A New Technique to Increase Amp Definition

[craiganderton]

I came up with this technique for the upcoming Helix book 1.2 update, but I wanted to share the concept with the forum prior to publication because I'm pretty excited about it. I haven't seen this kind of approach done before.

 

My sonic holy grail is "clean distortion," which is why I like multiband presets. They reduce intermodulation distortion, which can create a splattery, undefined sound quality. This technique is about a simpler option that nonetheless increases definition, especially for high-gain amps.

 

Place the same amp in parallel, and precede each one with a 10-band graphic EQ. Send alternate bands into the two amps. Sending frequencies that are further apart into each amp keeps the frequencies from interfering as much with each other. This interference is what creates intermodulation distortion. Note that the EQ's output levels are up +6 dB to compensate for less signal going into the amp.

 

 

The improvement is most obvious when you're playing, but these three audio examples should help get the point across. With each example, the first part uses this technique with the Fullerton Jump. The second part uses one instance of the Fullerton Jump amp with the same settings. Both parts are level-matched to create the same LUFS (perceived level) reading.

 

With Chords, you'll hear how there's more splatter with the single instance. Strums shows how strums are less defined with a single instance. Stereo takes advantage of the two amps to create a stereo spread. 

 

The difference isn't as big as using 3 or 4 frequency bands of distortion, but this technique does lead to a more defined sound quality (especially for amps with heavy saturation) and leaves enough DSP power to add other effects. I hope you find it useful!

[SaschaFranck]

Quite interesting, gonna check that out later on.

 

And fwiw, in case it's "wasting" too much CPU, we will all just create a TONEX model out of this in the near future (it's just too affordable to pass).

[soundog]

Very interesting. I can even hear the improvement on my laptop speakers! (And fwiw, I ordered a TONEX pedal; I couldn't pass either.)

[craiganderton]

On 2/19/2023 at 12:46 AM, soundog said:

Very interesting. I can even hear the improvement on my laptop speakers! (And fwiw, I ordered a TONEX pedal; I couldn't pass either.)

 

Cool! The amount of improvement depends on what you play and how you play it. For example, you probably won't notice any difference with single-note solos. But you'll hear a major difference with big, full chords that include 7ths, 9ths, etc. Basically, the more frequencies coming out of the guitar, the more youi'll hear a difference.

[codamedia]

 

@craiganderton..  I love the concept/approach, thanks for sharing that with us. I notice a huge different on my studio speakers. 

[craiganderton]

I've learned a lot about multiband techniques since writing the original book, and doing the multiband preset pack - it's a continuous process of refinement. This approach was one of those "what if...?" experiments that I stumbled on, and worked beyond my expectations.

 

There's always something new to discover in Helix!

[craiganderton]

On 2/19/2023 at 11:55 AM, MGW-Alberta said:

Very interesting indeed but I do have one request.

Could you please take your 3rd sample and sum the sides down to mono and then post that comparison?

Or in the alternate, edit your first post to include the 4th (mono) sample for ease of comparison.

I'd like to be able to listen to the difference without the stereo spread skewing how I perceive it because I run a mono rig.

 

I'm happy to say your request has already been covered :) The first sample and the third sample are identical, except for spreading the outputs of the amps in the third one. So, the first one actually is the mono sum of what's in the third sample. 

 

One of the reasons I came up with this was to have easy stereo/mono compatibilty for studio and stage. There's no delay or phase-based processing involved, so mono summing works extremely well. The 10-Band Graphic EQ has some minor phase shifts because it models analog EQ, but the audible results are of no real consequence. If it used a linear-phase design it would be perfect, but I don't think most Helix owners would be able to cope with the latency, or DSP power, that linear-phase EQ would require.

 

Maybe I'll do another audio example with complex chords, where the difference is more obvious between the processed and unprocessed sounds. But stereo summed to mono would still sound the same as mono.

[craiganderton]

Here's a demo that shows the difference better, because it uses more complex chords that send more frequencies into the amps. The first section uses this technique with two Grammatico Jump amps (and FWIW the inherent stereo is summed to mono). The second section uses a single Grammatico Jump amp. The difference is most noticeable toward the end of each section.

 

 

 

 

[SirNedwood]

It was my understanding that doing these large alternating changes on a graphic EQ causes phase issues. I'm not sure if that problem is solved in the digital realm, but I am skeptical. My trick for increasing clarity with distortion is to run 2 amp blocks, with a crossover split feeding into them. That way you can adjust the gain on the lows and highs separately.

[craiganderton]

On 2/19/2023 at 11:25 PM, SirNedwood said:

It was my understanding that doing these large alternating changes on a graphic EQ causes phase issues. I'm not sure if that problem is solved in the digital realm, but I am skeptical. My trick for increasing clarity with distortion is to run 2 amp blocks, with a crossover split feeding into them. That way you can adjust the gain on the lows and highs separately.

 

Thank you for the comment, it's worth discussing. You're correct that large alternating changes with a non-linear-phase graphic EQ can cause phase shifts. However, I've tested the 10-Band EQ technique vs. the Crossover Split, and unfortunately the Crossover Split introduces more phase shift. If you want to check it out for yourself, split a guitar track into two buses. Insert the Helix in one of the buses, and enable the Crossover Split. Sum the split's outputs back to mono in the Helix. Set both buses to the same approximate volume. As you vary the crossover frequency, you'll hear significant, audible phase shifts going on between the Helix track and dry track.

 

If you try the same setup using the 10-Band EQs, you'll hear virtually no difference when mixing the Helix and dry tracks together. If you mute one of the tracks, you won't hear much change either, except for level. The reason why we can get away with this is because as you said, the issues involve large alternating changes. There's no boosting happening, only cuts. You can reduce any possible phase shifts even further by cutting only -6 dB instead of -15 dB. However, the benefits of using -15 dB cuts outweighs any potential phase issues. 

 

The main problem I have with the Crossover Split is that it's not particularly useful for stereo because the highs are on one side, and the lows on the other. That doesn't mean it can't work, but using the dual graphic EQs gives a more even frequency distribution to the left and right channels. So, you can have wide stereo images, that also collapse well  to mono.

 

The other limitation is with traditional bi-amping, you're still feeding a continuous range of frequencies into each amp, which is more prone to producing intermodulation distortion. Having individual, separated frequency ranges is what reduces the intermodulation distortion, and provides the main benefit of this technique.

 

Yes, I am a nerd :)

[SirNedwood]

On 2/20/2023 at 6:03 PM, craiganderton said:

If you want to check it out for yourself, split a guitar track into two buses. Insert the Helix in one of the buses, and enable the Crossover Split. Sum the split's outputs back to mono in the Helix. Set both buses to the same approximate volume. As you vary the crossover frequency, you'll hear significant, audible phase shifts going on between the Helix track and dry track.

 

...

 

The main problem I have with the Crossover Split is that it's not particularly useful for stereo because the highs are on one side, and the lows on the other. That doesn't mean it can't work, but using the dual graphic EQs gives a more even frequency distribution to the left and right channels. So, you can have wide stereo images, that also collapse well  to mono.

 

That's really interesting doing that phase check. Always better to confirm with a test than just assume. As for the stereo thing, I must have glossed over that part in the OP because I didn't think of this as a stereo system. But I'm definitely going to try that out now.

[craiganderton]

On 2/20/2023 at 2:32 AM, SirNedwood said:

As for the stereo thing, I must have glossed over that part in the OP because I didn't think of this as a stereo system. But I'm definitely going to try that out now.

 

Thi is a little off-topic, but I was steadfast against stereo for live performance for, well, decades. However, I eventually realized I was going about it wrong. The object was not to use stereo to create a distinct left/right image, but to act more like using two different amps/cabs powered by different effects close to each other, which added together into a single, wide sound. The bi-amp thing works well for that.

[SirNedwood]

On 2/22/2023 at 4:24 PM, craiganderton said:

 

Thi is a little off-topic, but I was steadfast against stereo for live performance for, well, decades. However, I eventually realized I was going about it wrong. The object was not to use stereo to create a distinct left/right image, but to act more like using two different amps/cabs powered by different effects close to each other, which added together into a single, wide sound. The bi-amp thing works well for that.

 

So I used to have all the cabs in my presets as a blend of 2 mics on the same cab with a slight 25L/25R split, and didn't really know what I was doing. But recently, I've gotten better at understanding how important mic and cab is to the sound. Far more so than the amp. So now I listen carefully to find the best combo of cab and mic to shape the tone, I select my 2 favourite, and then make a dual cab block that is 100L/100R with those settings. I thought it would sound disconnected, but even with full separation like with headphones it just gives a really full sound. So now I'm re-doing everything in full stereo, and I have a smaller selection of my favourite cab combos.

[ossianott]

Tried this briefly yesterday, definately something interesting happening. I play metal so this might come in quite handy, but I have a stomp so I will have to experiment with the dsp-starved amp simulations I think. Thanks craiganderton!

Edited February 23, 2023 by ossianott

[craiganderton]

On 2/23/2023 at 4:10 AM, ossianott said:

Tried this briefly yesterday, definately something interesting happening. I play metal so this might come in quite handy, but I have a stomp so I will have to experiment with the dsp-starved amp simulations I think. Thanks craiganderton!

 

You can always split the amps in parallel, but then mix them and follow the mixer with a single cab. The new cabs do draw a lot less DSP, but using one cab instead of two might be enough to let you add some extra effects. 

 

As to metal, there's a sweet spot between "clean" and "filthy" amps where this technique is most effective. But it definitely does improve high-gain amps. Have fun!

[jester700]

This is cool, and much more DSP friendly than the 4 band multiband presets I've played with. I usually used a Boss DA-2 Adaptive Distortion pedal when going this route, but now I can pick from several amps/pedals. Still like the Boss, though. Never too many flavors...

[SaschaFranck]

Not sure how much this actually "improves" things (I mean, I don't even know whether I need improvements in that area), but it's pretty interesting for sure. Unfortunately I could never put this to use live, which I would find to be interesting (even when I still had the Floor, my patches were maxed out in terms of options (not CPU), with the Stomp it's completely impossible).

Anyhow, as said before, at one point in time I will most likely purchase a Tonex pedal and this technique will be just another one I'm going to put to good use, so thanks!

[MikeDV1]

I went to try this, haven't quite made it, I went to a shorter version.  I play Helix rack into a solid state power amp into speakers.

 

Always tweaking/messing to dial out harshness at high gain with various methods, along with different speakers, different eq, etc. etc.

 

I went to try this and by accident went amp > high/low eq > delay.  I pulled the eq down to add the split path (on Hx Edit) and there it was!

 

Ended up with amp > branch - one signal continued straight through the other went through the eq, and blended them back together before hitting the delay.

 

Ended up "keeping" the "harshness on the straight signal, cutting the highs on the other one, when they blended - for me - perfect!  This was before touching the amount of each return, both at full - I changed the balance slightly, but more good, just slightly different.  I basically did the same thing Craig laid out above, with one amp model.  Going to put the dual eq (one on each path) as craig suggests and try that today, but by splitting AFTER the amp, the same model pushes the same signal and is joined together in what seems like the same way.  

 

It immediately sounded so good with ANY high gain model, I had a hard time continuing to try different models because I just wanted to play.  Rolled in some clean models, but found it unnecessary, always felt very good with clean models.  I also have a Triaxis I'll switch into the mix occasionally, did the same thing replacing the amp model with the Triaxis and same good controlled results.

 

Thanks Craig for leading me to a very happy accident!

[craiganderton]

On 2/24/2023 at 5:53 AM, SaschaFranck said:

Not sure how much this actually "improves" things (I mean, I don't even know whether I need improvements in that area), but it's pretty interesting for sure.

 

The amount of improvement depends on several factors. With a clean amp or single-note leads with any amp, you probably won't hear a significant improvement because there's no intermodulation distortion. With tonic+5th power chords, the improvement will be noticeable, but not earth-shaking. You'll hear the biggest improvement with amps that have considerable distortion, when playing chords with 7ths, 13ths, 9ths, etc. that push more frequencies into the amps.

 

The audio example I did in this later post shows what I think is a considerable improvement in terms of clarity and definition.

[jester700]

Sorry for the necro reply, but I was just thinking that if we ever get a 1/3 octave EQ effect (a 30 band EQ), we could make a 3 path split using this technique and only passing every third band for each path. That passes a fairly narrow chunk to each distortion effect, and the chunks repeat on the octaves. So, even less IM, and the IM that's there should be more musical, as it correlates more closely with the natural harmonic series. I THINK this should work better than a 3 band split that just crosses over to each distortion with a single passband as Craig originally did.

[themetallikid]

I tried inputing this method into my STP tribute presets where I use the Ventoux/Placatar dual amp sound.  While I liked what it was adding in clarity, it also sorta removed the 'thickness' I was getting from the Placatar.  However, I did not spend a ton of time trying to 'fix' it.  

 

I like what its doing overall, just might not apply to my use-case greatly.  

[craiganderton]

On 10/22/2023 at 6:39 AM, jester700 said:

Sorry for the necro reply, but I was just thinking that if we ever get a 1/3 octave EQ effect (a 30 band EQ), we could make a 3 path split using this technique and only passing every third band for each path. That passes a fairly narrow chunk to each distortion effect, and the chunks repeat on the octaves. So, even less IM, and the IM that's there should be more musical, as it correlates more closely with the natural harmonic series. I THINK this should work better than a 3 band split that just crosses over to each distortion with a single passband as Craig originally did.

 

I've done that with Waves' StudioRack and placing their GEQ in each of three splits. It sounds pretty great, if...

 

On 10/23/2023 at 11:42 AM, themetallikid said:

I like what its doing overall, just might not apply to my use-case greatly.  

 

...and only if you want "clean" distortion (yes, I know that's an oxymoron!). It's a totally different type of sound compared to physical amps. I consider the technique a complement to traditional amp sounds, not a replacement. I use it a lot in synth-driven music, as the guitar acquires a sort of clarity that fits better in a mix with synths and electronic drums.

 

This is what I love about Helix - you can find something appropriate for just about any form of music. Although admittedly, I haven't had much luck with Gregorian chants :)