Need help designing a cab simulator

Started by Fancy Lime, April 30, 2021, 02:42:31 AM

Previous topic - Next topic

Fancy Lime

Hi all,

I want to design a home studio pedal platform for bedroom recording. Nothing too fancy, basically an amp + miked cab simulator but for guitar as well as bass use. I am having trouble striking a reasonable balance between flexibility and complexity.

For the amp part, I have almost settled on a Baxandall tone stack plus two Wien-Bridge semi parametric mid bands. Any distortion will be handled by pedals up front.

My problem is that I am not quite sure what I want from the cab sim part, especially in terms of necessary flexibility and complexity considering that there is going to be quite a flexible eq right in front of it. I will definitely need a tunable low pass. The obvious solution would be a Sallen-Key with a stereo pot plus maybe a second static SK. Maybe some cap switching option to vary the Q of the adjustable SK.

Then there is the question of high pass. How necessary is that really assuming subsonic frequencies are already filtered out by the coupling caps? And the resonant hump around the 100Hz mark that some cabs have. Necessary to implement, given that there is going to be a bass control in the EQ? Do I want to have some kind of Shape control for the mid dip in addition to two semi parametric mid bands in the EQ?

I guess my main question is: does anyone a good collection of frequency responses of different guitar and bass cabs with different miking options? Or an "impulse response" library with graphs for reference?

I think I know most of the great DIY cab sim designs out there but mine is supposed to work in synergy with the specific EQ instead of being "general purpose" or optimized for the typical mid-cut of most guitar amp EQs.

Thanks,
Andy
My dry, sweaty foot had become the source of one of the most disturbing cases of chemical-based crime within my home country.

A cider a day keeps the lobster away, bucko!

FiveseveN

Far be it from me to discourage building something, but I've yet to find an analog cabsim that comes close to its intended function.
Reason being those frequency responses are full of details (peaks and valleys) that matter and you can't fake with three filters, or two as you're planning. I'm only noting because I imagine "bedroom recording" involves a PC, i.e. something that can compute convolution.
Quote from: R.G. on July 31, 2018, 10:34:30 PMDoes the circuit sound better when oriented to magnetic north under a pyramid?

merlinb

#2
I'm sure there are a few websites showing multiple guitar cab response plots because I remember using them to design my own cabsim, although I don't have the links anymore. I'm sure you can find them with a bit of searching. FWIW my own cab sim is a one-knob design:
http://valvewizard.co.uk/cabsim.html

FiveseveN

Quote from: Fancy Lime on April 30, 2021, 02:42:31 AM
does anyone a good collection of frequency responses

Celestion has graphs on their product pages: https://celestion.com/product/vintage-30/
Then there's this thread on TGP that compares some IRs: https://www.thegearpage.net/board/index.php?threads/ir-cabinet-simulations-eq-shapes.1416811/
Compare to this analysis of analog cabsims: https://www.hexefx.com/diy/tech/cabsims
Though there's such a variety of cab tones, it would be hard to build a comprehensive reference, and then if you squint hard enough you can find some common features but lose the details that make a tone identifiable. I'd say find the cab(sim) you like first, characterize it yourself and go from there.
Quote from: R.G. on July 31, 2018, 10:34:30 PMDoes the circuit sound better when oriented to magnetic north under a pyramid?

Fancy Lime

Quote from: FiveseveN on April 30, 2021, 03:26:58 AM
Far be it from me to discourage building something, but I've yet to find an analog cabsim that comes close to its intended function.
Reason being those frequency responses are full of details (peaks and valleys) that matter and you can't fake with three filters, or two as you're planning. I'm only noting because I imagine "bedroom recording" involves a PC, i.e. something that can compute convolution.
Hi! Yes, IR plugins are an alternative I am considering. However, I like the idea of having a plug and play solution that works even without a computer for practicing. The details of real speaker response are certainly not attainable with analog means but the question is, do we need them for a "good" sound? I am not trying to simulate the actual sound of a real cab so much as getting it to sound "equally good" (whatever that means, subjectively). So, in a way, "cab sim" is not quite the right word for what I want. More like, "sound shaper that fills a similar role in the signal chain that we have gotten used to expect from cabs, so as to be able to use the other gear we normally use in conjunction with real cabs but without cabs, all the while not making it sound like ass". If that makes sense.

Also, thanks for the links, those are exactly what I was looking for. My google-fu was not strong enough. The graphs at hexe in particular give some great pointers as to what corner frequencies should be good starting values for experimentation with the filters.

Also thanks to Merlin, although I was well aware of your cab sim. Indeed, your article nudged me into trying my hand at an analog solution instead of lazily buying a commercial cab sim pedal.

Cheers and thanks,
Andy
My dry, sweaty foot had become the source of one of the most disturbing cases of chemical-based crime within my home country.

A cider a day keeps the lobster away, bucko!

temol

Quote from: Fancy Lime
I guess my main question is: does anyone a good collection of frequency responses of different guitar and bass cabs with different miking options? Or an "impulse response" library with graphs for reference?

https://docspike.com/download/axe-ii-graphs-revb_pdf

No miking options available but nice collection of frequency response graphs.


iainpunk

just some food for thought:
there was a time where i wanted to do the same as you, just a amp sim that plugs in to my HiFi system at home. i asked a teacher for some tips and pointers about speaker behavior, and he loaned me a book. about half of the book was devoted to amplitude dependent filtering, which kind of translates to soft clipping, with a filter instead of a fixed ratio of softness.
this non-linear behavior is something missing in 99% of speaker sim's tho, and i doubt it will make that much of a difference in a full mix, since they were quoting THD percentages of about 2% to 5% when staying within 70.7% of max rated power, given a front loaded closed back and bass-ported cabinet. i suspect open back cabinets having more distortion due to the higher excursion of the speaker cone for the same SPL.

cheers, Iain
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

temol

#7
If you're ok with a "neighbour" forum then you can also check this cabsim thread


garcho

Part of making something sound "real" i.e. a cab sim for guitar going direct, is all the phase and time dynamics that happen in the 4D physical world. Sound beating out of a paper cone in a wood box is complicated enough, put a mic, pre, tape/ADAC/mixer etc. and put all that in a space with a shape and walls with absorption/reflection  and it's basically impossible in the analog realm.
That being said, if you make an EQ, say something along the lines of Merlin's cab sim, followed by short "room" reverb that has some predelay control and whathaveyou, you'll be just fine. I've been an audio engineer for a long time, I've squeezed out decent fake real sounds with EQ and room verb that have fooled experienced musicians.

Spacetime is like, real, man.
  • SUPPORTER
"...and weird on top!"

Ripthorn

Quote from: garcho on April 30, 2021, 09:41:33 AM
Spacetime is like, real, man.

As an acoustic physicist, I can confirm that this statement is, in fact, correct.
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

deadastronaut

when i was tinkering with ''cab sims'' for months , i honestly breaded up around 20 of them...

some with graphs some without....what did i learn?, dont trust graphs.

trust your ears. i learned that the very very long way.

and ended up with the 'astro sim'. (which i use every day)

i see merlin has employed the set of 'T FILTERS '' which is what i went for too...nice and tweakable 8)

demo.

https://www.youtube.com/watch?v=fc6ilaVc9ho

https://www.youtube.com/user/100roberthenry
https://deadastronaut.wixsite.com/effects

chasm reverb/tremshifter/faze filter/abductor II delay/timestream reverb/dreamtime delay/skinwalker hi gain dist/black triangle OD/ nano drums/space patrol fuzz//

Ripthorn

I can confirm that the Astro cabsim is pretty great. I've tried a few others and prefer the Astro.
Exact science is not an exact science - Nikola Tesla in The Prestige
https://scientificguitarist.wixsite.com/home

Vivek


Vivek

Here is the SPICE graph of the Rockman X100 Cab Sim


Vivek

Quote from: merlinb on April 30, 2021, 03:43:16 AM
I'm sure there are a few websites showing multiple guitar cab response plots because I remember using them to design my own cabsim, although I don't have the links anymore. I'm sure you can find them with a bit of searching. FWIW my own cab sim is a one-knob design:
http://valvewizard.co.uk/cabsim.html


Which actual Cab did you target ?

iainpunk

#15
reading up on all this stuff again makes me want to build one of them loadbox/DI/cabsim units.
i really like the idea of variable scoop and a tilt control on the cab sim, with pseudo-resonant filters.
i'd also like to experiment with comb filters, are there any simple ones besides using a delay chip?

cheers, Iain

edit: probably going to opt for a simple one knob design with internal trimmers for the fine tuning.
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers