don't clip your EQs

EDIT: see my post on Page 3 - after some extreme testing, I do not believe the EQ's were clipping. I think I was creeping into Output clipping without realizing it. There may be some additional weirdness going on, but I was not able to replicate my initial issue while keeping my output levels clearly in the "safe" zone.

I have a couple patches rigs where I am using a few Studio EQ's (one patch rig has 4). I know how everyone says "try to make cuts only". Yeah, yeah, I know. But if EQ's weren't meant to boost ever, there would be no option to boost. Anyway, I found my tone was getting a bit harsher than it should be, particularly on the high strings in the middle of the neck. At first I thought it was the profile and started adjusting EQ to compensate - reducing the highs around 3 kHZ to get less crackly breakup. This helped a little but I was still left unsatisfied - something else must be going on.

I remembered another patch got nasty because I had a chain of Distortion and Compressor where they were both boosting volume and it sounded terrible. I forget which one was the offender, but the problem was that the first effect had a big volume boost going into the second, and the second did NOT like it. Once I turned the volume parameter down, the tone cleaned up and sounded smooth and natural again.

I tried the same approach with the EQ's - reduce the volume parameter of any EQ that feeds another EQ. Voila. Tone completely cleaned up. EDIT: this was probably due to lowering the output levels, not due to the EQ's. I did tweak to compensate for that initially, but I probably didn't tweak to match the initial output level.

Another thing that helps is if you are ever daisy chaining EQ's, make sure the first one has your cuts. If you have to boost there, only do your mildest boost.

A good setup is Graphic EQ (easy to maintain the sound you want with cuts only) into Studio EQ (to use the parametrics to boost/cut those specific areas that need work still).

"Why so many EQ's?" Because I like to party. No, actually I find EQ'ing is the crux of a good distortion tone. Normally, people use OD stomps or boosts to pre-eq their tone. This works great, but these units/models often do more than simple EQ. In some cases that's good, but sometimes it changes the tone too much. Also, their EQ'ing abilities are usually quite limited in comparison to the dedicated EQ blocks. On the other end of the spectrum (post-eq), bass, mids, treb, presence are great, but often not enough for me. I like bass but not boominess, treble and presence but not fizz. You want to lop off the ultra-high and low end, you're going to need one of the dedicated EQ's. And sometimes you want to get "between" those BMTP controls, or make boosts/cuts that are narrower or wider.

Don't fear the EQ. Just know the pitfalls.

I completely agree that the Kemper doesn't need nearly as much EQ tweaking compared to the Pod. And I actually relaxed my EQ'ing on the Pod quite a bit once I found some good cabs to mix in parallel.

But I'm curious about how one EQ will change how another works. I mean I understand that if I cut in EQ #1 then boost in EQ #2 I shouldn't expect neutral output - I would have reduced the SNR of the cut frequencies at # 1 and would be reducing the SNR of all other frequenies in #2 and thus getting a worse tone overall. But aural artifacts or differing functionality... would you mind explaining?

Will, thanks for the video. Gonna take a little time and reading to fully wrap my head around this. Seems like the example around 3:00 is what I'm most curious about - why is the filter causing ringing, regardless of which mode it is using?

I get confused with DSP when pondering a synthesized signal vs. a recorded signal. Shifting phase in a synth signal is simple, you just change that parameter in the algorithm (usually represented by Theta if I remember correctly). But how can you do that on a recorded signal? Do you Fourier Transform it into a collection of infinite sinusoids which you can then manipulate the same as synthesized sine waves? Of course in a real signal, such sinusoid waves are not going to be continuous over time. Seems like any adjustment to them can cause artifacts.

I feel like I understand linear phase EQ - rather than attempt to perform the above phase adjustment, you create infinite copies of the signal, all delayed by slightly larger and larger amounts as needed until the last signal provides 180 phase inversion for the lowest frequency desired (20 HZ?). Then you amplify these signals according to how much filtering you want to occur and mix them against the original signal. For instance, if you wanted to filter out frequencies centered around the 1 HZ frequency, you would only mix the original signal with a copy of itself delayed by 1/2 a second. To widen the Q you boost the signals more and more as you get closer and closer to 1/2 a second delay. To narrow the Q, you boost the same way but with phase inverted copies of those delayed signals.

This method does not have the possible artifacts with the phase-shifting mentioned above, but because it has to compensate for these variable delays, it can cause transient-smearing as the video noted.

Which method does the Kemper take? That's a good question, but considering that we know the Kemper can center a parametric EQ at a frequency of 20 HZ, which would require a .0025 s delay. Considering you can use at least 6 EQ's in the Kemper that's a latency of .015 s. The Kemper can run with a fixed latency of 5 ms (4.95 ms maybe?), I believe. So given this, I would expect that either the KPA always uses no latency filtering via the phase shifting method, or can operate either way when in variable latency but must use minimum-latency phase-shifting when in fixed latency mode.

As the video notes, the main danger of minimum latency filtering is that it can cause undesirable effects on correlated signals by creating phasing issues. In the KPA's case, there are no correlated signals - the KPA is only concerned with the mono guitar signal.

So how bad is the phase-shifting's effect on the signal? Sounds like experiment time! Fun, my first Kemper experiment. Here's the details.

I will take a couple samples of sounds that don't neatly fit into a nice continuous signal. So maybe a snare drum hit, human speech, and some disconnected instrument cable noise/hum. I will run this through a blank rig and a rig with 6 Studio EQ's - the maximum I can put on the Kemper. Each pair of EQ's will be designed to be frequency-neutral. So one EQ would cut at X HZ while the other would boost inversely. Each pair will focus on a different center frequency.

We should expect the rig with the 6 EQ's to possibly have more noise than the blank rig, although I'm not sure the degree to which this will occur. However, paying attention to the transients in those examples we should note the degree of artifacts introduced. I probably won't be able to test this until this weekend, but I will report back.

From my experience this is what i believed was happening and it shouldnt be hard to replicate - no doubt it would push the amp harder, but shoudnt that be compensated by gain? I am in no way dogging the kpa's eq's - including the tone stack they work very well. But i bet you can disable anythin but a pair of eqs and get distortion

tyler, this is absolutely true and something I stressed heavily in my Pod HD guide. FWIW, on the particular patch I'm talking about I used 2 Studio EQ's before the amp, and 2 more in the X and Mod slots. I found last night I could eliminate the 2nd post-eq by making smarter choices with the BMTP controls and in the other post Studio EQ. But the pre-EQ I need them both. Toggling them on/off is a world of difference and dials in the perfect Mark IV tone.

I also find that the "definition" control is basically like a pre-eq. Many rigs have this value set quite high by default, but I generally don't like the sound I get from that (of course there are exceptions). I prefer to turn definition down until I find the sweet spot where the tone is right between vintage and modern. Then I like to use EQ to change my distortion tone - I have more control over it. This is another reason I like EQ's over boost/OD effects. Whereas an OD will already have a specific sound and only give you a single "tone" knob to tweak, EQ offers much larger possibilities. For example, you can trim bass to tighten up the sound, but can find the exact cutoff frequency so the tone is still fat. You can also dip a little midrange to get more grind out of the tone, which is usually never available in a boost/OD.

Yeah, I actually tested this thoroughly yesterday with clips and patches which I'll post in a bit. I read on the wiKPA doc that Christoph said the internal gain staging isn't like other devices where you have to worry about clipping between blocks - just make sure the input and output aren't clipping, this way there's no need for indicators of where in the signal chain you were clipping. so I put that to the test.

To be clear up front, in "normal" usage, I couldn't get the EQ's to clip, so long as input/output were not clipping. So it is exactly as Christoph describes. Smart design - makes me love the unit even more. So what was happening in my case?

I think what was throwing me off is that looking at my interface's mixer (and I'm connected via SPDIF with no boost on the SPDIF send), I was not hitting red clipping levels, but getting close, and there was clipping on the KPA. If I backed off the volume knob, the clipping disappeared. So I think this is where I previously got in trouble - I think there were a bunch of things on that patch that was boosting the volume up too high, including the EQ's and the cabinet volume - for some reason these were not defaulting at 0 db. I was also turning up the mids on the tonestack.

I had to bring down clean sense pretty low to dial out any input clipping in entirety - the worst is when picking hard directly over the bridge humbucker. I never actually play like that, so I find I can get it around -8 without any issues, but I had previously had it on -4. I don't think it completely relates to the issue I was having, but as part of a thorough clipping test, I did not want false positives.

Also, I think one of my monitors is goofed up. I knew of one odd behavior it would exhibit, where the sound would cut out, but I've never heard it distort. Something VERY weird was happening on Friday, and I can't replicate it on headphones, so I don't think it's the KPA. I boosted the EQ's like I originally had them when I had issues. I would switch off the amp block - nice clean tone. I switch the amp block on and I get a weird distortion on top of the tone, like a ring modulator, or out of tune radio kind of sound. If I switched the amp block off, the distortion would persist on top of the clean tone for about 1 - 2 seconds then disappear. I had to replace the giant capacitors tied to the power supply on one of the monitors before. I don't remember which one - I have to take it apart to see what's wrong.

...

Now, I did do an "extreme settings" test and got some VERY weird behavior. It's post the clips and demo patches in a bit.

i'm on fw 2.2.0 beta btw