Okay, let's do this!
This is what makes your ears so suspect.
It is simply objective truth that the Kemper aliases, and it does so at a higher level than Helix, Quad Cortex, and the Axe FX 3/FM9/FM3. This is incontrovertible, having been measured and observed by many many people. The video that I and
@Jarick posted shows this extremely clearly. Here it is again:
You simply have no grounding in reality if you wish to argue this point.
So.... what did I actually say? Let's go over it again:
If I was to be super charitable, I would say that preferring the sound of the Kemper over Helix and Axe FX III, is akin to preferring the sound of an 80's rack delay over a newer unit from Strymon; because almost certainly the 80's digital rack delay won't have implemented high quality resampling algorithms, whereas the Strymon would have. The older rack unit would alias to high heaven (just as the Kemper does), but some people don't care, and it can certainly add to the character of a unit. I just personally don't think it adds very much to a guitar rig.
Let me walk you through the breadcrumb trail of my thought process:
1. Kemper aliases.
2. Kemper aliases more than Fractal.
3. 80's rack delays didn't have high quality resampling algorithms, thus they too alias.
4. I make the assumption that this is not true for a modern Strymon delay like the Dig pedal.
5. But I acknowledge that some people prefer the sound of the 80's rack delays like the 2290, over the modern stuff.
6. Which means that the aliasing doesn't bother them, and indeed it is entirely possible that this is part of the reason they subjectively prefer the unit.
7. Since this is true for rack delays, it is possible that it is true for amp modellers.
8. Therefore, when someone says they prefer the sound of a Kemper over the Axe FX III, my comparison to rack delays is compeltely appropriate.
I really don't see why this logical train of thought makes my ears suspect. The only thing I can come up with is that you wish to attack my character rather than deal with the argument. This is an ad hominem attack.
Are we to believe your ears over all those touring musicians using Kemper? Or for that matter, are we to believe your ears over all those who have had Kemper AND NAM AND QC AND AxeIIFx?
This is an appeal to authority fallacy. You're relying on the opinion of someone you consider an authority, instead of dealing with the actual argument put forth. It isn't my ears that tell me the Kemper aliases. It is measurements. Measurements that are very easy to do and you would imagine any electrical engineer worth his salt would be able to do them.
PS: Having a background in electrical engineering doesn't necessarily mean you have any understanding of digital signal processing, or digital audio full stop.
With respect to Aliasing, I happen to be a senior EE. Your explanation leaves quite a bit to be desired as well.
I kept my explanation general so it was easy for
@BenIfin to digest it, and he appreciated it. Not everything is about you.
Without going into the mathematical theory behind how it is possible, I will state simply that if you sample at a rate twice that of the highest frequency you wish to PERFECTLY reproduce, you have no issues.
This is way too simplistic to be a useful descriptor. There are many points throughout a DSP system where aliasing can occur, and the frequencies often go far above nyquist * 2. Therefore your approach is far too general a statement, and in quite a lot of cases, simply will not work well enough to get you the result you desire.
While different processing techniques can cause unwanted artifacts, those are not called Aliasing.
Which processing techniques are you referring to? I've talked about non-linearities causing aliasing, but I've also talked about resampling algorithms causing aliasing too. And yes, it is called aliasing in such a case.
There is an extremely well known paper called "The Quest For The Perfect Resampler" written in 2003. It is a good read and anyone familiar with DSP knows about it:
Let me give you a quote:
This is a technical paper dealing with digital sound synthesis. Among all known sound
synthesis techniques, there is at least one that will never be outdated: sample playback. We
present in this document a computationally cheap method to play back samples at variable
rate without perceptible aliasing. The main application of this technique is virtual music
synthesisers running on personal computers.
So, if your assertion is that a given resampling algorithm can cause side effects, but that isn't called aliasing... then I would simply point you to the academic literature.
FWIW, most internal processing these days is done with methods that provide insane dynamic range to prevent the algorithms from overcoming the mathematically limited storage of each sample. Again, I am not going to try to explain the math or the algorithms but will say that it is a fairly safe bet that the kinds of gear we are talking about here would employ these methods and make nearly all your assumptions incorrect.
This chunk is pure technobabble.
Why do you think dynamic range has anything to do with aliasing? Aliasing is a product of the frequency range. Not the dynamic range. You don't need to explain the math. All you need to say is that a plugin internally processes floats and doubles, giving you an approximate level of precision. Floats are 32bit, and give you approximately 7 decimal places. Doubles are 64bit and give you approximately 15 decimal places.
These two types of storage containers for a sample (IE: a number!!!) are not limited in any practical way. Maybe you were thinking of fixed point precision, but plugins haven't been written (at least primarily) with fixed point precision for over 10 years at this point; although fixed point precision still has its place.
In my previous post, I referred to something I hear as "Aliasing" as well, so I am also guilty. Having artifacts in the reproduced tone is not likely due to aliasing in any high end device today. It is much more likely an issue with the rest of the capture chain (microphone, placement, speaker, etc, etc).
The Kemper video we are referencing, it
IS aliasing. No question about it. But the more important point to make is this
the analog realm has no impact on how much aliasing a device creates. Aliasing is a result of insufficient sampling rate relative to the signal bandwidth. It is completely unrelated to bit depth or precision, and is not directly related to dynamic range either.
![[Nathan]](/data/avatars/m/0/613.jpg?1726885163){kind=avatar}
