graphic EQ as cabinet simulator?

[dacaumodo]

Hi all
perusing ROG and other sources about cabinet simulators, it occured to me that a cab sim seems to be mostly about EQing and cutting frequencies. Am I wrong? Is it possible to simulate a cabinet only with a graphic EQ, say Boss's GE7 (which, by a complete coincidence, I own)?

[bancika]

I've been thinking the same. It should be fine, just find frequency response of cab you'd like to simulate and try to match it. More bands you have the better...

[ashcat_lt]

I haven't done a whole lot of research in this area, but I have two concerns i'll throw out there for your consideration.

1)  Boosting a band on an EQ is not exactly the same thing as adding resonance in the same band.  This probably won't make much difference and is probably ignored in most cab sims.

2)  Are the extreme bands on your EQ in the right place, and are they shelving.  This will be much more important.  Again, I haven't recently looked at the boss eq, but just keep in mind that most guitar cabinets actually restrict the frequency range of the guitar, with fairly steep rolloffs on either end of the spectrum.

[bancika]

why couldn't you roll down bands on each side? On ROG site there's frequency response curve of Marshall cab I think. Try replicating it.

[lovric]

bands on most stomp EQs are wrong especially in the treble region. you try and see.

grab a LA3600 chip ($1) and make the proper one yourself.

[dacaumodo]

why couldn't you roll down bands on each side? On ROG site there's frequency response curve of Marshall cab I think. Try replicating it.

Yup, I'll try. Thanks.

[dacaumodo]

2)  Are the extreme bands on your EQ in the right place, and are they shelving.  This will be much more important.  Again, I haven't recently looked at the boss eq, but just keep in mind that most guitar cabinets actually restrict the frequency range of the guitar, with fairly steep rolloffs on either end of the spectrum.

I'm not sure what shelving means, but if it means that EQ bands are 'linked' (so that if you cut two neighbouring frequency pots, the frequencies in between are also cut) I don't know. Don't know how I could check either (not sure this can be done by ear alone)

[dacaumodo]

bands on most stomp EQs are wrong especially in the treble region. you try and see.

grab a LA3600 chip ($1) and make the proper one yourself.

Are you referring to a DIY EQ?

[GFR]

The filters in a 1-octave graphic eq are not steep enough to act as a (good) cabinet simulator. Perhaps with a 1/3rd octave one.

You can do a quick check, start with the EQ flat and then cut everything above 4kHz, all the way down.

[dacaumodo]

The filters in a 1-octave graphic eq are not steep enough to act as a (good) cabinet simulator. Perhaps with a 1/3rd octave one.

You can do a quick check, start with the EQ flat and then cut everything above 4kHz, all the way down.

Thanks, will do. Perhaps an EQ + hi pass & low pass filter would work then...but getting dangerously more complicated than a cab sim!

[Mark Hammer]

I wish I had the time to sit down and design one, but I've been saying for a while that an "ideal" guitar EQ pedal or EQ unit would have a variable lowpass filter, and a couple of bands of semi-parametric (frequency adjust but no Q adjust) boost/cut.

What people generally mean by "cabinet simulator" is not a simulation of just the box the speakers are in (even though you'll see terms like "ope-back" and "closed-back" used), but rather a simulation of the box with a certain category of speakers in them.  The cabinet will certainly introduce some low end characteristics in terms of resonances and lowest possible frequency, but will generally have no impact on the high end.  The speaker itself will introduce distinctive peaks and rolloffs.

While I wouldn't expect anything intended to be a single-weekend DIY project with a couple of dual op-amp chips and a small handful of passive components to accurate capture all the nuances of a "true" simulation of a particular 2x12 open-back cab with this pair of Celestions, looking at the frequency response curves of  number of speakers like these below:
http://www.tubesandmore.com/images/inv/p8r-4-graph.gif
http://www.tubesandmore.com/images/inv/c12n-8-graph.gif
http://www.tubesandmore.com/images/inv/p15n-graph.gif
...tells me that one could get *closer* to mimicking the sound of such and such a cabinet by having the control complement described.

[dacaumodo]

Mmmh sounds very good... something that would be both a very versatile analog EQ and an analog cab/speak modeller then? I'd buy one!  This plugged into a multiband compressor... but no, I'm letting my imagination run wild here!

Makes me think of this plugin:
http://www.voxengo.com/product/boogex/

[ashcat_lt]

I'm not sure what shelving means, but if it means that EQ bands are 'linked' (so that if you cut two neighbouring frequency pots, the frequencies in between are also cut) I don't know. Don't know how I could check either (not sure this can be done by ear alone)

nope, not that the bands are linked.  A shelving eq means that, rather than boosting or cutting a relative small part of the spectrum centered around the filter frequency, it affects everything below (in a low-shelf) or above (in a high shelft) the cuttoff frequency.  Pretty much like a hi-pass or lo-pass filter.

[Mark Hammer]

The classic "treble" and "bass" controls on a stereo amp are shelving-type controls.  They provide both boost and cut to content above (for treble) and below (for bass) the designated "corner frequency".  On some slightly fancier stereo amps and preamps, you sometimes see switches for selecting the corner frequency.  I have an old Luxman that labelled those "mild" and "intense", because moving the corner frequency closer to the middle range (and away from the extreme ends of the frequency spectrum) would result in more content being potentially boosted or cut by the tone control, hence "mild" change and "intense" change.

The overwhelming majority of shelving-type tone controls have relatively shallow slopes, generally 6db/octave. In contrast, tunable lowpass and highpass filters can have more severe/steep slopes, like 12, 18, 24, and more db/octave.  As well, the gain around the corner frequency can be altered to produce "resonances", such that there is a little peak just before the point where frequency response declines abruptly.

Not to beat a dead horse, but if you look at the frequency response of different speakers, you'll see a characteristic rolloff at the bottom, the top, a little bump in frequency response near the rolloffs, and a dip here and there in between.

A unit with 18db/oct tunable highpass (for taming the lows), and 12db/oct tunable lowpass (for adjusting the high end), and appropriately ranged resonance/Q for each, coupled with 2 semi-parametric resonant sections (with cut/boost and adjustable centre-frequency for each) would yield a really flexible unit that could not only provide pretty decent EQ for different voicings, but also provide the capability for reasonably decent cab/speaker simulation.  That's a total of 8 pots there, with more than enough tonal adjustment potential to easily rival a 15-band or greater graphic EQ.  I suppose if a person wanted to, it would be possible to forego some things like the highpass tunability to free up some critical panel space, replacing it with two toggles for res1/res2, freq1/freq2.  Given the awkwardness of tuning a 3-pole filter, that may actually be a better alternative in some ways.

[dacaumodo]

You got me drooling there (especially since I understood almost everything yo said!) Are we the only ones to think this would be one BRILLIANT project? The pots space problem is tough indeed but... High pass tunability is crucial in such a circuit don't you think? Or perhaps having 3 way toggles to add some flexibility...

Honestly it sounds so good (and yet so much beyond my tech abilities) I'd be ready to finance a prototype (I just invented this synonym to "can someone build it for me?" - neat, ain't it? )

Guillaume

[lovric]

beautiful reading on EQs regarding guitar is http://sound.westhost.com/project75.htm.

a LA3600 is 5-band equaliser in a single chip made by Sanyo. when you decide on your frequency response curve, you can do any dips, peaks and resonances with it up to +/- 18 dB. there is a schematic in data sheet which i used to start.

than, condor IS a graphic EQ, only with fixed frequencies. they (ROG) made it precisely because the stompbox EQs couldn't deliver since their resolution is just crude.

[Killthepopular]

I've just been doing this myself. You need a peak at 3k and a steep rolloff after that, so the best starting point is to set 3.2 at +15db and 6.4 at -15db. Then pretty much leave those there and adjust the other sliders to taste. It works surprisingly well. I've only been playing with it for a bit but I'm loving the sound. So far i prefer it to my Behringer GDI21 and my JOYO california sound as a cabinet simulator.

[Killthepopular]

If anyone wants to hear an example, the following song was recorded using just a GE7 and some distortion boxes going into the soundcard. Main guitar part is gretsch pro jet > metal zone > GE7 > audio interface. Can't remember the settings but the basic rule of thumb should be that 3.2k is your highest slider and 6.4k is your lowest slider. Obviously it's a direct sound, doesn't sound like a real speaker, but doesn't sound bad either. Like i said above, it sounds better than my joyo/behringer amp sim pedals. There's no EQ on the track except for the GE7.

https://soundcloud.com/user-152978877/a-toast-to-something

[GibsonGM]

The free Boogex VST is here: http://www.vst4free.com/free_vst.php?id=1534

Doesn't seem to drop out until you buy it, like the one linked above does.   This looks like a pretty useful plugin!

[Killthepopular]

I've studied a few frequency response charts to help me figure out what settings to dial in to get a classic guitar speaker type sound. I compared 3 classic celestion 12" speakers, the vintage 30, greenback and alnico blue. Here are some guidelines for dialling in your graphic eq for 12" speaker simulation based on what I've found. Works best with the boss GE7; MXR eqs don't seem to be ideally arranged for these frequencies.

1. The average position of the overall peak is 2.7k. Make sure your nearest slider to 2.7k is the highest.

2. There's a steep dropoff above this point so any sliders above your highest slider should probably be pulled down quite steeply.

3. The low end rolloff starts around 140hz, so your slider immediately below this frequency should be lower than your slider that is immediately above this frequency.

4. There is a dip around 1.6k, so your nearest slider to 1.6k should be lower than its adjacent sliders.

Follow these guidelines and you should have a good starting point for a frequency curve that sounds like a classic 12" speaker.