Hi cuts on IR's/cabs in modellers - I'm confused?

Beexter · July 2

It's quite common to see recommendations for using hi cuts on IR's/ cabs when playing a modeller through an FRFR speaker but.......

....if an IR is a "capture" of a specific mic, in a specific position on a specific speaker in a specific cab, in a specific room, why should it be necessary?

Surely if a particular guitar speaker doesn't generate much over say 8500 hz, surely the mic can't capture a frequency that isn't there in the first place and therefore it wouldn't be present in the IR? If a real world speaker was too bright, you typically adjust your tone controls on your amp/guitar/pedals etc, or change your speaker..

I mostly use York Audio IR's in both my Fractal and HX Stomp and I'm sure I read somewhere that Justin from York Audio suggested NOT using hi cuts.

I'm just trying to understand if I'm missing something fundamental, so curious to ask the Hive mind...

Musicwolf · July 2

I'm getting on a bit, so the upper limit of my hearing is going to be well below someone in say their 20s. For that reason I tend to set the filter at just above the point where I can't hear it doing anything (assuming that I'm not trying to change the sound). That way I know that I'm hearing what everyone else hears.

Bezzer · July 2

It's because people want the sound of the cabinet/speakers and not the sound of the microphone being used to pick up those speakers so try to EQ it out of the equation. Missing the point of modellers in my (admittedly extremely grumpy through stress today) opinion

Fiftyshadesofjay · July 2

What @Bezzer said.

If it’s okay to stick a 57 in front of a speaker and get great results with minimal tweaking, you generally don’t need extreme high or low cuts on a modeller.

I think it’s normally people without much experience with mic’d up tones trying to make a FRFR sound like a guitar cab.

Nerine · July 3

Yeah, people go extreme with this shit because YouTubers have banged on about it ad nauseam for ages without really thinking about it.

It’s led to to a lot of people having very thin, boxy sounds that have no appreciable weight and air because it’s all been filtered out because they were told to…

Lewy · July 3

I've wondered about this myself. Does something about the modelling process add something that isn't there in real life? How much "frog DNA" do they have to use to make it all work in the first place?

I've concluded that it doesn't matter. People are very probably wrong about why the things they do work for them, but if they work then happy days.

High and Low cuts on cab blocks....at the end of the day it's just EQ isn't it. Do we ever really hear the raw sound of a SM57 on a speaker without any EQ, either live or on record? And of the sounds we do hear live or on record, would they be pleasing to the ear on their own for practice? The cab block is just a convenient place to do it for me.

Nerine · July 3

Lewy said:

I've wondered about this myself. Does something about the modelling process add something that isn't there in real life? How much "frog DNA" do they have to use to make it all work in the first place?

I've concluded that it doesn't matter. People are very probably wrong about why the things they do work for them, but if they work then happy days.

High and Low cuts on cab blocks....at the end of the day it's just EQ isn't it. Do we ever really hear the raw sound of a SM57 on a speaker without any EQ, either live or on record? And of the sounds we do hear live or on record, would they be pleasing to the ear on their own for practice? The cab block is just a convenient place to do it for me.

I do think there’s place for HPF/LPF on modelling stuff, but I’ve heard people go as drastic as frequency limiting it from 100Hz to about 3kHz.

That’s stupid.

Roland · July 3

Beexter said:

It's quite common to see recommendations for using hi cuts on IR's/ cabs when playing a modeller through an FRFR speaker but.......

....if an IR is a "capture" of a specific mic, in a specific position on a specific speaker in a specific cab, in a specific room, why should it be necessary?

Surely if a particular guitar speaker doesn't generate much over say 8500 hz, surely the mic can't capture a frequency that isn't there in the first place and therefore it wouldn't be present in the IR? ...

A guitar speaker does generate higher frequencies, but you are less likely to hear them. Sound dissipates with distance. Higher frequencies are absorbed more by the air they move through, and other obstacles in their path. The higher the frequency the less energy it has to start with. An example of this is that the sound stage front at Download is reasonable balanced, but a mile or so away all you can hear is a low grumble.

When sound is reflected the reflection and original sound will be out of phase. The higher the frequency the more likely it is that some degree of cancellation will take place.

IRs are generally close mic’ed to avoid picking up reflected sounds from the room. Lack of absorption and cancellation mean that close mic’ing is going to give a brighter sound than what you’d hear a metre or so away.

To make an IR generated sound through an FRFR speaker sound more like an amp in the room you need to cut higher frequencies. Ideally this would not be a steep 18dB per octave cut, rather it should be progressive, but steep cuts are easy to talk about in YouTube videos. The other thing you need for amp-in-the-room is early reflections from walls, floor, and ceiling.

In the Fractal Cab block you can introduce room effect, and it’s interesting to hear how this affects the frequencies which your ear interprets. To hear an example listen to this video just after 20:00:

https://www.youtube.com/live/7YJMVwcF4SQ

Roland · July 3

Beexter said:

… I'm sure I read somewhere that Justin from York Audio suggested NOT using hi cuts....

You’ll have to ask him about his own reasons. My take on this is that the very high frequencies are necessary to the quality of the sound, albeit at reduced volumes compared with other frequencies. A hard cut will exclude them, degrading the sound.

CD sample rate is 44kHz. DVDs are 48kHz. The highest frequency you can record on a CD is 22kHz, and that is an on-off square wave. Since most people can’t hear above 16kHz, and many of us considerably lower, it was reasoned that 22kHz was a good upper limit for CDs. It turns out to have been a wrong assumption. Our ears don’t use Fourier analysis to separate out the frequencies in a sound. So even if you can’t hear a pure sound wave at 18kHz your ear and brain know if the frequency is missing. Age and loud music mean that my hearing is 9dB down above 4kHz, but I still know if a guitar sound has a hard shelf at 5kHz.

So I agree with York Audio. Don’t hard cut at 5kHz, or other any frequency. Instead use early reflections and a gradual EQ slope to bring an IR’s higher frequencies down to a more natural profile. As a band it’s worth doing this collectively. What really pisses off my ears is if each musician has applied separate reverb algorithms so that they sound as though they’re playing in different rooms. I don’t mean reverb as an effect, such as a Strymon might give, but those early reflections.

Beexter · July 3

Thanks for all the responses. I looked up the frequency response of a Celestion Vintage 30 and was a bit surprised.. whilst it's frequency response drops off a cliff at 5,000hz, it is still pushing out measurable frequencies up to 20,000hz, so whilst you may not specifically hear them, they are contributing to the overall sound as @Roland commented.
Therefore, by cutting anything over say, 8,000hz, you are fundamentally changing the core tone of the speaker.
I get that we don't listen to a guitar cab on full tilt with our ear right on the speaker (which is essentially what a mic is doing) so the sound we actually hear in a particular space is influenced heavily by the environment in which we're hearing it.
In the "real world" if you are playing an un-miced cab, you adjust "your" tone via amp and guitar and if that doesn't get you what you want, you might consider a different cab/ speaker.
Of course, when it comes to micing a speaker, that brings a shed load of other factors into play.
If , however, you have spent a long time honing "your" sound (likely dialled in directly and not through a miced cab) wouldn't you want it to be replicated as closely as possible?

Why try polishing a turd when you could choose a different speaker like you would outside of the digital domain?

A modeller provides a myriad of ways to "EQ" the sound yet the hi/lo cut on the "cab" approach seems the most often mentioned and I'm still no closer to understanding why that is so.

I really should get out more....

Lewy · July 3

On this video, Rhett Shull and Rick Beato compare a real Marshall (mic'd up and through EQ) with an HX stomp (through the same EQ) and there's an interesting bit at 12m25s where they talk about removing the high cut he usually uses on the Stomp to make it sound more like the amp.

mrkb · July 3

In the Fractals you aren’t applying a hard cut, it rolls off at a user selectable slope ie 6 /12 etc dB per octave. So if you have a 6db roll off Starting at 5kHz, that’s 6db at 10kHz and 12dB at 20kHz - so it’s not that severe. Not sure what slopes other modellers use.

Bidley · July 5

mrkb said:

In the Fractals you aren’t applying a hard cut, it rolls off at a user selectable slope ie 6 /12 etc dB per octave. So if you have a 6db roll off Starting at 5kHz, that’s 6db at 10kHz and 12dB at 20kHz - so it’s not that severe. Not sure what slopes other modellers use.

I don't know the exact slopes, but HX is similar. The cuts in the cab block are less severe than the hi/lo cut eq block.

PolarityMan · July 5

I feel like helix in particular just has a bit of nasty shit up there that you need to get rid. Having said that for high gain it's not uncommon for me to apply a roll off even if im working entirely in the box jsut to tame the fizz a bit sometimes.

crunchman · July 5

mrkb said:

In the Fractals you aren’t applying a hard cut, it rolls off at a user selectable slope ie 6 /12 etc dB per octave. So if you have a 6db roll off Starting at 5kHz, that’s 6db at 10kHz and 12dB at 20kHz - so it’s not that severe. Not sure what slopes other modellers use.

QC is similar. I've got 6dB per octave on most of my presets, but you can select 12dB, and possibly 18dB, but I've never gone that high.

The QC does seem to need taming like that. I don't do that with my Kemper. My Kemper is a toaster from before they released the stage version, so the QC is better for live. Being able to have 2 completely separate signal paths is good live as well. I have an ES Les Paul that I've had a Fishman Power Bridge fitted to. There is a stereo output jack, and I can use a Y cable to make use of the separate signal paths - one for the conventional pickups and one for the piezo signal. I'm experimenting with acoustic IRs for the piezo signal.

Howdy, Stranger!

Categories

In this Discussion

Become a Subscriber!

Hi cuts on IR's/cabs in modellers - I'm confused?

Comments