights the issues including pros and cons of various different pitch shifting methods. Basically, it says that you can pitchshift mall intervals like a semi tone without knowing the input frequency but even here there sound quality can be less than wonderful to formant shifts, phasing, transient smearing etc. As I already said, the detail is way over my head (I just skimmed anything with a mathematical formula involved ?) but the basic cdesign ncepts are really interesting.
Certainly a good read. Thanks for sharing that. It's mostly an academic discussion at this point but in case of the phase vocoder it doesn't analyze the exact pitch of the input signal (which is why it can also work on drums, polyphonic music etc.). However it's true that the process involves latency - just not from detecting pitch :).
I found an alternative to the 7 and 8 string behemoths. The 5 string Schecter Cello Blaster.
They are rare but not too expensive.
