Controlling high frequency of heavily distorted guitar tone (metal)

Question

I'm constantly experiment with metal guitar tone for years. If I get the amount of distortion/clarity from the ampsimulator, the next stage is much harder: controlling the high frequencies; handling low frequencies seemed easy to me. Problems are the following:

1. I've created a decent tone, that sounds great, but it has some annoying high frequencies (almost like a static hissing starting from about 6 kHz)
2. After a load of cutting (details below), I can either choose a tone that sounds odd due to much of cutting, or a sound that is dull (cutting brutally kills upper frequencies), or leave it mainly untouched which is too harsh and noisy.

This is how I work currently with Waves REQ:

1. I cut out gently some low and high frequencies (lowpass: about 100 Hz, highpass: about 9 kHz)
2. Cutting out some frequencies near 2 kHz (I boost a narrow frequency, then I am searcing where it sounds the ugliest - great tip from a guy from youtube)
3. Cutting out some 200-300 Hz if it has a strong "inside a bin"-like sound
4. Probably I boost around 400 Hz a bit for making the fundamental tones a bit more audible (great tip from Colin Richardson)
5. Then the problems arise: I am left with fiddling with the highpass, leaving me with dull or still a harsh, but less annoying sound. Or I am cutting more frequency in the high region (above 4 kHz) and I am left with some odd, unnatural sound (especially compared to the original, which was too harsh, but natural).

What kind of strategy you use for controlling the sound of a heavily distorted guitar?

How do you "debug" the sound?

(Later this day I can upload some stuff if you want to help fix me the stuff, though I'm interested in your setups/strategies).

Some details about me: I am recording guitar at home through a Line6 UX1 for many years now. I'm recording both a dry and a decent wet tone from the card. I am feeding the former through an amp simulator (like Amplitube 3).

Update Since its a broader question and many of you had great tips, I suggest you turn one of the answers to a community wiki and therefore I can accept it as an answer.

Answer 1

Ok some insight I got from a high-end Mixing Engineer from one of the best recording studios in Germany:

1. With distorted guitars: First get the sound right, then thing about eqs and stuff THEN:
2. Be careful with the high pass: Some Guitars have needed power in the 70-90 hz area (punch, impact). It is often useful to boost this area to give the guitars more impact. highpass at 50-60 hz should be enough to get rid of subsonic mud.
3. Mud/Body area everything from 150-700 hz is a critical area. you really need to balance the frequencies dependent on the other instruments and the role of the guitar.
4. search some good frequencies between 0.8 and 2 kHz and boost if needed to define the guitar in a mix (especially when multiple guitars are in use)
5. 3-4 khz is a very very important are too look out for. Usually u find a ear hurting whistling around there. Notch or cut it a little bit to get the unwanted aggressiveness out of a distorted guitar.

6. 8-10 kHz controlls brightness but collides with hh, snare and "esses". I would cut the area since the hh should dominate this area.

7. Automate Eqs or create multiple eq setting fore different regions depending o nthe role of the guitar in different song positions. If the guitar is just used as a backing rhythm you might eq it quit hard (muddy area). But at one point there is this cool guitar riff and the singer/ other instruments stop. So you might want to give the guitar more low mids and highs to give it some body and character.

8. Eq the guitar in the mix (all other instruments playing) and without visual reproduction of the frequency spectrum. solo and mute other instruments and see if the distortion is hiding a lot of needed frequencies ...
9. Be sure to use panning to separate different guitars
10. use compression but don't overdo it. compression should flatten out unwanted peaks. but if you overdo it you will get a wall of noise with no place for other instruments.
11. Volume automation is important to bring the guitar to the foreground in solos/fill ins etc.

To sum it up: Think of the role of the guitar -> then use the right amp/cabinet/distortionfx -> then use eq and compression to get rid of unwanted frequencies

gl

PS: oh and don't do a generally lp at 9 kHz you will use a lot of open-ness to the sound

Answer 2

There's a point of paramount importance the answers so far haven't even mentioned: as with anything nonlinear (and metal-guitar distortion is the most nonlinear you get in audio production) it makes very much a difference whether you EQ before the distortion, or – with exactly the same settings – after the amp plugin. That's why you can't just simulate different amps by adding even the most cunning FIR post-EQing.

Indeed a clever combination of EQing before, after, and in between distortion stages is pretty much what makes up the whole character of an amp. (There's also yet more complicated effects that have to do with power supply drop etc., but those are secondary and can reasonably be approximated with compressors.)

Properly designed distortion/amp plugins, such as iZotope Trash, take care for this and (instead of offering a ton of complete amp simulations that may sound close to the analogue originals, but offer no more tweaking possibilities and none of the "real-world-feedback" benefits) focus on giving you a useful set of primitives: essentially just clipping functions and saturating filters. This gives you control over all the crucial parameters.

Now, to the particular problem you describe. This may actually be a technical problem you shouldn't tackle with EQ at all, but grab right at the root.

• Actually is sounds a lot like you've run into aliasing, which is a particular nasty digital artifact distortion plugins are rather susceptible to. The correct way to deal with it is to increase the oversampling rate, an option any plugin worth its salt should offer. Try that first!

• As you say "almost like static", you should of course make sure it is not, in fact, static! USB audio interfaces have a tendency to pick up quite a lot of nasty switching–power-supply bursts, and even if those aren't obvious in the clean signal they will get dramatically boosted by distortion (as distortion also act as a kind of compression). To adress this, many plugins have noise-gates built in, which unfortunately makes the problem again harder to localise.

If neither of these fixed it then it's in fact a sound problem.

Now, a thing with high-distortion sounds is that the treble frequencies originate predominantly from the distortion itself, not from the guitar. So what I'd try first is, see what happens when cut the > 6 kHz range pre distortion entirely! It's quite possible that this won't change the sound much at all except for slightly duller attack character. But OTOH, the thing that annoys you might in fact be gone, because the guitar already has a strong component up there that interferes with the distortion's harmonics. This can especially be an issue if the guitar has strong inharmonicity, typical of short scale lengths (Les Paul etc.) or when the pickups have very strong magnets and are too close to the strings. Again, points you should cover before starting to fiddle with any post-EQing.

If you can't get rid of those annoying frequencies by any of the above means, you'll have to use some EQ though. The procedure you've used at point 2. (boost narrow -> search problem -> cut) is correct, but it's typically not very useful with distorted sounds because, if there's a narrow band of problem frequencies in the original signal, the distortion will smear that problem out over the entire spectrum. Therefore, you should try this technique before the amp plugin. Perhaps 2 kHz is actually the problem, yet the distortion projects this upwards and therefore the result sounds bad mostly above 6 kHz. Cutting a small band before the distortion has the benefit that the result will never sound as "hollow" or "tinny" as if you cut afterwards, because the distortion adds back some frequencies into that gap.

If none of this helps, then it's simply the amp plugin's fault. Try a different amp model, or a different simulator plugin.

Answer 3

You don't "debug" the sound. You adjust it prior to recording so that you don't have to adjust or it's clear what kind of small adjustments will need to be made. You can't polish a turd.

Use the amp's EQ (and other) adjustments first and then adjust what's left after recording, if you can't get the sound just right just by using the amp's controls (and mic placement, if recording for real).

To me it sounds that you're equalizing a bit too much, which might signal that the initial sound/tone is not that good or not what you're looking for.

Answer 4

are you using external IR's (impulse responses) or are you using the ones within the sim? Often the sim ones especially in guitar rig I have found are not as good due to lack of available positions and mic types as third party ones such as Redwires.

My method is to make sure the di signal to the computer is clean as a whistle first without any amp in the sim. If it's not a noise gate or deesser can help.

Then I set up my amps how I want them to sound usually based on a reference from an existing artist. Then I remove any cabs from the amp sim and in the plugin in chain add Redwires irc2 I then mix a bunch of different mics and cabs often again based on an existing artists recording method or live setup. Redwires have impedance curves and other subtleties which makes a difference.

Then I will usually add a multi and compressor if I feel it requires it and apply shelving eq to both end to make the guitars sit better with the bass etc.

I will always multi track my guitars at least 2 tracks often more and usually each track has either subtle amp setup, mic setup or eq changes. Sometimes these changes are drastic so as to be like totally differnt guitars depends on what I am going for.

I will also sometimes apply a tape saturation or similar plugin to emulate some analogue warmth.

Failing all this I go to my brothers house and retrack the guitars using his axefx2

Hope this helps, good luck.

Answer 5

A harsh distorted sound is often caused by harmonics which are too dense. Applying a lowpass removes these harmonics making the sound softer, but this throws the baby out with the bath water,

The distorion adds harmonics to the signal and the higher harmonics get very close to each other, forming non-musical intervals.

You can mitigate this effect by applying an impulse-response. The frequency response of (most) IRs is a series of peaks and valleys, which also get denser at higher frequencies. Thus an IR "combs" out many of the high frequencies making them less dense and less harsh.

In the real world, i.e. using physical amps, speakers and rooms there is quite a lot of IR between the distortion and your ear, so you typically have less of a problem with harsh distortion.

To apply an IR electronically you can try to use some "erly reflection" effect, or even better a sampled IR from a real amp in a real room.

Answer 6

Maybe try multi-band compression or depending on the frequency, a de-esser. You'll need to make sure that it's not affecting certain notes much more than others but it can be more transparent than an eq and you can set your attack and release to clamp down hard or to let a bit of those frequencies through before it starts working.

Answer 7

I put my experience on the matter on this lil print. 1.Guitar can be humbucker and even dull one.Not fully dull but half dull like Angus Young style 7500k pickup which I have in my SG. That produces very solid distortion when put through signal chain. 2.Pre Equing is most important practical tool and the point where you need to invest thought. But you can get the real kickstart by being aware about Blue Cats free triple eq. It has all those tree requirements to create basic pre-EQ for guitar distortion: a)HP ie. low cut b)1 parametric zone with bandwidth c)LP ie. high cut. Use it to scan correct hump and low and high cut for the pickup and subsequent distortion.

Shevliaskovic and leftaroundabout with joint answer are so right about the philosophy of distortion.I belong in to that party if somewhere. One can learn from real amps and there should be proper tools to use extracted primitives. Distortion is such magical great thing on its own and should be considered and researched as such. Forget the amp sims and use you own created signal chain. But Izotope Trash cannot be recommended because demo version is so much crippeled that you cannot test if real one has proper most important component which is as mentioned pre equing the distortion.

Answer 8

There are many ways but they are not trivial. Some of the IR libraries are "mix-ready", meaning someone else figured out how to phase out the harshness out of the tone. After all these years it seems the trick is in cabinet selection + mic placement, but that is a very hard topic; not for the everyday guitar player.

Personally, I started buying some good quality "mix-ready" IR libraries and then do some tricks:

1. Select a few "favorite" tones - most likely the loudness (the peaks) of these won't match, therefore...
2. I save them as FX presets/FX chains with the peaks adjusted to be the same (same peaks and the guitar has no presence => you have a weak IR).
3. Now I set up a rough mix for the song I work with, a one with guitars quite present.
4. I start swapping my favorites; now that they are normalized, I can hear what they do in the context of the mix.
5. Make final adjustments with EQ, saturation (in extreme cases Softube's Dirty Tape did wonders in rolling off the nasty high frequencies).

Bonus tip: if you are working with bass (especially distorted bass) you can further refine the tone if you select the bass IRs while you have the guitars and a bit of the mix running. Once I have even turned off bass cabinet IR as the raw distortion complimented so well the sophisticated miced guitars.

For more in-depth explanations, I have the selection of these videos:

Using phase of two microphones to filter out nasty frequencies:

Answer 9

try vandal from magix with redwirez impulses and mixIR 2. i apply a LP around 5.4khz -12db with the redwirez impulses because they sound very fizzy for me also vandal cabinet simulation is very good

Answer 10

What you really need is a 12" guitar speaker which will automatically roll off those frequencies for you. A tube amp would also be ideal, but really the lack of a speaker is a bigger deal. Whatever mic you sing through is fine stick it right on the grill move it towards the center for a thicker sound, towards the edge of the cone for a brighter sound with more attack.

If the issue is neighbors, you need to get a low spl (read:quieter) speaker, which is going to be pretty cheap (under $50). You want something like 90 or 92 spl, light magnet is fine because you can't turn it up that loud anyway. Cannibalize the couch and build a pillow fort. Make sure you rehearse your parts so you can nail them quickly.

If the issue is not neighbors, here's a little secret: Amp simulators and even worse cabinet simulators, do not even vaguely sound or play anything like a real amp or speaker cabinet. Save up for the real thing.

If direct is the only option for whatever reason, first run your guitar through a Digitech Death Metal Distortion pedal ($50 new) and use the top left output. The "speaker simulator" on that output is probably a capacitor and a resistor, but it sounds infinitely better than all the high plug ins.

Answer 11

like to hear where your at now with it. here is what my advise is

find the cabinet type you like (1x10,2x10,4x10,1x12,2x12,4x12,ported open close etc.) and speaker size and tone that you like (each speaker breakup diff and tone is diff) next find the microphone you like (sm57 boosts lot of high makes it harsh and scooped, u87 is full bodied etc.) next find out what time of mic preamp you like( to amplify the mic and add coloration ex. neve api, ssl ,etc. next finf the type of recording console you like (simulated or actual there is never api ssl brit american etc. many many options next find the type of tape you like to print the sound on (either virtual or real tape recorder)

you see all these things are critical and unique to add up to be the whole tone after all thisis known (your preferences) and you set it all up (real or virtual) all you will need to do after in any circumstance to change toneis add a eq (real or virtual) ex. neve 1073 or ssl or whataver with analog sim and finally a light compressor either real or virtual like tube pultec etc. now once you have all these in order (either real or virtual) you will have the ability to adjust fine details to get ANY pert of the tone perfect for your tase as well as have all that detail of analog steps nonlinear (real or virtual) these is how to eliminate artificial notch filtering and all that mess which ends up with artificial fake sound so to review

1. amp
2. cab type and speakers
3. mic or mic's 4.mic preamp 5.analog console 6.analog tape

then if you NEED to adjust more 7. tube eq or analog eq 8. tube compresser or analog compresser with 1:5:1 attack of 100 ms release of 300 ms to start off then tweak.

whether vst or actual gear this is the chain and adjustments to get any guitar or bass track or extended range perfect no matter what the goal is to set each stage up right then at end go in fine tune to get PERFECT final tone that the accumulation of all that non-linearity analog sweetness smooths anything out no matter if you want boosted high or low and warm the final frequency response of lows mids and highs will be smooth no matter their volume level no matter what specific details you choose for each step.