cab sim mod question -- Lart's "simple cab sim"

[pee-j]

hi,

I'm very new to cab sims, I mean the entire genre... usage, what to expect, etc...
I'm building my first one, Lart's simple cab sim -- see yeraym's post in "Any cab simulator circuit worth building?"

but before I build it, I tried some modifications in LTSpice (which is a new experience for me, too)
and using 3906/3904 pairs, I came to a finding that changing C7 and C9 to 10nF, 3.9nF, respectively,
will produce a nicer "curve" that goes steeply down after 4kHz,
and doesn't produce a high bump-up in the same region (around 2-3-4kHz)...

is this a good idea?

I mean, I can change 2 caps, back and forth, no problem, but I don't have the basic experience of using cab sims...
any maybe what I'll like is something like the "charming fruity taste in wine", which shouldn't be there at all, other then the grape's taste, according to Bordeaux farmers

#2
of the the 3 curves published by yeraym, the one with 3906/3904 transistors appears to me as better than the other two...
cause it goes down sooner around 4kHz..
but again, what "better for me" means might be anything from very bad to totally wrong

#3
would you even mod this cab sim?

thanks in advance,

Peter


this is from yeraym's post:


the LTSpice file is the work of Yeraym (vulcanofx)
in fact if you were to download it, you should go there, or click this link  cause otherwise it seems I can't attach it...

[deadastronaut]

with cab sims i would recommend going by what your ears like.

breadboard it,  plugin an overdrive ,  out to your mixer/sound card,  then tweak....

i wouldnt get hung up on graphs either... 

[GibsonGM]

Yes, I think you might want to mod it by ear on the breadboard!  If you have a steep rolloff so low, 4kHz, you might lose some 'sparkle'...admittedly, much above that area is just harsh and not desirable, but depending on your guitar and amp, and quality of overdrive/distortion you prefer, you might need some adjustment to avoid losing TOO much above 4kHz.    4-5k can be a sensitive area for guitar 

[pee-j]

thank you both,

very much!
the graph hangup advice is extremely liberating, especially from the creator of astrosim!

and  it is likewise a great introduction to this field to be told about the sensitive nature of the 4-5kHz range..., GibsonGM

[GibsonGM]

Do you have a DAW, Pee-J?  A way to record your guitar and affect the sound (even audacity)?   It can be VERY educational to record some guitar (either with your own rig, or VST amp sim) and mess around with an EQ.  All can be had for free!

With guitar, we hear that there is not a lot of desirable information MUCH above say 5-7kHz, just shrill noise, harmonics of distortion that are very unpleasant in places.  So in the studio, people tend to start rolling the guitar off below 10kHz (AND they roll off bass quite a bit; more on that another time).  Might be at 7 or 8k, remembering that the slope of most EQs isn't so sharp that it cuts off immediately (so you still have a lot of 5k content).    Where one places this cutoff frequency is selective, and changes based on what guitar, amp, amp sim, mic, pedal etc.  You may be surprised at how little info you need that is up at 10k and above!!   There are areas engineers go looking at (esp. with amp sims), 3k, 5k, that are known for nasty shrill - tho you won't dig into that TOO much with a discrete cab sim, that would be too picky!  

I think it's worth playing around with in a DAW to really hear what's going on, AND you could input your experimental cab sim into it, and mess with THAT too, changing things with the EQ first, then trying to build them into your sim circuit.    Just a thought    

[FSFX]

You may be surprised at how little info you need that is up at 10k and above!!

I think that is generally true for most of us working with guitars and pedals and in the case of things like cab sims. However I remember listening to an interview with the late and great Rupert Neve who claimed that in his designs, having a response beyond audio was important as those components did influence the final sound. Presumably it is the contribution they make to producing intermodulation products which are within the audible range when they encounter non-linear systems such as tube amps.   

[GibsonGM]

Perhaps trying low-passing guitar at 7 or 10kHz or to taste, and comparing to not LP'ing at all, might lead one to their own conclusion (?)

I get scared by 'things causing IMD' in my amps or other sound systems, LOL.  Harmonics notwithstanding, but which are within the audible range, and decrease in power remarkably as you go up the freq. scale.  YMMV 

[FSFX]

I get scared by 'things causing IMD'

I think we all get a bit scared of things causing IMD but that is exactly what diode clippers do in fuzz pedals to produce the 'fuzz' sound.
If they just produced harmonics or clipped the waveform in a way that someone once described it as 'giving the signal a hair cut' then the resultant sound would not be so rich in content.

[GibsonGM]

That's very true..but if it was coming from 'super audible' frequencies, I'd imagine that there would be little or no control over what's going on.  I don't notice anything 'missing' or lack of 'airiness' by putting a cap to gnd after clippers.  Anyway, the energy present in such products is probably very small.    I'd just mess with the idea in several ways (as we all have, accidentally...if we don't low-pass!), and decide what sounds good, and have at it, LOL.

[Rob Strand]

What I found with the Lart is the response is sensitive to the transistor gains.   Without measuring the response it's difficult to know what you have actually built.   Nonetheless the Lart is a good place to start.  It sounds good with little effort to build.

I would also recommend going over the very long thread (or threads?) on the FSB forum.  A guy went through many designs.  However, be warned:  unfortunately some of the designs, even some good ones, had serious build errors.   That means the sound samples are not good representations of the design.   

At some point I did a review myself of many cabsims and they didn't match-up with the sound samples in the thread.   I built the designs in LTspice and tested them with sample input files - also gain tweaked the levels for best apples to apples comparisons.

In that thread, there were a few designs which were complex and had a few notches in the mids.   To me these sounded more like real speakers.  (At least on the days I was looking into it  )

is this a good idea?

I mean, I can change 2 caps, back and forth, no problem, but I don't have the basic experience of using cab sims...
any maybe what I'll like is something like the "charming fruity taste in wine", which shouldn't be there at all, other then the grape's taste, according to Bordeaux farmers

of the the 3 curves published by yeraym, the one with 3906/3904 transistors appears to me as better than the other two...
cause it goes down sooner around 4kHz..
but again, what "better for me" means might be anything from very bad to totally wrong

Those sharper cut-off designs that popped up a few years back definitely have a different character to the common low-pass filter designs.   If you use a lot of distortion or play metal they give better articulation in the high frequencies.

I would say judging any design based on a few numbers is almost impossible to get an idea what they will sound like.   It's not wise to impose any preconceived ideas regarding peaking.   These things are very much a whole package thing.  There's cases where even changing the peaks 1dB is noticeable.   In general small changes to the high frequency response are noticeable.

As far as tweaking the shape of the high frequencies to suit your taste - it's best done with a distorted tone.   A clean tone is much more forgiving with cab sims.

You can also run into judgment bias problems because overall sound depends on the cabsim and the EQ in the signal chain.  If your distortion pedal isn't 100% to your taste you might end up putting unnecessary EQ in the cab sim to compensate for the pedal - so you need to try many pedals.    If you do your testing on crappy headphones which roll-off the bass you can end up adding too much bass boost in the cabsim to compensate.