The Aliasing Problem

Hi Armin,

no, I didn't mike the cab, I don't have mics at home. The "cabbed" version is recorded straight from the Main out of the Kemper (cab on), the "uncabbed" one from the Monitor out (cab off from the output settings). It is no wonder that the cabbed version shows way less aliasing than the other one, as most high frequencies are cut out by the cab sim.

I know that digital alias is unavoidable. On the other hand, it is well known that preamps generate far higher harmonics than the usual 44.1 KhZ converters, and since the Kemper provides a "direct" output (that is, cab sim off) probably a higher rate DA converter would have solved the problem already (well, provided it covers most distorted preamps harmonics...I am not talking about artificial frequencies produced just to prove that aliasing is still there, of course).

What do you think of the "anti-aliasing filter" cab idea? I think it could be really beneficial. Do you know how one could make one?

Quote from Armin

Did you mike the Mesa cabinet for the sound sample?

The first part fo the clip shows a lot of aliasing - the second is fine.

remember each digital distortion creates aliasing - it's only the question how much ...

And I think the KPA works fine.

Display More

Quote from Rescator

Sorry to hear for your fingers problem Laimon .
Why don't you try with smaller strings gauge or tune down to D ?

Also, I find that soaking finger tips in pure (medical) alcohol can help with calluses stiffness .

No worries Rescator. In fact, a posteriori, I found your joke actually pretty funny in the moment I didn't though, because I really made an effort to record that clip with the fingertip pain.
Btw, when that comes - and it comes way more often than I would like, once or even twice a month - it stays for a few days before it fully recovers, and it takes rest (well, both finger rest and general rest) and as much oranges and spinach I can take. I will try the alcohol too though, one never knows.

Quote from Michael_dk

I am thinking this is an interaction between the hum and the frequency of the notes being played. Sounds like hammering.

I don't think that is it, the clip has a lot of hum (I discovered later that plugging the guitar from the alternative input in the back generated that, while plugging it in the front does not. And yeah, when plugged in the front of the unit, with no (relevant amount of) hum the problem was still there, and actually even more evident as not masked by the annoying hum.

Quote from Quitty

I take it back, you are correct - this is some serious aliasing. Slightly evident in the second track as well, as you said.
Weird, i haven't noticed anything.

Have you opened a support ticket?

Thanks Quitty. It's nice to have a confirmation that I am not "hearing things"
I don't really know how to open one, I might try now.

Hey Paul, nope, that's not what we are talking about. Yes, there is a lot of hum in the clip, but we're really talking about aliasing. If you have the chance, send the clip without cab sim through a power amp and guitar cabinet, and you'll hear it loud and clear. Also in the clip with cab can be percieved, but you will need maybe some good earphones and listening (all the more because of the hum partially covering it)

PS: as I wrote in a post just just above, the hum was somehow due to connecting my guitar on the alternate input in the back - connecting it to the front solves the problem. Cables are Mogami, normally dead silent. I had just reworked the whole routing before making those and I was caught off guard by this hum myself.

Quote from paults

This is not digital aliasing - this is harmonic amplification of the 50hz/60hz hum/noise that can be heard before the guitar is being played.

It is very loud - a distorted profile will amplify and add harmonics to this constant sound. Notes played (and especially bent notes) will cause the harmonically rich constant sound to be consonant and dissonant with the notes that are played.

Using the Kemper's gate, or adding a gate before the Kemper will not fix the problem. When the guitar is being played, the Kemper will still amplify the noise. There is a bad connection or loose shield on a cable (seems most likely a guitar or pedalboard cable). It reminds me of the sound that an amplifier makes when no guitar is connected to the other end of the cable.

Quote from Quitty

The rear input is noisier. It's acknowledged by Kemper, don't use it unless you have to.

By the way, just to be clear - this is aliasing, but really not necessarily digital. Aliasing should exist on your tube preamp as well if you compress the signal hard enough.
Normally, the cab cuts out most of that, but i've never compared the Kemper's no-cab out with a proper preamp - i would imagine the Profiler is capable of producing much more range.
Poweramp and cab is *not* the way to test a studio unit, generally speaking.

Uhm? To me aliasing only make sense in a digital context, because it's an erroneous sample interpolation of frequencies higher than half your sampling rate; hence, no sampling, no aliasing.
How can it occur in an analog setting?

Quote from Trazan

I hear that aliasing in your clip. I can not reproduce it on my Kemper though. Tried it on 3 of my high gain sounds, cab sim off, no aliasing.

Have you tried with Keith Merrow's KM-ANDROMEDA profile? I used that one for the clips. (with that one I get aliasing easily already on the 12 fret of the G string, absurd!)
Of course it depends also on the guitar, I might well imagine pickups and buzzing from low action generates a lot of that.

BTW: I already received a reply for my ticket, they asked more info (which I am gonna provide later at home), and said:

To shape the frequency range for the monitor output you can use the dedicated Monitor EQ that you can find in the output menu.
You could also try to profile a static lowpass filter and use the resulting cabinet in your rigs if you want to cut high frequencies before going from the Monitor out to your poweramp.

I have tried the first, but I didn't find the EQ fine-grained enough for this.
On the other hand, I would definitely try the second option, which seems the best one to me. I didn't see options for profiling something as a cabinet, though?

Quote from Quitty

I'm sorry, i was terribly unclear.

It's often difficult to differentiate aliasing errors from some clipping artifacts.
Aliasing is purely digital and lower in frequency than the fundamental. Clipping artifacts vary, but cover the range of aliases as well.
So you are correct, i'm just not sure we're hearing proper aliasing, just something that sounds like aliasing noise.

No problem, I just wondered if there existed some analog phenomenon somehow referred to as analog aliasing.
What you say is correct, it might be something else, sure. All points lead to aliasing for now, though: first, because I heard it several times in the past and it was really alike, second because the higher the note played, the more prominent the effect (in line with the fact that harmonics must surpass the range of proerly sampled frequencies).
Anyway, later today I will provide all infos to Kemper and also try to create cab profile out of a lowpass filter, and keep you posted on the result.
I am just surpised that no one is being able to reproduce this in their own setup?

Quote from paults

At gigs, I go as high as the 24th fret high E. The Kemper sounds just like the amp I profiled, all over the fretboard. I don't claim to have the same kind of ears as the producers who are using the Kemper for major label projects, but they are getting results they like with it, too.

I agree the clips (especially the second one) have non-musical distortion artifacts in them. But, there is enough input noise to be amplified and distorted, and the second clip has the treble boosted too high to be a practical tone for a mix.

A non-musical result can be achieved with distortion and a lot of high end with a tube amp, too. Or, just plug you tube amp into a full range PA speaker or monitor, and crank up the gain.

Do you go to a guitar cab or PA? Cause in the second case the aliasing it's admittedly very hard to hear.
The input noise is really not a factor, and as I wrote before by plugging in from the front of the unit I solved the hum problem but not the aliasing one (which in fact was even clearer as not masked by the hum)
Too much treble: see below

Quote from nightlight

Can you provide a clip without the hum in the background? Seems to me like this is a case of the excessive brutalz resulting in digital pandemonium. Also, what firmware version are you using?

I hope I can provide a new one hum-free later today.
Firmware is 2.4.2, completely up to date.

Regarding treble and excessive brutalz:
a) that's not really a good point, right? This profile are made from amps and sound settings that people actually use, and sound good in the real thing. I am sure that it is in the interest of everybody to be able to capture also those sounds. After all, saying that I shouldn't use that much treble is a bit like saying "hey, maybe you should play country instead of metal, it will be all better"
b) I picked that profile particularly because it was (out of those I tried) the most affected (aliasing can be heard for over the highest octave), to make a point. For myself I would use less extreme ones, but it is audible also there.