"Crush-n-Blend"

[BradM]

I just stumbled upon this thread in a search. If I wanted to implement this passively would I just have the 10kA pots and the 4.99k resistors?

thanks,
Brad

 

[SSLtech]

Don't even attempt this passively.

Sorry, not all lunches have a free option.

Keith

 

[BradM]

Can you explain a little more? I don't have enough of an electrical engineering background to understand what the pitfalls might be.

thanks,
Brad

 

[matthias]

I hope to find enough time to finish one of my next planned projects... a parallel mix controller...

do you have any other useful features in mind that could be implemented..
maybe a hp filter for the wet signal?

 

[SSLtech]

[quote author="Brad McGowan"]Can you explain a little more? I don't have enough of an electrical engineering background to understand what the pitfalls might be.[/quote]
Well it does go pretty fundamental, but it's like saying I'd like to build a passive compressor. -Could you do it? -Yes of course. -Would it be any good, or anything like insertion loss-free? -No.

Same exact argument.

Keith

 

[amorris]

damn guys you just gave away my best trick!!

why else do consoles have 80 faders?

 

[drpat]

Cowbell returns!!!!

 

[matthias]

hello !

I just made some simulations with the crush'n'blend circuit provided by keith ...
in case my simulation is correct, I found out that the "blend curve"
isn't really linear...

http://twin-x.com/groupdiy/albums/userpics/crush1.GIF

the curves on the scope shows the sweep of the mix pot from one end to the other end.
you can see, that in the middle the signal amplitude is 1,5 as high than on the single ends..

I think all the pan-circuits you find in the most common mixers also have this issue..
if you turn full left, the signal isn't as loud, as in the mid position..

so what do you think.. is it possible to make this circuit more linear??
or do you think the performance is good enough..

mat

 

[Kit]

You could use a 2-gang pot for better panning.

 

[clintrubber]

Dutch Blender

Never enough time and the to-DIY-list is endless, so if projects can be combined it'd be nice.
And it's the Netherlands here, we're cheap :wink:

I'm talking about combining a M/S-matrix circuit and the Crush'n'Blend.

So it occured to me that this circuit

http://www.uneeda-audio.com/mat03.pdf

could easily do Crush'n'Blend as well, with the re-configuration done by a simple toggle and/or some re-plugging:

#1 It should be prevented that Mid & Side each reach both U1 & U4.
#2 It needs normalled inserts for the G-SSL(-etc)-returns before U2 & U5.

That's all, signal-polarities are as desired (even number of inversions).
Maybe the panning-law now wil feel differently, but I've used this kind of panning before and it felt OK for an FX-blend. And it's obviously not balanced in and out, but neither are my desk-inserts.

I'll add a pic of the intended mods later.

It's more the idea than a real net-saving I guess, although the overhead for drilling holes, adding PSU etc is of course now only required just once.

Bye,

Peter

 

[SSLtech]

Matthias,

Didn't see your post until now, but basically R1 & R2 (should be same values at all times) will affect the center gain to edge gain relationship) whereas R3 and R4 should affect the mid-left and mid-right parts of the curve...

If you still have the simulation stored, I suggest massaging those resistor values till you get a smooth match, but the difference should certainly be well under a dB.

-Considering that the "mashing" of the compresor/limiter will have a stupendous modulating effect on the gain, I'm certain that any static error in tha balance law will be completely eclipsed, but I don't have a simulator here, so I'd be intereted to see what some different values would do: I came up with those component values in excel...

Keith

 

[matthias]

thank you all for your help!

I didn't really finish this and in the meantime I was working on some other things...

i can remember that I made some other simulations with a circuit that is very similar to the one "clintrubber" posted.

the tracking with a dual pot turned out to work really good.

but if you want to blend stereo material a circuit with one pot/channel
would be better...

I'll come back to this soon and will post my results.

mat

 

[clintrubber]

[quote author="matthias"]the tracking with a dual pot turned out to work really good.

but if you want to blend stereo material a circuit with one pot/channel
would be better...[/quote]
I had in mind to use a 2*12 rotary switch for the M/S-matrix, giving 'perfect' tracking.
I think I could live with a 12-position blend-control for dry & crushed (G-SSL etc) as well.

Bye,

Peter

 

[matthias]

of course...

in case of implementing this into the gssl I would use a second vca and sum into the same i/v converter

then you can easily adjust the levels with the control voltages.

 

[clintrubber]

[quote author="matthias"]of course...

in case of implementing this into the gssl I would use a second vca and sum into the same i/v converter

then you can easily adjust the levels with the control voltages.[/quote]
Sure, 'local' blends can be done (although I recall various discussions here about blends that seemed to work but in the end weren't fully doing it). But when you'll only be crushing by g-sll it'll be the simplest route.

And as Keith pointed out the C'n'B is universal, so can be wrapped around about anything, and that can come in handy once you've got developed an appetite for dry+wet-FX-combinations.

And I like the 'short' path for the dry-component.

Bye,

Peter

 

[rrs]

Just wondering to best way to power up this Crush-n-Blend circuit?

I'm thinking of using a external 12v power transformer, how would I wire it up ??

Help much appreciated.

 

[clintrubber]

[quote author="rrs"]Just wondering to best way to power up this Crush-n-Blend circuit?

I'm thinking of using a external 12v power transformer, how would I wire it up ??

Help much appreciated.[/quote]

Better use a TX with a bit more volts, like 15 or 18 (and make it a dual 15: 30V with a centre tap, or two 15 v windings.
Reason: likely the stuff you want to blend has a +/-15V dc supply as well, so let's be prepared for the max.audio-signals it can give.

The complete supply can be like the G-SSL schematic for instance, see the www.gyraf.dk site..

If you know how to do it and if your TX has enough power you could power it well from say a single 3V winding as well, but it requires more complexity. I mean, the mentioned method is the easiest, and if you're a bit deeper into this you'll know
how to use a single 15, or a single 9 Vac winding as well, but let's keep away from the voltage doubling stuff for now etc.

 

[Larrchild]

I think for a grin, you should invert one opamp and listen to the difference signal between compressed and dry! :grin:
Could be the next big sound.

 

[vertiges]

:shock: :shock: :shock:

:green:

eD)))

 

[MHanson]

Just for the sake of curiousity. Have you thought about adding any buffers for loading problems instead of just y'ing off the two units?? Most current stuff won't care I'm sure, but . . . .

Otherwise, I loved seeing this. I was just contemplating building something like this in the next couple of weeks. You just made my life easier. Thanks.

Michael