MIX BUS SATURATION

[Biasrocks]

As a circuit designer, with a long time interest in popular myth, I am not aware of any significant mechanism in linear solid state audio paths that significantly alter the sound's character with level, away from the extremes of clipping or noise floor.

The key word you used is 'linear'.

I have found that many times it's the non-linear behaviour of a piece of gear that achieves a 'sound'. A fairly common example might be, pushing an API preamp beyond the needed gain level to get some harmonic distortion.

I prefer older designs in this respect. Things that are discrete and have lot's of iron and sometimes tubes. I'm sure the original designers didn't intend for the equipment to be used this way, but many times pushing these designs beyond what the designer intended is what gives up the 'magic'.

Ultimately, I let me ears judge what I find pleasing, as should everyone.

I learned long ago to take 'internet legend' with a large grain of salt.

Regards,
Mark

 

[SSLtech]

Measuring the mix buses of several brands of console (Neve, Amek, SSL, API and others) I have never encountered ANY 'gradual' or 'gentle' or 'progressive' behaviour onset.

Could everyone please stop mentioning transformers and tape machines? -I don't remember any transformers in the summing bus... not even on an API. The transformers are at the OUTPUT... and most people who seem to 'bake' the buss, generally turn it down with the fader rather than having the needles pegged firmly and constantly against the endstop... Since the fader is before the output transformers, this renders that avenue essentially irrelevant.

I have to make a service call to a local studio with an early API console sometime soon. -I've looked after their gear for the last 17 years or so. They've had a few PSU modules go down over the years, and each time I've repaired them, I've always trimmed the rails to 16.0VDC.  -On that particular console I've even engineered a few records over those years, including two Deep Purple albums... matter of fact, I can describe how we recorded and mixed that one, since I think it most assuredly counts as a 'rock' album

All of the saturation and distortion has come from the OTHER side of the control room glass.  -Every last bit of it. -the first album I did there was tracked simultaneously to Studer Digital 48-track DASH and simultaneously to a 2" 24-track, but in the end the sheer inconvenience of waiting for machines to lock up during the lengthy overdubbing process (the album was recorded over about nine months) meant that we just used the DASH machine instead. -We contemplated locking up the 2" analog reel with the drum tracks on, but it simply wasn't deemed necessary... it sounded just FINE coming from the digital machine to everyone involved. It was mixed on an SSL 6000 E series (not even a G-series computer!).

We never saturated anything in the control room. Leslie amps? -Certainly. Guitar amplifiers? -You bet your bottom dollar. We used some outboard tube compression (A Tube Tech CL-1B and a Manley Variable-Mu) and some clever trickery as required here & there...

..But 'standard practice to saturate the mix bus'?

Absolutely not. -And I question that there's much to be gained by trying. -My measurement gear seems to back up what my ears tell me. -It doesn't sound much different, then all of a sudden you're out of headroom and you wish you weren't.

 

[JohnRoberts]

Amen...

JR

 

[barclaycon]

It's my recollection that vintage Neve designs employed a voltage summing arrangement using transformers.
Not a perfect arrangement by any means - as the bus impedance and channel output level is affected significantly by the number of source channels connected.
But it had an effect on the sound.

In fact, Waves have just launched a series of plugins called NLS, which attempt to mimic this 'Non-Linear Summing' effect.

 

[SSLtech]

I should say that I have no knowledge of that... such an arrangement would indeed possibly have some effects on summed signals... also, wouldn't such an arrangement mean that -in addition to nonlinearity from variable loading- the effective throughput gain would be influenced by number of assigned feeds also? It depends how it's implemented, I'm certain.

However, every one of the APIs, Neves, Ameks, SSLs, Harrisons etc which I've ever worked on have had virtual-earth summing with bucketloads of negative-feedback onto the summing node. With such massive NFB, the output signal tends to be essentially 'perfect sum' of the inputs, up to the point where the output swing can no longer swing (eg: when it meets a rail) far enough, or fast enough (eg: running out of slew rate) to correct for any nonlinearity.

 

[Biasrocks]

To be clear.

I wasn't saying you push gear all the time, everytime.

There are situations, in my working environment that have benefited from pushing gear beyond what would be considered normal operating parameters.

The mix bus I would put very low on the list.

On the API console.

I have a schematic for the 2488 console which shows three transformers on the Channel module, Input and Two outputs, specifically pre-amp out and on the Channel Output which goes to the bus assigns. The ACA schematic doesn't show transformers, only connectors which I'm assuming route out to transformers contained in the consoles frame.

I'm not going to say all API's are the same, especially since Keith says the one he works on lacks output transformers aside from the Main Output's. Jeff's console works with output transformers on the ACA busses and elsewhere I'm sure.

You can see Jeff's recreation of the ACA board from his console on his site.

http://classicapi.com/catalog/product_info.php?cPath=22_46&products_id=81

You can see a Jeffs console with the transformers for the ACA mounted to the frame here along with
a description of how the summing network is configured.

http://www.groupdiy.com/index.php?topic=33038.0

Regards,
Mark

 

[SSLtech]

There are situations, in my working environment that have benefited from pushing gear beyond what would be considered normal operating parameters.

Agreed, though the single greatest example is analog tape. All other things exhibit effects which I would characterize as being orders of magnitude less significant... including pretty much all transformer preamp stuff I've encountered.

The mix bus I would put very low on the list.

Very definitely agreed.

 

[Jean Clochet]

It's my recollection that vintage Neve designs employed a voltage summing arrangement using transformers.

Yep.  Pretty much all of the ones that are today considered 'classic' anyway.  Bus amps were the 40dB modules such as the 1272 and, later, the 3415. 

Not really advocating mix buses should be pushed here.  Just that, on those desks, there is a difference when it's done. 

There was also a mixing scheme used sometimes by Dick Swetthenham (Helios) which used what he called 'free grouping'  wherein any channel amplifier could be selected as a mix bus amp.  Users of some of those desks have said that 'pushing it' would be part of the process of getting the sound.  Not talking about every Helios desk here as not every one used the scheme.  Also, it's debatable whether Dick didn't actually get the transistor biasing a bit off on some early desks which tended towards a bit more distortion than was sometimes wanted.

IM (and others) HO of course. 

 

[Jean Clochet]

P.S.

Not a perfect arrangement by any means - as the bus impedance and channel output level is affected significantly by the number of source channels connected.

Both Neve and Swetthenham got around this to an extent by forcing the bus low with a terminating resistor which made the differences only a few dB between 1 or 2 channels only on the bus or all 24 channels.  Still a difference though and required more gain make-up regardless of channels assigned which added noise.

 

[Andy Peters]

I found that at some point it give me the "rich" sound, my client give a CD called Velet revolver as a mix references. I can't get the similar sound with my normal mixing process.

That sound is a result of the "mastering."

Ironic quotes definitely required.

-a

 

[sr1200]

What ever happened to "if it sounds right, it is right"?  I dunno, I've always made it a point to understand the "limitations" of my equipment (and understand what its like to exceed those limitations be it bad or good) so when a particular situation arises where that result of "misuse" is warranted, it can be used.  I can't for the life of me understand anyone that says "this is how it SHOULD be done".  It's the way that person WOULD have done it... and if Hendrix played his guitar the way his teacher said he should or if Mozart had followed the "rules" his employer set forth...
Just make it happen, if it sounds like ass, everyone is going to know it, if it turns out sounding amazing some idiot somewhere is going to assume that THAT METHOD is the way ALL recordings should be done. 

And if it sounded great leaving the studio, and it comes back sounding like dog fecal matter... don't use that mastering house ever again.
This concludes my rant... Im sure it pissed somebody off.

 

[simonsez]

I found that at some point it give me the "rich" sound, my client give a CD called Velet revolver as a mix references. I can't get the similar sound with my normal mixing process.

That sound is a result of the "mastering."

Ironic quotes definitely required.

-a

I know, but try to explain that to the boys... :-X

When the "right" mastering engineer doesn't available in my place, i should try to get my mix as close as the reference that they want, except the level. I think the saturation sound is a part of present rock mix, i can't get individual track saturation when using 16 ch summing mixer.

 

[0dbfs]

Saturation, clipping, and distortion can of course be used as special effects across a whole mix, individual elements, or anything for that matter.

A general rule of "do no harm" tells me that you want to keep your mix buses in the analog realm at or about +4. This allows for headroom which prevents clipping and allows the signal to therefore be "punchy" with regard to transients' etc.

The contemporary practice of "hitting the mix bus hard" as a rule is mostly a fad led by sales, marketing, and myth. There may be instances where that makes sense but in most cases maintaining headroom is preferred.

Your mix bus is not a guitar amp.

I also hear people say to not try to get your mix to sound like a mastered project/cd/recording. This is also bunk in my opinion. Why wouldn't I want my mix to be anything less that done?

Cheers,
Jb

 

[JohnRoberts]

Your mix bus is not a guitar amp.

Cheers,
Jb

+1

JR

 

[helterbelter]

Regarding that velvet revolver mix reference : you don't necessarily have to OD the mix bus.

The last mix I did (a while ago btw), I made 2 stereotracks of the mix.
1 normal mix, tapped of the masterbuss insert send, and 1 with additional inserts (gssl and an aphex exciter, plus maybe an eq or whatever, can't really recall). The first for the mastering engineer, the second one as listening copy for the band.

Funny thing, the second one was used for the mastering of the  cd, but that's beside the point. The point is, you don't necessarily have to choose what to do. Do both.

 

[maxiemixer]

There are many options regarding mix bus "saturation" that is in fact "adding some sort of distortion". I mix in the box for long enough to say that there is nothing really "wrong" with the digital mix bus of my HD system apart from the fact that it sounds different from an analog desk. That said the thing is very genre dependent. For rock I usually like to get some "saturation" with a careful use of selected plugins. I like ac1 in the console 2 setting drived a little (no more that 2 db reduction, a tiny bit of aphex AX (the waves take on this) and finally a couple of db of reduction with the ssl comp. I set the slowest attack and fastest release as a start, at 2:1 and then play a bit with the release first. For me this is enough to get that elusive "analog" sound without getting too much distortion. the cumulative effect of that is a light transient compression and a more controlled midrange. In that way I can push the mids harder without getting way too hard. I've been using for many (too many...) years the ssl e ang g series and that way I get the same general feeling. Of course is NOT the same thing but I like better the option of recalling instantly my mixes and jumping from one song to another to care much about the hassle of go back to the analog mixing. And my clients are not really ready to pay the extra money needed for a full analog path /(at least for now... I hope to get rich and famous so I will be able to charge thousands for a mix).
I also investigated in the "summing bus" option but the advantages are not enough for me to make things so much more complicated. But I found my way of implementing the "Michael Brauer" multi master technique so I definitely get interesting results in having different busses for different group of instruments, with different compression etc. And this is very addicting. Once you try it it's difficult to go back.
But in the end, in the digital domain, the most benefical thing that I discovered is to work way out of the digital full scale. in fact I now keep my recording and mixing level to a steady -18dbFS and leave the crunching of the bits to the mastering stage, possibly to a first class mastering house. And I like what I hear.
In the end a good recording starts with a great performance of a great musician/singer, a careful mic positioning, a good preamp (I like solid state, class A ones - the most important spec for me is linearity ut to 80KHz at least, that forces other specs to be up to that), and a good mic (I can live with SM57 if needed), and a good room. And a good monitoring environment so I can judge what goes to tape (or to the DAW) That's another reason why I like the PTHD: I can listen trough it with almost no latency and compensate for the crappy 192 (Actually I like mine even if there is som warmth in the 200 Hz range. For monitoring now I use the mytec DAC96 that my mastering engineers love, since they don't retouch my spectrum too much.
And lastly. I now know for sure that you get better results when you move fast and leave your istinct the main role.
Don't waste you time in the quest for the best way to saturate the mix bus. Improve on the skills that make the recording richer and more emotioning and leave the loudness war tho the deaf...
And please believe me: the best song on the radio is the one that has been left unsquashed...
Good luck!

 

[dirtyhanfri]

...the most important spec for me is linearity ut to 80KHz at least,...

 

[sr1200]

you know... for music for dogs... and bats...

 

[dirtyhanfri]

you know... for music for dogs... and bats...

Ok, I don't worry about then, I'm working just for a few (apparently) humans...

 

[JohnRoberts]

Linearity a couple octaves above 20khz is not unrealistic. One obvious example is close mic'd cymbals, that could generate plainly audible, lower frequency IM distortion components from above 20kHz content not being passed cleanly. 

Further, slew related limits are not always a sharp knee, but often a gradual deterioration in linearity as we approach slew limiting, so some headroom is not a bad thing to have.

JR