Posts by hm2

Cool. Which amps? Profiles on the exchange, or ones you've made?

I didn't see the thread in question, and it sounds like that's probably a good thing.

To the point of the original poster's question: I love the cab-swap feature -- it's one of many examples of the kind super clever, out-of-the-box (pun intended) engineering that the KPA in general exemplifies. It's a great way to get in the ballpark of what an amp would sound like running through a particular cab. However, it is not really an accurate swapping of cabs in the way that the profiling process is a literal, and often highly accurate, direct measurement of the sum of amp, cab, and recording chain.

In general, when cabs are swapped to new amps, cabs win. You can test this very easily. Just profile the same amp at two different tone and gain settings without changing the cab or mic, and swap the cabs on the resultant profiles. The profile tones will change -- sometimes substantially. The cab from the brighter profile will port this brightness to the darker one, and vice versa. In effect, the amp tones are to some degree being blended.

There's nothing wrong with this. It's still a super cool feature and I'd rather have it than not have it. I just don't think it should be taken as literally as the profiling process in terms of its accuracy, and I think the manual should in fact be clearer about this. It took me a couple days when I first got the device to realize that the manual was, in fact, implying that you could actually move a cab to a whole different profile.

Moving forward, I'd love to be able to model amps and cabs separately -- not as a replacement for the current implementation, but as an extension of it. Rigs for which separate amp/cab profiles were not taken can still use the current method. And very conveniently so.

Incidentally, when the KPAs start shipping with power amps, this would be an ideal time to implement cab profiling. Just jack the output of the power amp into a cab, run some profiling signals, and bingo. I'd also love this process to also capture room ambience the way impulse responses do, but I'm not holding my breath on that.

In either case, looking forward to it...

Quote from gorillabendz

I saw that you fixed this issue by rolling off the low-end, isnt that correct hm2 (this is exactly what I do by default when Im recording guitar bits with some of the profiles)?

My issue is actually a little different -- I'm getting too much bass *compression*, not necessarily too much bass. This can't be fixed afterward with eq. I've had some success filtering out lows during the "refining" phase, in terms of influencing the refining algorithm to run less low end through the distortion. But it's not perfect. I have tried this on a couple high-gain amps and some profile better than others. So it seems to be amp-dependent. I need to do a little more testing on high-gain amps. If I find the results are repeatable, maybe I'll submit a ticket.

In your case: If you think the distortion characteristics in your profiles sound correct, then you're set. The exact *amount* of bass may or may not be accurate, but can always be tailored on your Allen&Heath with a twist of a knob!

Quote from r_u_sirius

PS:
my profiles came stock with your Kemper??? are you sure?

Ok, that's funny!

I just saw "Christoph" the other day and figured they must have been stock. But it's quite possible I installed them on the first couple days when I had no idea what I was doing.

In any case they sound good! I don't have a VH4 in front of me, but since it's a high-gain amp I've played, I'm interested in heading back to a studio that has one, to run a few profiles -- just to see if it's an amp that happens to profile easier than mine, which I've had some trouble with.

Did you run through preamps, or just the Heil directly into the Kemper? What was your gain setting for the Ch3 profile? I have some test clips I did at about 12 o'clock the last time I played one, and that felt about right for high-gain rhythms.

Quote from gorillabendz

Now what I always thought was that the KPA is simply a xerox machine - and it copies what it 'hears', so - the excess/muddy bass I was hearing was a function of the profile, and how the mics were setup (rooms have a sonic signature and mic'ing is an art - as I have learnt the hard way ) when the amp was being profiled.

Is this a correct assumption?

Hi! Thanks for the details. If I understand you correctly, you're saying that the bass response of your profiles sound different depending on the environment you play them in? If that's the case, the good news is that's most likely a monitoring issue, and not really a Kemper issue at all. As has been pointed out, room response can have a big effect on frequency response, particularly in the bass. Room modes can easily punch a big hole in key frequencies that you'd never notice without measuring.

If you want to kill an afternoon with a fun project, download a copy of a good room acoustics analysis app and run a few impulses in your room. I've used Fuzzmeasure ( http://www.supermegaultragroovy.com/products/FuzzMeasure/ ). You'll need a cheap measurement mic. The Behringer ECM800 is a good one ( http://www.amazon.com/Behringe…-Microphone/dp/B000HT4RSA ) . You'll see right away what's going on in your space, and it's fascinating.

Btw Allen & Heath makes a great board, and the Ultralite is a great interface. I wouldn't worry about your gear in that respect at all!

Quote from old crow

Thanks for all the research, but it's not coming from the "character" setting. In fact if I dissable the stack section it's ( weird phasing )still there. If I dissable the reverb section it's perfect, the modulation / phasing is gone. As much as I use Wills reverse settings in a small room for headphones, we still very much need the "space in headphones" that is coming , soon I hope.

Ha -- no worries. Bring on the 'space effect'!

Cool post -- thanks for the details. I do think nailing high gain with the KPA is tricky, and I posted some clips from a Cornford Hellcat profiling session in another thread. I don't know if the algorithm can really suss out how many gain stages are involved, and what's happening between them, or if it's just measuring the overall amount of clipping and dialing that in over the top. It *is* amazing that it gets as far as it does, considering the variety of preamp designs that exist, and the various methods they use to keep things like bass response under control as the gain is dialed up through successive stages.

The VH4 profiles that came stock in my unit, by "RU Sirius", were actually modeled by Christoph -- or so they say in the credits. Unclear if they used SLO type circuits during modeling, or if those circuits simply model better for whatever chance reason. I am however sure that if these guys really sat around with a bunch of high gain amps and went super micro picky on it, they'd nail it. But you do have to be a bit of high gain aficionado to want to do that.

If you've got any A/B Diezel clips I'd love to hear them -- I've played VH4s and will try to get my hands on one to model if I can.

Quote from Will_Chen

You should experiment more with the character knob.It only creates an excessive comb filter effect if the cab has that type of tone fairly prominent to begin with (ie the profile was not close miced when created). I dial up the character when I want more resonance out of the speaker as if it was room miced when the profile was close miced, I feel it is fairly effective at achieving that. My "space" emulation has zero to do with the speaker character, but taking one of the reverbs and turning the tail to zero and cranking the mix up close to 100%/100%. The combination of both is about as close to the illusion of playing in a room as I could dial in monitoring by headphones. My goal is not to recreate "the classic guitar recordings of last fifty years", but to simply get tones pleasing to my ears.

I've experimented quite a bit with the character knob (see above). In small doses, I like it -- it's a cool effect.

Re: the "amp in room sound", I think I was quoting your instructions when I said "crank the character knob". You're in the WiKPA document, page 104. Although not attributed -- you should do something about that!

In any case the poster above was asking about the phasey sound, and it's the character knob that does that, not your reverb settings. I know you know that, I was just 'splaining him.

Cool technique, thanks for sharing!

Quote from Per

I'm not sure that's not what the character control is doing (amplifying the peaks and valleys of the frequency match), do you have any reference for that?

I do! Here's a cool little test I ran:

[Blocked Image: http://img850.imageshack.us/img850/3005/character15.gif]

Amplifying the peaks and valleys is immediately what it sounded like it was doing, and as it turns out -- that's *exactly* what it's doing. It's progressive, starting with the higher frequencies first and working its way down. It mostly leaves the mids alone until you get to 5, by which point it's more of a special effect anyway.

Cool stuff!

Cool -- will give 'em a spin!

Quote from old crow

I use Wills method all the time but it usually adds a weird phasing / modulation.
I'm really looking forward to the headphone space.

Agreed. This is because cranking the "character" knob in the cab settings is just amplifying the peaks and valleys in the speaker's frequency response. This kind of hyper notchy sound is exactly what happens through phase cancellation when you mike up an amp in a tiny room like a closet. We associate this phasey sound with space, hence the illusion. But it's not actually adding space in any real sense. You can get the same effect if you tone match a guitar sound in in Logic's match EQ, pushing the slider past 100% -- all the peaks and valleys get taller/deeper. It's not "realistic" emulation of space in any literal sense, but if it takes the edge off in headphones, so be it.

For better or worse, the classic guitar recordings of last fifty years have almost all been close-miked, "one-dimensional" sounds that are a little scratchy to listen to in phones without a whole band around them.

Quote from Guitartone

Now that you're happy with it and have it saved, out of interests sake, have you tried your Profile with a different cab, perhaps a Tills cab?

What, cheat on my amp with some other guy's cab? Are you trying to get me in trouble??

Just kidding. No, I haven't. Are you saying, search for author "Till" and take the cabs off his profiles? Any profiles in particular?

The Hellcat profiles are up on the rig exchange as "Hellcat 664" -- those are the tone settings, bass/mid/treb. I gave you an SM57 and a Royer 121. Both on a pine cab with H30s. These are mid-focused speakers that mix well with more scoopy / modern sounding speakers like Vintage 30s. For example if you were to cut your rhythm tracks on a V30, you could play your leads on an H30, and they'd fit together like puzzle pieces.

Thanks!

Quote from DonPetersen

those are some pretty drastic measures, that shoudn't be necessary at all to get a accurate profile. I'm still convinced that something odd is going on there.

I thought so too, but the other Hellcat clip I posted, "Hellcat red", is the "Conford Hellcat Red" profile from Einar Nysted on the rig exchange. And when the gain is matched to the profiles I ran (7.5), it exhibits the exact same issue. So I'm inclined to think this is just the way it is with my amp. Of course, if you aren't recording test clips and A/Bing, you'd probably never notice.

I'm sure this varies amp to amp. I downloaded some Engls off the exchange, and they sound pretty clean. But those amps are famous for their extreme bass tightness at high gain -- it's almost as though they are internally high-passed. So they probably profile well.

Nevertheless if anyone else owns a Hellcat and wants to run a few profiles on the red channel at max gain, that would be interesting to compare.

Anyway, fascinating device, and a learning experience to boot!

Hi all -- thanks again for all the input. Good news, I've got bit of a hack that works pretty well for nixing the compression issue: MXR 10-band eq pedal during refining. Here's the skinny:

Amp:
https://soundcloud.com/graphics8/hellcat-amp3

Kemper:
https://soundcloud.com/graphics8/hellcat-kemper3

The dynamics are now very similar, and personally I'm hard-pressed to tell the difference. This was done with the MXR between the guitar and the Kemper during refining. All three bottom bands -- 31hz, 62hz, and 125hz -- full cut, -12db. The 250hz band only a quarter cut, about three db. Bingo.

Note that with the eq, you're not looking to alter the frequency response of the resulting profile. You're really just looking to influence the dynamics/distortion modeling in the refining phase. For whatever reason, filtering out the bass during refining does not create a big hole in your tone -- the resulting profile still has plenty of sub. It simply causes the refining algo to remove sub from the distortion path, and that creates more headroom, less compression, and life is good.

Fyi this was done with my preamps inline, at a return level of -22, and a short refining phase of about a minute. It's also highly repeatable -- I ran this multiple times in a row on the same settings and got pretty similar results each time. The "definition" parameter wandered a little, between 6 and 9, but the frequency response was identical, and the results always sounded the same. I'm pretty much convinced at this point that the manual can be taken at face value on the refining process, in that nothing particularly special needs to be played here, and that super long playing times are not really necessary.

Don -- thanks for your attention here and no intent on my part to sound overly negative. I have to laugh that I can post a thread with "problems" with a space-age device from Germany that exactly recreates the sound of my expensive boutique amp from England so I can play it in my living room at whisper volumes at 3am. Next I'll be complaining about the price of gas in my Lamborghini. As Louis CK once said, everything's amazing and nobody's happy!

I'll upload the final profile to the rig exchange for those who'd like to try it out.

Thanks!

Wow, thanks for all the details Per.

I've put in a little more time on this, and more or less just by repeating the process, I've got a handful of profiles that are much closer. Here's the new A/B:

Hellcat Amp2
https://soundcloud.com/graphics8/hellcat-amp2

Hellcat Kemper2
https://soundcloud.com/graphics8/hellcat-kemper2

The squish is still there on the chugs, but less evident. This was basically luck of the draw. As a test of all the suggestions above, I ran a bunch of profiles with and without preamps, and with various lengths of refining times. There were close matches among all of these combinations. This one happened to be straight in, bypassing the preamps, with a short refining time -- one or two minutes, again playing just the test riff.

With no further processing in Logic, here's what the match curve looks like:

http://img90.imageshack.us/img…hellcatkempermatcheq2.png

Still a big rolloff necessary in the bass, but quite flat above that. With a simple high-pass in Logic at 100hz, 18db slope, this is what the match curve looks like:

http://img24.imageshack.us/img…llcatkempermatcheq2hi.png

Flat. In the Kemper clip above you're hearing the hi-pass.

So. What's the takeaway here? I've probably profiled this amp about twenty or thirty times now at the same settings. To those who may be daunted by the widely varying reports of the profiling process, here are my seat of the pants conclusions:

1. I don't see a huge correlation between preamps / no preamps as far as accuracy is concerned. All you guys with your Neve 1073s, more power to you.

2. There doesn't seem to be much correlation between refining time and accuracy. I did one for 7 minutes that was over-compressed squashed, yet this one above was only a minute or two. My best guess is that once the box hears what it is looking for in the refining process, further playing is simply moving those values around. This is good news, since the manual has almost no instructions for correcting things you hear during refining anyway. You may as well just profile the amp a handful of times, with some basic refining each time that reflects the types of riffs you plan to play. Because there's some analog voodoo in the process, you'll get different results each time. Pick the one that's closest.

3. When it comes to frequency response, the Kemper's match eq is really accurate from 200hz all the way to 10k. Preamps or not, long refining or short, that part of the match curve is usually pretty spot on. Kudos.

4. Bass accuracy and compression seems to be its achilles heel. I ran a handful of crunch profiles, and bass content below 200hz was almost always overblown. My guess is that this is simply more evident in high gain profiles, where excess bass results in excess compression -- much the way feeding too much kick drum into your mix bus compressor will take down your whole mix. I suspect certain amps are tougher for the Kemper to profile in this respect. The Hellcat may very well be one of those, since in reality it exhibits very controlled bass response at high gain with no squish at all.

To be fair to the Kemper, overblown bass is an issue in real high-gain amps as well. Amps like the 5150 bleed off bass in between gain stages to keep this under control. You can't simulate this by turning down the overall gain, or lowering the bass eq on the amp. The Kemper's "definition" control may work this way, as it seems to reduce the squish somewhat. I can't really tell. But the most accurate profiles I did were the ones that came out of the oven with lower definition values *and* less audible squish to begin with. This is where the profiling process gets black box. In the clip above, definition was set at 7.5 after the profiling process and that's where I left it. Again, luck of the draw and simply a result of running multiple profiles.

Moving forward, I'd love to see the profiling algorithm improve when it comes to crisp bass response at high gain. Or at least some more exposed parameters for adjusting preamp compression, or thinning bass response inside the preamp, to alleviate the sag.

Nevertheless, quite a disruptive little piece of technology. Again, kudos.

Thanks!

Hi Christoph. I once had it stick at its highest setting -- I kept getting "amp too noisy" errors and then I noticed what was happening and reset it. Not sure what caused that, but it hasn't recurred.

When I'm running through my JLM pres, the return level is down around -20. Is 0db return level preferable over a negative return level? Presumably there is a preamp in the Kemper to handle directly connected mics. What return level is minimum gain on the internal pre?

Big fan of Sneap's approach, and super chill nice-guy attitude. I've seen his Kemper video, but I can't really tell from the Testament clips he plays if he's getting the same sort of artifacting I'm getting. Given his ears, I would have to think he'd hear it if he was. It could be that this issue is specific to the Hellcat.

Will try the preamp-less approach and report back.

Thanks for the speedy response guys. So you're saying the compression issue is not typical of high gain profiles and can be avoided?

Re: Miking directly into the Kemper, will give it shot once SnowPocalypse II blows over. That being said, the JLMs are nice pres and part of the sound of my recording chain. Are we saying that optimal profiling really only happens when jacking a mic directly in the back?

Hi-

First post, thanks for having me. Still getting acclimated to the Kemper, and have not yet nailed my main amp, a Cornford Hellcat. In particular, I'm having trouble getting accurate dynamics when then gain is on.

Here's a quick A/B. Signal chain:

Les Paul Classic -> Hellcat, red channel, max gain -> pine cab -> H30 -> SM57 -> JLM Audio preamp -> Kemper -> Lynx Aurora -> Macbook.

Amp clip:
https://soundcloud.com/graphics8/hellcat-amp

Kemper clip:
https://soundcloud.com/graphics8/hellcat-kemper

What I'm hearing here in the Kemper track is over-compression, particularly in the bass. Whereas the amp track is stiff and chuggy, the Kemper track sags. This squishy sound is not unlike what happens when you reamp a track too hot into an amp. However both of these clips were live -- no reamping.

In addition, the Kemper is picking up a ton of sub. The '57 does not put out much below 100hz, but the Kemper seems to think so. You should be able to hear this in good headphones or studio monitors. Additionally, here's the Match EQ curve in Logic:

http://img542.imageshack.us/im…/hellcatkempermatcheq.png

Aside from the huge bass roll-off, the rest of the curve is actually very close. Reasonably flat from 200hz to 1k. And dead flat from 1k to 10k. You can click this curve on and off in Logic and the mids and upper mids sound essentially identical.

So... If this were a real amp, I'd assume I was simply going into the amp too hot. The Les Paul has a 500T in the bridge which is a very hot pickup, but I have the "Clean Sens" at 4 and I'm not getting any red lights.

Dropping the gain (or "Dist Sens") doesn't help - that just reduces the overall saturation, not the bass strangulation that's happening here.

Definition is already at 9.7 -- nowhere to go from there.

Filtering out bass using "Studio Eq" before the stack just eviscerates the sound. And frankly should not be necessary if the profile is correct.

As an additional test, I installed the "Hellcat Red Channel" profile from the rig exchange, increased the gain to 7.5 to match my profile, and swapped cabs with my profile. Here's what that sounds like:

https://soundcloud.com/graphics8/hellcat-red-kemper

Still pretty squishy.

Thoughts? Is anyone else having this issue with high gain profiles?

FYI:

For the "refining" phase, I simply played the same riff a few times. From the manual, Wiki, and forum, there seems to be endless speculation and no real consensus on what the box actually needs to hear during refining. My own experimentation has produced widely varying results -- "definition" values and frequency curves all over the map. For the mean time, I've settled on using the test riff itself to introduce the fewest variables.

Moving forward, more transparency on the refining phase would be welcome. If any guitar will suffice, as the manual states, then the Kemper should probably just play its own refining signals to eliminate the needless fumbling in the dark. By comparison, creating IRs of speaker cabs is a clear-cut, repeatable, and highly accurate process.

Thanks!