Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[guerillamixer]

Not being a tech, and mixing entirely in the box at the moment, my knowledge of what goes on inside an analog console is limited to say the least. As I'm currently experimenting with bussing and summing inside a DAW, I have a few questions for the assembled tech masterminds, here goes...

Does channel bleed mostly result from the summing buss?

Do solo/PFL systems contribute?

Does the amount of bleed vary the further up you go? Eg, does Ch2 recieve more than CH16 of CH1's signal?

Does the bleed signal vary in a consistant manner? Eg Low freq roll off or a regular drop in gain by 3db per channel?

I realise that summing methods vary from console to console, and different designs would bleed differently, but any  advice will be much appreciated.

Thanks,

Andrei

[nother]

Physical relations, like ch 1 being closer th ch 2 will often be a factor, also don't forget the patch bay.

[Bill Yacey]

Bleed in an analog desk is usually due to capacitive coupling on the buss lines running between all the channel strips. If I understand your question, you are experiencing bleed within a DAW setup. This bleed must be occuring in the analog portion of the circuitry, so I would suspect again capacitive coupling or poor analog electronic switches. Are you sure the bleed is happening internally, or through cabling and patchbays as someone suggested?

[guerillamixer]

Sorry, I haven't made myself clear...

There is no bleed in my DAW of course, but when  I used to mix through a desk,  The channel bleed was low but audible and  I felt at the time that it contributed to the sound of the mixes in a not insigificant manner.

I've been experimenting with creating bleed inside a DAW, starting of at 80db down for closest channel then dropping by 3db per channel after that.  The same mix, with and with out bleed, had an greater  volume level with bleed of course , but also  felt spacially a little different.

My questions relate to identifying what makes up that contribution in an analog console, or even in specific designs of console,  to fool around with this idea in a little more informed fashion

[moogus]

SHHHH!

Crosstalk is the secret of the 'analog sound'.

Dont spread it round, ok?!?



Well, its not that simple, but a big componant IMO.



M

[John Klett]

crosstalk is created over several different paths...  noise moves around similar paths...

electrostatic...  capacitive leakage across wires in ribbon cables and traces on circuit boards will be loaded toward higher frequencies.  If you view two circuit board traces that are next to each other as plates of a capacitor the equivalent resistance to crosstalk between them will go down as frequency goes up.  This is on common characteristic of crosstalk in analog desks.

inductive...  every wire - even a straight wire - has an inductive component.  If you view two circuit board traces that are next to each other as wire turns in a transformer the equivalent resistance to crosstalk between them will go down as frequency goes down...  I have not found this to be as dominant a source of cross talk as electrostatic coupling (high frequency) in analog desks but it is a factor.

resistive...  this is pretty huge.  Circuit "Common" (or Ground, or Zero Volts) should be as near zero impedance as possible and all source (sink) to a central point that can be referred to as a master ground or master common...  the lowest impedance point in the console.  This can be a buss bar or a central lug or some other central distribution point...  not that I call it a distribution point.  Circuit common is almost always zero volts and is almost always going to be tied to ground.  If a group of four channels are tied properly to a common but the common they are tied to has some resistance to ground then the channels will more easily bleed signal to each other than they will to ground.  This type of crosstalk tends to be full bandwidth.

combinations...  an inductance in the path to ground will allow low frequencies to have a low impedance common circuit reference while high frequencies will see a less low impedance reference...  so highs will be more likely to cross talk.

an aggravator...  one large problem in many consoles is the insert point and unbalanced INPUTS in general.  If you take any device that has a actively balanced transformerless output and plug that in to an unbalanced input such that the "low side" of the active balanced output is feeding in to ground you will create a crosstalk problem.  The reason is that active balanced outputs are two amplifiers driving - say, pin 2 and pin 3 of an xlr - they are driving the pins - the hi and the low - out of phase, but in all other respects, identically.  This is a differential signal and can be called balanced.  When you plug one of these outputs into an unbalanced input you will short one of the two amplifiers to ground and in many cases the remaining amplifier will shift level up by 6dB the compensate for the loss of half the signal.  So...  the amplifier shorted to ground does not turn off and does not stop driving audio current to ground.  It will distort because it is driving a short.  What happens in the distorted audio current travels on an imperfect ground and some of it will distribute into parts of the console, like buss amplifiers, where you can hear it.  A current passing through anything with a resistance will create a voltage and in this case it a distorted harmonically rich noise voltage.

In many mid range consoles the crosstalk will be high frequency as more or less "clean" capacitive couple leakage, mixed with some distorted midrange active-output-half-shorted-to-ground crud, mix with some more of less full range stuff.

So...  if you want to replicate that you would make the crosstalk in your workstation by taking the signal and boosting the high end by 6dB per octave or so and then mixing that with the same signal that has been passed through some sort of exciter...  if it was an analog device I'd say you wanted to oldest Aphex Exciter you can find but, in the land of plugs you are on your own.  You would want to have the high end leak stuff down to around -70 or so at 10K and the gritty distortion (which you'll have to mess with to get it just right) maybe at -60 in the upper mid-band.  Generally you'd cut all the low end off the grunge.  Transformer input consoles don't have the grunge problem so if you want a more Neve-like crosstalk you bail on the grunge thing and keep all the crosstalk really clean and low

[guerillamixer]

wow.

Thanks John!

[Audioboffin]

Crosstalk? Q: What's in common with every channel strip in a console ? A: The power supply. The first place to look if you want to reduce crosstalk is improving PSU decoupling "at the point where the rail meets the strip" if you like.

On the other hand, some pwer supplies are already regulated, and adding regulation on each channel strip may drop the final rail voltage to below that where the regulators can properly regulate, and as a result the rails get noisy... so the answer is to then bump up the output voltage at the actual PSU, so the channel strip regulators can actually regulate. What I'm trying to say is some of these issues can take a lot of engineering to reduce or eliminate. Something you probably don't ever have to consider when "mixing in the box".

[Brian Roth]

I've been chasing a bit of bleed in a LARGE Amek desk, and while the jury is still out, I keep running into what appears to be a common denominator.

If the "low" end of any channel's fader isn't PERFECTLY tied to the ground reference of the fader buffer stage, then some signal will leak.  Worst case is when the low end of the fader has a totally open circuit...the fader will only have a few dB of adjustment range!

In less extreme examples, I've seen cases where merely un-mating, then reconnecting, the connectors in/out of the fader will clear up "bleed".

Bri