Today, I want to answer one question: can vintage profiling tech recreate the sounds of vintage amps?
Read on to find out.
Table of Contents
- What Am I Trying To Achieve?
- Why Do I Want To Do This?
- Why Am I Making My Own Profiles?
- What Tweed Tones Am I Going To Profile?
- Methodology
- Criteria
- Technical Notes
- DanDrive Tweedy 5B3
- Mad Professor Sweet Honey Overdrive
- Mad Professor Amber Overdrive
- Mythos Lark
- Do All The Profiles Sound The Same?
- My Conclusions
- The Kemper Can Profile Tweed Tone
- The Kemper Doesn’t Make Super-Accurate Profiles …
- … But Refining Profiles Sure Does Help!
- The Profile Accuracy Isn’t An Issue In Practice (With One Caveat)
- All My Tweed-Tone Profiles Are Usable In A Mix
- I Would Use The Kemper For A Featured, Solo Part (But With Care!)
- Go All-In, Or Don’t Use The Kemper At All
- Final Thoughts
What Am I Trying To Achieve?
I want to know if my Kemper can create profiles of my tweed-tone rig.
Why Do I Want To Do This?
That’s the same as asking: why can you find a Kemper in professional recording studios all around the world, and at the heart of many live performance rigs?
Because the answer is the same: it allows you to pull up the exact same sound weeks (sometimes months) later. That’s something that I cannot do with my hybrid rig.
When I start recording my music, I know that some of these songs aren’t finalised yet. I’m going to be going back and changing arrangements. I’m going to be going back and adding new sections. And accurate profiles will allow me to do that without having to re-record all the guitar tracks.
If the Kemper can’t do this for me, then I need to finally move mine on. (I’ve been saying that for almost seven years now, so don’t be surprised if I still don’t …)
Why Am I Making My Own Profiles?
Last year, I did a comparison between the Kemper profile and my real Tweed Deluxe amp. Spoiler alert, it was not favourable.
That blog post, though, relied on a third-party profile that I’d found on the Rig Exchange.
The Kemper website and Rig Exchange are currently offline. There’s a holding page that says it’s all offline for “security reasons”. They have my sympathies there.
That means that I can’t find a better profile (I can’t even try that profile again). And anyway, the Kemper’s original mission was to profile your own gear. I’m far more interested in that.
What Tweed Tones Am I Going To Profile?
I’m going to make profiles of up-to four different tweed tones:
- my DanDrive Tweedy 5B3 through my vintage tones preset,
- my Mad Professor Sweet Honey Overdrive (SHOD for short) through my vintage tones preset,
- my Mad Professor Amber Drive through my blackface tones preset,
- and my Mythos Lark through my vintage tones preset.
I say “up-to four” because I’ll probably stop if the Kemper struggles with any of these.
Methodology
For each rig, I’m going to create five different audio demos:
- my hybrid rig, with all delay and reverb added in post
- my hybrid rig completely dry and mono (ie the signal that will be profiled)
- the Kemper’s initial profile (ie before profile refinement), completely dry and mono
- the final Kemper profile after refinement, completely dry and mono
- the final Kemper profile, with all delay and reverb added in post
I’ll be comparing the two demos that have effects on them.
I’m adding delay and reverb in post because there’s a huge gulf in effects quality between the Axe-FX 3 and the Kemper. I don’t want that influencing the outcome.
I’m actually going to throw away the first attempt at each Kemper profile. I use that first attempt to figure out what the ‘Return Level’ setting in the Kemper should be. Once that’s stable, then it’s worth trying to make a profile to evaluate.
Criteria
How will I decide whether or not the Kemper profile is close enough? I’m going to be looking at two things:
- The overall EQ of the profile must be very similar to my hybrid rig.
- The overdrive characteristic of the profile must be indistinguishable to that of my hybrid rig.
Whether it’s low-end, low-mids, mids, high-mids or treble, I need the profile to have the same energy in the same frequencies. It doesn’t need to be exact. The profiles are for recording music, so I can accept the profile if it’s lost a bit of low-end. But it’s got to sound close in the frequencies that matter in a mix.
It’s also got to sound like an overdriven tweed tone. The profile’s got to capture the raspiness / fuzziness of the hybrid rig. It’s got to reproduce the note attack. And there’s no room for error here. These are characteristics that will be heard in a mix, and which also affect my performance. If it’s not spot-on, it’s not good enough for me.
One thing I’m not assessing today is playing feel. Frankly, at this stage, it doesn’t matter. I need to create Kemper profiles that sound convincing first. Once I’ve got those, then I can spend some time with them to form an opinion on what it feels like to play through them. But only then.
Technical Notes
My signal chain is:
- guitar
- into the Kemper
- out of the Kemper into my pedalboard
- from my pedalboard into the Axe-FX 3 (for amp and cab only)
- back into the Kemper
- out to my audio interface
- and into my DAW.
The guitars that I’m using aren’t important; they’re not part of the signal chain when the Kemper makes a profile.
I’m not reamping the same performance. Each audio demo is a live playthrough. Partly because I’m not setup to do reamping, and partly because I think it’s important to capture how I respond to playing through the different tones.
All demos are recorded through the Kemper – even the hybrid rig that is being profiled. (I didn’t want to constantly switch cables around.) If the passthru signal in the Kemper (aka the Reference Amp signal) is filtered / coloured in any way, that will affect those demos.
I’ve disabled all effects inside my Axe-FX 3 presets, and I’ve configured Output 1 to sum to mono. I’ve also disabled the noise gate in my presets.
I’ve also disabled any compressors that I might normally use, including the Mary Cries. The Kemper’s profiling method simply isn’t compatible with a compressor. (Yes, I found this out the hard way!)
Unless stated otherwise, I’m applying no EQ inside the Kemper: no EQ changes when making the profile, and no EQ changes (either to the Amp block or by adding a Studio EQ block) after the profile has been made.
Audio gear normally outputs signals at one of two levels: -10 dBV or +4 dBV. If your audio interface supports both (mine does), then you need to set it to expect whatever line out level your other audio gear produces.
I’ve had to put both the Axe-FX 3 and my audio interface into -10 dBV mode. Unfortunately, the Kemper doesn’t seem to support +4 dBV mode. For recording, I have the Kemper’s Master Volume set to -12 dB.
Finally, I did a full factory reset of the Kemper before starting this session, and I updated it to the latest firmware (10.2.2.49572). This makes sure that the Kemper is operating as it’s meant to.
Right, let’s get into it.
DanDrive Tweedy 5B3
I’m starting with my DanDrive Tweedy 5B3 because I’ve just wrapped up my First Impressions post for it. Everything’s already dialled in and ready to go.
I’m profiling:
- Ceriatone Centura
- DanDrive Tweedy 5B3
- my vintage-voiced pedal platform preset
Here’s the hybrid rig, with delay and reverb added in post:
and here’s the final Kemper profile, with delay and reverb added in post:
I think they’re pretty close.
- The overall EQ in the profile isn’t bad. Yes, the profile is missing some of the low-mids, but everything else is there.
- The Kemper has definitely captured the character of the overdrive. I think it’s added a little extra gain compared to the original, but otherwise I think it’s all there.
Now let’s take the effects away, so that we can really hear any differences.
That’s definitely helped. I’m hearing clear differences between all three of them.
The unrefined Kemper Profile sounds thin to me. It just sounds like there’s holes in the sound, compared to the hybrid rig that I’m profiling.
Refining the profile seems to have fixed that, while adding a bit of extra gain and/or top-end (I’m not exactly sure which). While it isn’t spot-on, I suspect it’s close enough if used in a mix.
Mad Professor Sweet Honey Overdrive
I’ve picked this as my next pedal because it’s my go-to tweed tone. This is the signal chain that I care about the most. If the Kemper can’t profile this, then what’s the point?
I’m profiling:
- Ceriatone Centura
- Mad Professor Sweet Honey Overdrive
- my vintage-voiced pedal platform preset
Here’s the hybrid rig, with delay and reverb added in post:
and here’s the final Kemper profile, again with delay and reverb added in post:
I think that’s very close indeed.
- The Kemper profile is missing a bit of upper-mids emphasis, but it’s quite subtle. I didn’t spot it on the first listen back.
- The overdrive character and note attack seems spot-on to me. If there’s a difference, I can’t hear it in these demos.
Are the effects hiding differences between my rig and the profile? Let’s listen to the dry audio.
That’s a bit more revealing, I think.
I’m hearing a significant difference in the mids between my rig and the unrefined Kemper profile. For some reason, the Kemper seems to have boosted the mids a bit – or does it just sound that way because the low-mids are missing? Either way, refining the profile seems to have sorted that out. Not perfectly, but pretty close.
I still can’t hear any differences in the overdrive character or note attack. I think the Kemper’s nailed that.
Mad Professor Amber Overdrive
I’ve been using the Amber Overdrive as one of my reference pedals for dialling in my new pedal platform preset. It’s a challenging pedal, and it’s taken me almost a decade to finally find a signal chain that suits it.
I’m profiling:
- Mad Professor Amber Overdrive
- my blackfaced-voiced pedal platform preset
Note that, this time, I’m not boosting the drive pedal with the Centura at all. The Amber Overdrive doesn’t need it.
Here’s how my hybrid rig sounds, with effects added in post:
and here’s how the final Kemper profile sounds, with effects added in post:
I can’t hear any real difference between those two.
So let’s listen to the dry tracks, to see if that helps reveal any differences between my rig and the Kemper profile.
Woah. Now I’m hearing clear differences between my rig and the profile.
To my ears, that undefined profile isn’t remotely close. Both the tonality and the overdrive characteristic are way off. It’s so honky, as-if the Kemper just can’t hear many of the key frequencies. And so horribly, horribly boxy too.
Has to be said, I think refining the profile has brought it right back into contention. I’m still hearing some honk, but it’s nowhere near as prominent now. And the overdrive characteristic and note attack has been added into the profile.
Mythos Lark
I’ve picked this because it’s a different flavour of tweed tone. It’s chasing the old Gibson amp sound rather than the old Fender amp sound.
I’m profiling:
- Ceriatone Centura
- Mythos Lark
- my vintage-voiced pedal platform preset
I’m curious to find out whether this profile ends up sounding like all the others that have the Centura in the signal chain.
Here’s how my hybrid rig sounds, with effects added in post:
and here’s how the final Kemper profile sounds, with effects added in post:
Hmm. I’ve got mixed feelings about that. It’s definitely tweedy – and usable – but does it really capture the tonality of the Lark? I’m not sure it does.
Let’s listen to the dry audio, and hear what’s going on.
The Lark’s an interesting test because it’s all mid-range. There isn’t much low-end or top to hide behind. With the Centura adding additional Klon-style mid-push to the tone, any profiling inaccuracies are going to be magnified.
And that’s exactly what I’m hearing.
Do All The Profiles Sound The Same?
What do the profiles sound like, when compared to each other? Has the Kemper heard the same thing every time?
Three of the profiles use a Klon klone pedal as a clean boost; only the Amber Overdrive profile doesn’t have it. Also, those same three profiles use the same speaker IRs.
As a result, it’s no surprise that the Tweedy 5B3, Sweet Honey Overdrive and Mythos Lark profiles do have a similar sound. Especially if you’ve spent quite a bit of time listening to all the demos I’ve made for this blog post.
I think that’s an important criticism of the Kemper’s signature tonality. I find it ear-fatiguing to work it, because it’s such a narrow sound.
My Conclusions
Now that I’ve gone through all 20 (!!) audio demos, what do I think?
The Kemper Can Profile Tweed Tone
First and foremost, I feel that the Kemper profiles that I’ve made today do sound like tweed tone. No-one’s going to mistake these profiles for (say) a JCM 800 or a 5150.
In particular, I think it gets both the overdrive characteristic (that raspy / fuzzy nature) and note attack close to spot-on.
The Kemper Doesn’t Make Super-Accurate Profiles …
When I take a listen to the dry audio, I can hear clear differences in tonality. Worse still, these differences are in the mid-range, which is the most important part of a guitar’s tone.
The frustrating part is that there’s only one tool available to try and fix this.
… But Refining Profiles Sure Does Help!
In every example, it was worth me going through the profile refining process. Every profile got better, and the Amber Overdrive profile went from unusable to very usable.
The Profile Accuracy Isn’t An Issue In Practice (With One Caveat)
I know that the profiles aren’t accurate because I can A/B them against the original sound. Without those comparisons, could I tell that they’re not accurate? Almost certainly ‘no’, because I wouldn’t know what the profile should have actually sounded like.
That’s why – as long as I use them right – I don’t believe that the accuracy issues are going to be a problem in practice. Except for one instance that I can think of, and that’s down the Kemper’s signature tonality rather than an accuracy problem in specific profiles …
All My Tweed-Tone Profiles Are Usable In A Mix
When I added in delay and reverb, every refined Kemper profile sounded very usable. Effects sure do hide a multitude of sins!
I Would Use The Kemper For A Featured, Solo Part (But With Care!)
The Kemper has long been notorious for how it never captures the low-end of a guitar profile. That can make it difficult to go with a Kemper profile for a sparse mix where guitar will be featured, especially when playing unaccompanied.
If I was going to do this, I’d carefully audition the Kemper profile first, and see how much low-end could be restored in post (for example, by using a Pultec EQP-1A emulation) – and whether or not that low-end sounded natural enough to use in this situation.
Go All-In, Or Don’t Use The Kemper At All
I wouldn’t record a track using a mixture of the original signal chain and the corresponding Kemper profile. I think you would hear the difference when switching between the two.
As a result, I’d only use the Kemper profile if I did 100% of the guitar track with it.
Final Thoughts
That didn’t go as expected. But it also did.
I did a test run a couple of nights prior, to prove that I had everything cabled up right. It … did not go well, to put it mildly. In fact, the results were downright unusable. So I went into this session expecting the same.
I’m pleased that I got clear and usable results this time around.
Got some humble pie to eat over that one: I’ve been telling people for the last few days that the Kemper can’t do this. I was definitely wrong about that.
The results still have that signature Kemper tonality to them, though. I’ve seen Kemper dealers pass this off as “mix-ready” tones, and they’ve probably got a point. I might even take it a step further and say they’re “radio-ready” tones.
Still, I’m disappointed that Kemper hasn’t addressed the shortcomings of their profiling approach by now. They’ve had well over a decade to do so. They have improved other areas, such as the onboard effects, but that’s mostly shown just how far the hardware has fallen behind now.
To put the age of the Kemper hardware in context, it’s the equivalent of still using an iPhone 4 today. I can’t think of any other piece of digital tech that has remained the same for so long.
Fair play to them for longevity – and for keeping their tech out of landfill. But can we please have an updated profiler that produces much more accurate results? I think it’s time.