Treble detail in profiles woes

    Quote from ckemper

    Per, can you link me to those examples again?

    I have only found the one without noise.

    Hi Christoph,


    Every single example was recorded this way in "Profiler" mode immediately after recording. The only one that was recorded in "Browser" mode was the merged profile. I'm not sure why you don't believe me on this.

    When is the background noise loop recorded exactly? I'll make a new profile and clap my hands or something at that point in order that we can clearly hear to check that the profiler is in fact recording the loop and upload that.

    Here's every single example so far :

    Kemper (DAW->Kemper->DAW) : http://peranders.com/music/demostuff/Kemper_per_reamp.aif

    Amp (DAW->Kemper->Amp->Mic->Kemper->DAW) : http://peranders.com/music/dem…ic%20Signal_per_reamp.aif


    Kemper (DAW->Kemper->DAW) : http://peranders.com/music/demostuff/Kemper_per_reamp2.wav

    Amp DAW->Kemper->Amp->Mic->Kemper->DAW) : http://peranders.com/music/demostuff/SM58Amp_per_reamp2.wav


    Profiling Session itself (Kemper:DAW->Kemper->DAW, Mic:DAW->Kemper->Amp->Mic->Audio Interface->DAW) : http://peranders.com/music/dem…per_profilingsessions.zip (mic captured at the same time as Kemper output via audio interface rather than profiling A/B obviously).


    Both Kemper and Amp for a DI and Mic'd (Kemper:DAW->Kemper->DAW, Amp:DAW->Kemper->Amp->Mic->Audio Interface->DAW, DI Amp:DAW->Kemper->Amp->Kemper->DAW) : http://peranders.com/music/dem…/KemperStudioVsMerged.zip

    Resulting Profiles : https://www.kemper-amps.com/wb…5-permesacrunchmands-zip/

    Edited once, last by Per: Adding signal chain routing info ().

    Ok, to prove that my Kemper is working and the noise loop recording is working this time I jammed my iPhone up against the mic playing music and made a profile. This actually resulted in a profile that was much more brittle/bright than the amp!


    Anyhow, the results are here, all levels are un-adjusted, i.e. exactly as they come from the sources.


    The Kemper (A) (DAW->Kemper->DAW) : http://peranders.com/music/dem…_Profile_cousin_noise.aif

    The Amp (B) (DAW->Kemper->Amp->Mic->Audio Interface->Kemper->DAW) : http://peranders.com/music/dem…m_Kemper_cousin_noise.aif


    Additionally I recorded the mic line output during the Kemper re-amp directly from the audio interface i.e. amp->mic->audio interface->DAW. To be absolutely clear I have done this before in the "Merged" profile example zip file and during the "Profiling session" example as there was no other way to capture the mic at the same time as the Kemper only outputs the Kemper signal. Elsewhere (the first two profiling sessions) it was amp->mic->kemper->DAW. In this session the signal chain was amp->mic->audio interface->kemper->DAW.


    The Amp recorded simultaneously during the Kemper reamp via the audio interface (DAW->Kemper->Amp->Mic->Audio Interface->DAW) : http://peranders.com/music/dem…9_Silver_cousin_noise.aif


    And just to ensure we have complete data (outside of the profiling session itself which I'm happy to redo if required), I've attached the resulting profile of pain.


    PER-Mesa_Cousin.kipr.zip


    Everything here was recorded the exact same way in the other examples. I'm going to edit my previous post and add in the signal chain for each entry so that it's 100% clear.

    Per,


    The cabling seem to be correct.

    However, it is hard to comprehend for me, that you made this profile with the audio track playing in the background, disturbing the listening, and the profile now comes out with a hard distortion.


    I don‘t doubt that your Profiler is working well.


    And a lack of high frequencies (the thread title) doesn‘t seem to be the predominant issue ...


    Just a recording of the regular A/B after the profiling would help.


    What amp was it again?

    Edited once, last by ckemper ().

    Hi Christoph, maybe we’re not talking about the same thing with A/B recording? To me that means immediately after making the profile recording the Kemper and the amp while the Kemper is still in profiling mode and using the Kemper/Reference Amp soft buttons (used to be labeled A/B) to switch between both.


    That is what I have been doing. Is that not recording of regular A/B?


    I cannot explain any difference in noise floor however in my last post I gave you all the files at the levels they appeared at, in others I matched rather levels in the DAW (except where I said I didn’t). You can hear that the profile is louder than both the mic direct from the audio card which is expected as mic level is much lower than line level, and “B” which is interesting as I would expect the Kemper to match up levels much closer, I had manually matched them but this was the third profile attempt because I couldn’t believe that the sound was meant to be that bright and it seems the Kemper auto adjusts the levels after each profile. Is it possible you’re not taking into account the additional gain used on the mic tracks to bring them up to level in the prior examples?


    By the way I would very much like to be able to request support for line level return volumes if it is possible. That way I could use the levels indicator on the UAD Console app in a meaningful way when routing the mic signal through the card.


    The amp used is a Mesa Express 5:25 in class A (5 watt) mode., in this example I still had the THD Hotplate attached to it from the previous session as honestly it seemed to improve the outcome. The second amp (in the second example from the thread) was a small Marshall practice amp, not a tube amp so I didn’t expect it to be that accurate although it was surprisingly so outside of the frequency response.

    If you could link me to that A and B file again from profiling mode, without iPhone backing track and without attempts to match the levels. The background noise level should thus match.


    I am in the road just with an iPad, thus I cannot open zip files.


    Thanks!

    Sure.


    All of the examples thus far have A and B files from profiling mode though.


    But I've made a new one quickly for you with a new profile attempt, and I've placed all the A/B onto a single track in sequence.


    First there's a section of re-amp (so you don't also have the guitar strings sound in there on the "reference amp" side) then there's me just playing the same riffs A (Reference amp, then Kemper).


    The order throughout is Reference amp first, then Kemper second.


    The Kemper A/B Track (me just switching the a/b switch after calling out what I'm playing) : http://peranders.com/music/demostuff/AB_Kemper_AB_Switch.aif

    The DI track (used for the first portion only) : http://peranders.com/music/dem…B_DI_Guitar_for_Reamp.aif

    The Mic only for the whole session as recorded from the audio interface : http://peranders.com/music/demostuff/AB_Mic_Only.aif


    The profile is attached here : PER MESA AB - 2019-07-15 22-17-39.kipr.zip


    The levels are as recorded, not adjusted at all.

    Edited once, last by Per: Edited to fix the links ().

    I have listened to the A/B clip.


    While i cannot judge the first half, since the reference amp has (again) a different level (you can tell by the background noise level), the second half seems to be a valid A/B comparison.


    I think the profile is spot on. The biggest difference I hear is that the reference amp is mic‘ed in a quite reverbarant room, the mic catches lot of room reflections. The profiler does not catch those, still it will take their frequency response into account. The high frequencies seem to equal beside this. Especially the highs of the cleans are perfect IMHO. What do you think?

    The trouble with the "second" portion of the AB is that when it goes over to the Amp you're also getting me strumming away in the background, in the first portion as it's a re-amp you're getting effectively what you'd get if I had a nice studio with a control room to work from, i.e. nothing but the amp sound, so that should be much closer.


    Anyhow regardless of that I agree, the room verb is a big part of this.


    As I mentioned in nightlight's thread one thing that made a big difference to getting closer to the sound was using a convolution reverb in DAW (although I don't have an IR of my room, I found one that's close enough in Reason labeled "Studio" which is ironic as my space is completely untreated with a tile floor!). I wasn't really able to match using the Kemper onboard reverb.


    I would seriously love if the Kemper would optionally capture a second or two's IR during profiling and had a basic convolution verb built in to the cab section that could use this, as I think the verb is a huge (final missing piece) portion of the sound, not just for the room but for the cabinet (as opposed to the speaker) with profiles. Also it would finally allow me to do distant mic'd sounds accurately.

    I am fine with a KIPR format 2.0 being a larger file size too, I don't need 1000 profiles on my Kemper, especially with the new Rig Manager coming (and I'd be fine with sticking a USB drive in to store background hiss loops and IR files). So long as the Kemper can pre-load up the required data quickly enough before switching in "Performance" mode then I think this is fine.


    I'm going to carry on doing experiments. I don't know why the Kemper vs the Amp when re-amped would have different volumes. Nor why they wouldn't match up perfectly on frequency response. I also don't know why I lost all of the top end in the first profiles where the mic was plugged directly into the Kemper return, while now I'm looped back through the Apollo (no plugins are involved by the way) I'm getting slightly too much top end in the profile, although that last one I believe may be down to the rooms reverb.

    At the end of the day I just want to be able to make a good example for the Gearslutz thread to demonstrate that the Kemper does in fact match (or even exceeds) the real amp when it comes to pick dynamics, and to not show the Kemper in a bad light.

    Quote from Per

    I'm going to carry on doing experiments. I don't know why the Kemper vs the Amp when re-amped would have different volumes. Nor why they wouldn't match up perfectly on frequency response. I also don't know why I lost all of the top end in the first profiles where the mic was plugged directly into the Kemper return, while now I'm looped back through the Apollo (no plugins are involved by the way) I'm getting slightly too much top end in the profile, although that last one I believe may be down to the rooms reverb.

    At the end of the day I just want to be able to make a good example for the Gearslutz thread to demonstrate that the Kemper does in fact match (or even exceeds) the real amp when it comes to pick dynamics, and to not show the Kemper in a bad light.


    As I said in my first post, I am sure that you haven‘t lost top end while profiling. But for the A/B comparison you have left the profiling mode, resulting in the background noise being discarded and probably the volume being normalized by the profiler. The missing background noise makes sounds duller by psychology, a widely known effect, and the reason why we recreate the noise during the initial A/B comparison. We even simulate the latency of the reference amp, caused by DAWs or whatever, for creating the same feel. Of course, the latency is also discarded later.


    We have no plans or technic to capture the room response while profiling, and as you know, it is not a major part of the majority of profiling situations. I would strongly recommend to use the Natural Reverb to recreate a desired room, where you have the mix and time of the room in your hands.


    You‘ll be able to make good profiles and demonstrate them simply by following our standard route, which I am pushing since days on this thread :)

    I have not left profiling mode for the AB examples throughout this thread. Again please believe me on this, what on earth possible benefit could I have for lying? I want to get the best I possibly can from my Kemper.


    I also 100% disagree about the importance of the verb, and I think my examples here demonstrate that it is important to the sound beyond 1D frequency response and makes a significant contribution to the final sound. I believe this is the case regardless of how “lively” the room is too unless you are profiling in an anechoic chamber.


    As I explained I was unable to achieve a close approximation (or even similar) sound to my particular verb using the onboard reverb (using natural verb). It’s a similar problem to the one that the Kemper itself solved, trying to match a programmed effect to a real thing is not always easy and takes a lot of time.

    Per, just a thought but have you tried moving the speaker and mic to a different spot in the room? If your room is tiled and untreated it is possible that there is all sorts of phase cancellation happening at the point the mic is "sensing" due to reflections. If the Kemper is capturing this (even though it isn't capturing the time domain information) it might be possible it is affecting the quality of the profiles. I am only thinking aloud here but I can see how frustrating this issue is so thought it might be worth testing.

    I’ll give it a go, but shouldn’t that sort of natural comb filtering simply be present in the mic frequency response?


    I do believe Christoph about both the impact of the rooms reverb (despite him then discounting its value as soon as I requested a feature based on it), and the impact of background noise on the perception of a sound, even if in this case the Kemper for some reason heard a different amount of noise than the source signal presented. But even with background noise included it’s not a complete match and it just bugs me.

    Quote from Per

    I’ll give it a go, but shouldn’t that sort of natural comb filtering simply be present in the mic frequency response?

    I would think so too but at this stage any straw is worth clutching :) I just wonder if the lack of any time domain information responds differntly to the full mic in the room with natural reflections which arrive later could somehow have an impact.

    Quote from Per

    I also 100% disagree about the importance of the verb

    An amp without built-in reverb always produces a 100% dry sound. The "room reverb" you hear when you play the amp (and cab) comes from the room you're in. If you like this particular "room sound", the easiest way to get it is to playback the dry recording (or live playing) through speakers and record the "room" with a (far) microphone on a separate track so you can mix it to your taste. No need for an impulse response. Try it, experiment, discover. You can even setup a small speaker in a staircase and record the staircase sound in addition to your dry profiled sound. You'll be surprised what you can achieve by this and you'd be even more surprised if you knew how often this kind of playful trickery is done in studios. :)

    Yes the room sound is important, not just because it adds verb, but because it affects the perception of the frequency response. It doesn’t just come from the room, but it’s also why your cabinet sounds the way it does even when you switch speakers in it.


    I have recorded the Kemper including a mic in the room before and it works great, but doesn’t this defeat the whole purpose? No-one is going to take a room around with them on tour, only the profile-makers have access to their own rooms, and most importantly I don’t always want to be recording making a bunch of noise, otherwise I’d just be using, you know, my real amp.

    Quote from Per

    No-one is going to take a room around with them on tour

    And no one will be able to hear "your room" on tour. People hear the venue's room in the first place. If you e.g. want a small room effect on a specific part, it certainly will be part of the sound but in a large venue it will blend with the massively huge natural reverb and delays of the hall. As an example, listen to the Neal Schon solo from Bell Centre Montreal, someone posted in another thread. :)

    Once your "small room" goes to places it's one flavor of many to make a tasty melange … never a 100% authentic reproduction. Use this to your advantage. That's all I can add to the "room" part of the discussion.