Question for Experienced Hardware Techs - How Much Difference Was There Between Individual Channels in Classic, High-End Consoles?

[cpsmusic]

Hi All,

As I mentioned in another thread, I had an interesting discussion today with a well-respected mixing engineer. One topic that was touched on was Brainworx's TMT (Tolerance Modelling Technology) which according to their website

Due to the hundreds of parts in each channel strip, there are subtle differences between each channel strip in a console. These differences are responsible for the beautiful imaging and the characteristics that we know and love from analog consoles. And that's why Brainworx came up with the idea to implement TMT to their channel strips. To have the possibility to add little differences between channel strips in the box and to make your mix feel more organic and less sterile.

I've always been pretty sceptical about TMT. What I'm curious about is how different channels on classic high-end consoles were/are? I imagine that over time, as manufacturing processes improved, and prices decreased, channel-to-channel differences would have decreased. But it doesn't say a lot for the console manufacturers! What about quality control and component matching, etc.?

What I'm curious about is whether anyone who's worked on these consoles and repaired or maintained them has first-hand experience of this? How were they calibrated? Did they vary from channel to channel?

Also, I realise that component ageing can play a role in this, but I'm primarily interested in new and well-maintained units.

Cheers,

Chris

 

[ruffrecords]

The single most important factor in minimising the differences between channels due to small variations in component is negative feedback.

All components have tolerances and slightly different characteristics. The effect of these differences is reduced by the amount of negative feedback in each circuit block. This might typically be 30dB or 40dB which will reduce the effects of component variation by 30 to 100 times. So 5% component tolerance is reduced to 0.5%.

Only components outside the negative feedback loop will not be subject to this. So coupling capacitors and input/output transformers will affect the sonund.

Cheers

ian

 

[jacomart]

My opinion is that often these are commercial gimmicks, in the end you will make the adjustments you deem necessary on each channel strip section (gain, comp, eq, saturation, etc.) adapting it to the type of signal it will have to deal with and no one will be equal to the other. Rarely, if ever, will you have two channels with identical settings and the same signal running through to notice a substantial difference (If ever there will be).

Cheers
JM

 

[scott2000]

As I mentioned in another thread, I had an interesting discussion today with a well-respected mixing engineer. One topic that was touched on was Brainworx's TMT (Tolerance Modelling Technology)

I missed this....What did they have to say about it?

off topic but that Knifonium plug is on sale now for $40.... The real one looks insane...

 

[jacomart]

I missed this....What did they have to say about it?

off topic but that Knifonium plug is on sale now for $39.... The real one looks insane...

Here

Cheers
JM

 

[scott2000]

Thanks...I'll watch it ....

Sounds maybe like a similar theory to un-linking stereo channels in a compressor to get a different feel ..

 

[user 37518]

My opinion is that often these are commercial gimmicks

Absolutely agree, I can't believe that plug-ins like these are now a thing, after decades of audio design new BS still comes out every day! it shouldn't surprise me....

Whenever I hear something like "organic and less sterile" I immediately know its something that the Marketing department created.

And yes, most analog consoles have differences between channels, it is not something desirable if you ask me....

Whats next?

I have an idea:

"People back in the 60's and 70's used to smoke in the studio, this created a thin layer of nicotine and tar on top of the circuitry, faders and pots, which made the equipment sound more 'musical' and 'upfront' giving it that old school vibe, which is why with the help of our friends from Phillip Morris we modeled this effect with our advanced 'Marlboro Convolution System' (MCS)"

Perhaps I should put that in my CV and apply for a job at Brainworx

 

[pucho812]

the only time I have ever seen every channel strip behave exactly the same was when the desk was brand new.
This was because any channel strip that didn't test 100% out to spec was replaced before it went in. This was because whoever purchased it spent upwards of 6 figures on the price.
But like everything else, overtime, strips drift and usual factors like heat and so forth can play a huge roll in what is happening and how much each channel behaves like the other.
Most common desks to have issues between channels is the Neve V series consoles as they were so hot temperature wise that it ultimately would cause problems.
None of the things I describe are things we want, but when analog was the only option we just had to work with it as is. Digital emulations of problems we had to deal with are not favorable to me. I would say they are up there with the price of vintage gear, even the crap we didn't want when it was new is now vintage and expensive.

 

[scott2000]

I have an idea:

"People back in the 60's and 70's used to smoke in the studio, this created a thin layer of nicotine and tar on top of the circuitry, faders and pots, which made the equipment sound more 'musical' and 'upfront' giving it that old school vibe, which is why with the help of our friends from Phillip Morris we modeled this effect with our advanced 'Marlboro Convolution System' (MCS)"

Perhaps I should put that in my CV and apply for a job at Brainworx

lol...

 

[JohnRoberts]

It is the intent of console designers to make every channel as similar as possible.

I have advised people right here when troubleshooting old consoles to perform null tests between channels to look for outlier response deviation between channels. This technique can help identify old dried out caps and the like. Most consoles can perform the null using an internal bus. If you don't have an input polarity switch you can make a polarity flip cable.

JR

 

[porkyc]

Going back to the original question; No.
I've commissioned a lot of new MCI and Harrison consoles and a few SSLs over the years and this never came up, even with some of the "golden-eared" fader pushers.
Mind you, the fact it lit up was often the main criteria!

 

[pucho812]

When a console was new, there was a lot of work to make sure channels were the same. After years of use, some will fair better then others due to location in the desk and how often they get used. This effects switches, pots and mechanicals. Even in a used desk, there usually is minimal sonic difference between channels unless something is really wrong with the channels.

 

[hodad]

I'm not sure I buy the marketing hype, but I approve of the concept. If you want the most accurate emulation, then you'd want to allow for some slight component variation. In real life, there will be subtle variations from unit to unit of a "classic" compressor, or from channel to channel of a "classic" console. The variation's likely to be more significant in something with transistors or tubes, or with carbon film or even carbon comp resistors. Some of the differences are likely to be extremely subtle, and I don't know that they're going to add "magic fairy dust" to your ITB mix; but if you want verisimilitude, modeling subtle component variations is going to get you one step closer.

 

[Michael Tibes]

I use those Plugin-Alliance Plugins a lot, but I have honestly never bothered with the TMT. I'm not sure if I'm missing out on something, but doing my adjustments and then stepping through tiny variations doesn't tickle my interest.

The original consoles - when new - were pretty high quality and I wouldn't expect any serious differences between the channels. But then again they were developed to work with tape machines, so tiny variations were part of the whole concept and the consoles were probably the most exact element in the chain. The pots on SSLs were very expensive custom made high-end parts many with specials like center taps to improve performance, I believe Neve wasn't any worse. A single SSL pot did cost more than a PA plugin on sale, a whole channel far more than the whole bundle...

TMT might have a huge impact if they model a real world ok-ish maintained 30+ year old console which will most certainly have at least audible and relevant differences between the channels due to aging. If that is what they did, they should probably label their variations like 'broken comp' or something alike ;-)

What I'm curious about is whether anyone who's worked on these consoles and repaired or maintained them has first-hand experience of this? How were they calibrated? Did they vary from channel to channel?

I never thought about that back in the day, looking back I'd say yes, they did - but only very little and much less than the variations between analog tape machines. When something broke in a channel we would just swap the channel with a working spare and I can't remember anyone ever complaining that the 'magic channel' was missing. Calibration was mostly done at the factory, I don't recall having done any systematic calibration on SSL 4/6ks or the Neve V. Most circuits were anyway opamp-based, I only recall calibration pots from some eqs - and those were factory calibrated and sealed.

As I already said, I use these plugins a lot, but for me the 'magic' is not in the TMT variations.

Michael

 

[JohnRoberts]

I believe I already said this, but consoles were designed to deliver similar performance channel to channel. In circuits where it mattered precision components could be used. No need for trims, except perhaps on level meters.

For context look at the the variation between microphones and loudspeakers. They make tape recorders look good.

JR

 

[Kingston]

I have an idea:

"People back in the 60's and 70's used to smoke in the studio, this created a thin layer of nicotine and tar on top of the circuitry, faders and pots, which made the equipment sound more 'musical' and 'upfront' giving it that old school vibe, which is why with the help of our friends from Phillip Morris we modeled this effect with our advanced 'Marlboro Convolution System' (MCS)"

I know you are just trying to be funny but this is a real thing. I've discussed with a sixties radio studio engineer who came across this in a bad way. Back then everybody smoked everywhere, inside, in studio, during radio interviews, the hosts themselves smoked. At some point the Neumann mics of the radio station had to be serviced due to excessive tar residue. New capsules basically.

Apparently everyone complained the mics sounded too bright with the new capsules.

 

[hodad]

I believe I already said this, but consoles were designed to deliver similar performance channel to channel. In circuits where it mattered precision components could be used. No need for trims, except perhaps on level meters.

JR

Yes, and I can show you discrete amplifier modules from a "legendary" console that use carbon comp resistors. Despite that, console modules track fairly closely--which I guess is a credit to the design & construction--but you might see a tenth or two dB variation at various frequencies between modules--and that's without the inductor-based EQ engaged. And almost certainly there are many more variations between modules that I don't have good enough equipment to measure.

If you're talking about an SSL, there's probably next to no difference between modules.

And yes, variations in mics & speakers are certainly far more significant.

 

[Gold]

I remember helping the chief tech calibrate the gains on the Neve 8048 and the broadcast Neve about twice a year.

 

[JohnRoberts]

Zeroing a console with mechanical VU meters can take a while (and involve trims). It is easy to see small fractions of a dB deviations above or below 0VU (see not hear) with a mechanical VU meter. LED meters rarely if ever offer even 1dB resolution around 0VU. I preferred 3 dB steps full range (for other reasons).

JR

 

[Brian Roth]

It's been too many years, but I seem to recall the Otari Concept One desk meters had 1/4 dB steps just above and below 0 VU.

Bri