Posts by tgs

Been running a bunch of more tests now, been trying to change various things in the method.

* Tried with a different amp (Mesa Trem-o-verb)
* Tried with less distortion
* Tried with almost clean sound
* Recorded a new DI track with chugging on various notes so it's not just one bump in the frequency range, so to speak
* Recorded a parallel signal from the mic pre to rule out that the Kemper did something weird when playing back the return signal
* Checked frequency response when playing a track with no chugging at all

I keep coming back to the same results. As for the EQ curve, I modified it a bit, still a bit different but I only use it to get close enough in the most problematic range. I changed the HPF from 12dB/octave to 24dB/octave to get it more accurate. Apart from that, the main problem area is always around 100-120Hz, there's a peak there that the Kemper simply doesn't capture when I run the tests. Regardless of amp, cab, mic, etc. That, and the "addition of information" below 90Hz.

Mesa amp, Mesa cab, SM57, recorded with new DI signal as described above (by the way, the Ozone frequency plots do NOT correspond to the scale on either side of the diagram; unfortunately I can't zoom in on the curve either):

EDIT: Sorry, the scale on the left DOES correspond to the frequency response and I CAN zoom in if needed.

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/NewDI_kemp-amp.JPG] 

EDIT 2: The problem area zoomed in:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/NewDI-kemp-amp-Zoomed.JPG]

The modified EQ:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/NewEQ.JPG]

The frequency response reading of the amp vs the Kemper WITH eq applied:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/NewDI_kempEQ-amp.JPG]

Also, if there's any doubt as to why I'm doing this:
I'm trying to narrow down a problem and troubleshoot. So far, my findings have been very conclusive, the same problem over and over. I'm trying to approach this in an as scientific manner as possible and I'm trying to rule out various things. Therefore, it would be extremely helpful if someone else could try to replicate this, using the same methodology. If they come to the same conclusion then the problem is uniform, and it could be a flaw that could possibly be addressed by Kemper in a future update (possibly by giving us the option to apply a similar corrective EQ curve in the profiling process; the fact that I'm getting nearly identical frequency discrepancies would support such a possibility). If others can NOT replicate this, then it indicates that something is wrong on my end and I'll have to keep trying to find out what it is. Saying things like "you don't want that much low-end anyway" doesn't change the results.

If my methodology isn't clear enough, let me know and I'll explain exactly how everything is connected and how I proceed!

Quote from DonPetersen

could you please check if the 'Direct Mix' parameter is set to something other than '0' in your profile?

Checked. It's at 0.

Quote from ckemper

3 dB is a lot, actually a lot more than your JMC profile would need.
By carefully refining you can achive a deviation of less than a half dB with most amps.
A good final refining is made by playing in those regions and styles where the deviations are revealed.

Have you tried to profile your Orange again with less amp volume?

It is a lot perhaps, but to say it's more than the JCM needs is not correct. I reached these results after measuring, which is very important to keep in mind. After applying the EQ as pictured, the Kemper sounds closer, and it MEASURES closer. Please also don't neglect that I had to boost AND apply HPF. If you apply the same curve as I did, you should get the same result. I can make a screenshot of that measurement as well if anyone's interested.

Oh believe me, I've tried to refine by playing what the "problem" is. Just for the heck of it, I've now also tried refining by playing palm mutes for a good 20 minutes. The result in the low-end is...exactly the same.

Tried with less amp volume, yes, as described in an earlier post. That's also what I did now with the palm mute session of death.

Quote from paults

But, even if your profile is 1, 2, or even 3 db down around 100hz to 200hz:

Every recorded guitar track ends up with less low end than the amp has, and in a live mix, a guitar with as much going on in the 100hz to 200hz range as in higher frequencies would also have the low end rolled off in FOH, because that frequency range needs to be available for the Bass and Drums.

That's completely besides the point. Even if it would be customary to roll off the low-end (which is a far too blanket statement) I do not want the Kemper to make any mixing decisions for me! Actually in that respect, with my test results the low-end rolls off nicely from around 90Hz with the amp while the Kemper adds something below that. I'd be far more concerned with that than anything else.
For the record, I'm not a "guitar player" so to speak. I'm a studio and FOH engineer. To to address this:
"your Engineer is going to roll off those frequencies in order to allow the other instruments space in the mix"
I AM the engineer.

To try to address some more points...
* I've tried starting with a similar profile.
* I've tried re-starting the profiling after initial (and second and third attempt etc). Same result.
* I don't currently don't use the SPDIF. I'm sending the DI through a LittleLabs Redeye 3D. When I run the tests I'm first playing the DI through the profile, then switch to Reference amp. It's exactly the same signal being used both times.
* I'm aware of phasing affecting the sound while listening through monitors and having the amp play in the other room (I deal with that to a certain extent every day in the studio). But that wouldn't account for different measurements. If ever in doubt I usually pop on a pair of closed-back headphones. In those I can hear the difference even better. But measurements are measurements, it's not only up to what I'm hearing or perceiving.

By the way, here is a reading with pink noise playing through the profile vs through the amp:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/pinknoise.jpg]

Quote from crankyrayhanky

I wish I had ozone to check this out on my end...my ears do not perceive any low end bump as OP states, but IDK

There are other options than Ozone for capturing the frequency response, probably some free ones as well. I'd love if some others could make the same kind of test and show the results. We might be on to something interesting!

Some more tests done. Moved cab to different room: same result. Turned down the volume a lot to drive the power tubes less and try to get less room influence: same result, in both rooms. Hmmm...

Here's the kicker though: Tried with the PSA-1.1. The result was MUCH closer, practically identical.

Quote from Jeffsco

Because at that exact spot in the room...there IS more of that frequency due to a buildup of sound waves at that frequency. You need a signficant amount of bass trapping to level out Sub Bass issues. You'll get Peaks and nulls throughout the room. I'd suggest moving the amp and mic to another part of the room and profile it again. See if the frequency response doesn't change. If it does change...you know it's just a room anomaly due to the peaks and/or Null at that frequency and not the Kemper

Ok, to clarify: in the first test it was a Mesa cab in one part of the room. In the second test it was a 1960BX in a different part of the room. Same result.

Quote from Jeffsco

Does the room your Profiling your amp in have issues in regards to these frequency area's? If you hare sampling in a part of the room where there is a higher cancellation of these frequencies...you'll see a skewed response in the bass region but reaching up to 200-300 hz.

It's a quite well-tuned room, but it could still be a factor. Would be great if more people could do the same kind of test so we can compare the results. Still, why would the Kemper profile add so much stuff in the sub region, that really doesn't make sense to me.

EDIT: I'll try to profile a PSA-1.1 later tonight if I have time. If it's still the same problem then it's something fishy going on...

Quote from crankyrayhanky

Interesting eq curves...I assume you have & tried different amps (3+)? I found some amps profile much better than others
With so many variables, I'd like see a 2nd person with ozone try to repeat this procedure

3 amps so far, two cabs with different mics. I'll try some more when I have time. I totally agree that different amps profile better. But it's the low end response issue that seems to be the same regardless.

Quote from Quitty

Have you tried refining without playing any low notes at all?

Yes I have. Unfortunately I didn't have time to be as thorough as I wanted, but I tried a few different things:
* Only open strings
* Only palm mutes
* Only dissonant chords
* Dissonant chords + palm mutes

In all cases, the results were very similar.

Alright, some further testing done. I recorded some DIs this time to make the test more accurate. I tried some different things, increased power sagging before refining, played dissonant chords, open strings, palm mutes... The results were practically the same though. What I found was that one specific EQ curve always made the profile get closer. I'll need to try this on different cabs and amp settings before I can draw any conclusions (and I have a studio to run unfortunately, hehe), but here are my findings so far:

Captured the frequency curves with Ozone; blue line is the amp, yellow line is the Kemper profile:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/kemper_curves.JPG] 
As you can see the Kemper has much more stuff going on under around 90Hz. It's also lacking in the 100-200Hz region. Some discrepancies around 600-1500Hz but it's not as noticeable.

I applied this curve to the Kemper signal afterwards:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/kemper_eq.JPG]

The result comes much closer. Not all the way but much closer. The Orange still breaks up very differently but I guess that's because I'm driving the tubes hard. Difficult to capture, I get that. With the Marshall it's much closer.

Now, you might want to say "so just apply that EQ setting then". Indeed, if I'm in the studio working with a DAW, fine. But if this discrepancy is consistent for all profiling (I'll need to run some more test, as I said) it would be great if it could be applied on the Kemper side already.

EDIT: Ok, couldn't help myself. I ran one more test. Different amp (5150), different cab (1960BX), different mic (MD421). The result:

[Blocked Image: https://dl.dropboxusercontent.com/u/8687414/Kempertest/kemper_curves2.jpg]

Looks familiar? I then applied the exact same EQ setting as above. Still, not exactly the same, but MUCH closer. Hell, as far as I'm concerned - let's say the Kemper had one more feature in the profiling options like "Clean up low-end". In that case the profiling could stay the same as before, people would get the same results if they don't engage that function. How about that?

Quote from ckemper

Refine in the region where you think it's required the most.
In your case do palm mutes.

With palm mutes that actually seems to make it worse. Actually like tylerhb describes, just strumming open strings often get me closer. But I'm still lacking the low end...

I made some tests with EQ yesterday, I'll try it some more today and post my results. I was a bit surprised by what I found actually. What I get is not a HPF on the Kemper signal at all. It actually has MORE subs, BUT it lacks some dBs around 100-120Hz.

Quote from Quitty

However, it's always the same inaccuracy - it skimps a little on the sub-bass.

Exactly! That's my point. It feels like there's a HPF engaged. I wouldn't even say sub bass, it's higher up than that as well. Guess I'll measure it some day if I feel a bit bored.

I understand perfectly that the Kemper can't capture the sound to the very last detail, but when it consistently has the same problem then I'm leaning towards calling it a design flaw. I thought I was doing something wrong at first but it seems like I'm not alone at least. If anything can be done to improve this in the future, that'd be greatly appreciated!

I'll try the power sag thing that you suggested and see if it makes any difference for me. I've fiddled with that setting before and it made very little difference for me then, but I never tried it while refining.

Quote from ckemper

I think the JMC is as close as it can get.
It took me a few seconds to hear the difference even.
This can be corrected by adding one or a half a decibel in the bass region.

But the question is: Could you tell A or B after you left the studio for 10 minutes?

The Orange is worse.
Did you push the power amp hard?

I agree that the JCM800 is very close. I've also used a profile with the same setup for an album earlier this year and it worked really well.

If I could tell which is which after leaving the studio, well my answer would have to be: it depends. If I'm listening on the monitors with lazy ears I think it sounds pretty damn close. If I listen on headphones I can clearly hear the difference, the low-end just isn't there. On open chords - not a huge difference. On palm mutes - big difference.

The Orange is driven pretty hard, not extreme but pretty hard. I suspected that that might be the problem.

What troubles me though is that it captures some things really well, but the low-end is a different story. I tried with two amps and it was the same problem with both. I can try a few more but I suspect I'll have the same problem.

I'm not trying to bash the product or anything, I love my Kemper. I just wonder about if I'm in a situation where I want to fix up a part that was recorded earlier and where I have taken a profile. If I recorded with the Kemper, no problem. If I recorded with an amp, I wouldn't want to replace it with the profiled signal because the low-end would sound very different. Sure I could add a bit of low-end with EQ, although as described it's not really the same. But wouldn't it make sense if the Kemper simply captured that "right" in the first place?

I'm having some trouble with profiling and I was hoping someone could give me some pointers on how to improve.

My main issue is that the low-end doesn't sound right at all. The Kemper sounds high-passed and palm mutes don't translate well. I also noticed that right after profiling, the Kemper profile is noticably quieter than the reference amp signal. After refining, they are closer in level but during refining I can basically hear the mids getting louder but the low-end seems to stay put.

I made a few samples where I'm switching back and forth - Kemper first, then reference - to show what I mean (tuned in drop C so I could play with one hand - I normally play in D). First up is my JCM800 -> Mesa cab -> single SM57. As you will notice, right after profiling the difference is quite big between the Kemper and the reference. After refining it's much closer, but pay attention to the low-end:

https://dl.dropboxusercontent.com/u/8687414/Kemp…_pre-refine.wav
https://dl.dropboxusercontent.com/u/8687414/Kemp…post-refine.wav

Next is my Orange Dual Terror, and here is where I really can't get it to work properly. Right after profiling, the difference in level is very noticable. After refining, again it's much closer but the low-end is just so much weaker. It also breaks up in a very different way, kind of "uniform" vs "non-linear" in lack of a better description.

https://dl.dropboxusercontent.com/u/8687414/Kemp…_pre-refine.wav
https://dl.dropboxusercontent.com/u/8687414/Kemp…post-refine.wav

I've tried a few things already. Tried with an external mic-pre, tried with mic directly into the Kemper. Same result. Tried changing a few parameters in the "Amplifier" section of the stack to make it break up better but couldn't get it closer. Tried adding some bass in the tone stack but it's not the same; the palm mutes don't react the same either way.

Any tips on what more I can try to get closer? Especially with the Orange amp - or is that one just really difficult to profile?