Passive summing boxes necessity vs wants

Was looking at a new passive summer with make up gain recently, was 16 channels  made up of 20 opamps and 4 transformers.  I am wondering how necessary is it using the opamps for an input buffer?

For passive summing I would say not at all. But you might like the sound of the extra opamps, which is a harder question to answer.

If you are gonna use all the extra opamps it might be interesting to make a switchable passive / virtual earth summer.

If you want a passive summing box, keep it passive and don't mess with the extra headache of powering those op-amps.  Of course, you could do a single-ended, off-the-shelf, wall-wart style power supply on the cheap...  But then with op-amps, it's kind of an active box now, isn't it?   

well is  passive summing with make up gain considered active?

pucho812 said:

well is  passive summing with make up gain considered active?

Click to expand...

Yes, because it is.....

There is precious little difference between the noise gain of a virtual earth inverting summer , and a passive sum with equivalent post gain stage.

I do not understand the popularity of passive sum boxes... perhaps because it's relatively easy to DIY, and then add you favorite flavor du jour mic preamp.

JR

JohnRoberts said:

pucho812 said:

well is  passive summing with make up gain considered active?

Click to expand...

Yes, because it is.....

There is precious little difference between the noise gain of a virtual earth inverting summer , and a passive sum with equivalent post gain stage.

I do not understand the popularity of passive sum boxes... perhaps because it's relatively easy to DIY, and then add you favorite flavor du jour mic preamp.

JR

Click to expand...

I can sum it up for you. No pun intended.  People who mix entirely in the computer have been sold this idea that in order to get maximum results, you must mix down to several stereo pairs of things and then go out of the computer in to a box that takes these pairs  and sums them into stereo either passive with make up or via virtual earth.  Then you take that stereo pair and go back into the computer to record it.  I personally feel they could easily get similar results by going out of the computer and back with just the extra conversion stages or even running  out into line amps and back into the computer.

Why don't you visit "them"?  "Their" wishes, needs, and opinions are important to me because my gig is to enable artists to be as creative as possible with the least amount of hindrance.
In a production sense standardizing stereo stems in some way make a world of difference when fine-tuning mixes for release.  So the mixing options are 1 bounce the stems mix to stereo ITB, 2 bring the stereo mix OTB and re-record it, 3 equally sum the stems with a passive/make-up box and back in, and 4 go through any number of active consoles that may be around and back in.
When you prefer one method over the others maybe you find you are one of "them" ;-)
Mike

FWIW I look at this from the perspective of circuit design and how the electrons behave.

The original complaint about mixing inside the box was that digital summing was somehow compromised. (It isn't or shouldn't be.)

Mixing outside the box makes it easier to apply legacy external efx.

The customer is always right even when ........

JR

pucho812 said:

Was looking at a new passive summer with make up gain recently, was 16 channels  made up of 20 opamps and 4 transformers.  I am wondering how necessary is it using the opamps for an input buffer?

Click to expand...

That doesn't sound very passive to me.

No need to get micro-agressive, here.  We don't want Ethan to have to create a Safe-Space on the site!

There are differences between passive mixing (with gain make up) and virtual earth (active) mixing. For passive mixing:

1. Interaction between channel controls depends on the driving source impedance and the number of channels summed.
2, Crosstalk between buses driven from a common channel also depends on the driving source impedance and the number of channels summed.
3. To maintain a constant gain, unused inputs need to be shorted.

For virtual earth (VE) mixing:

1. Interaction between controls and crosstalk depends principally on the VE amplifier open loop gain (which is frequency dependent.
2. Inputs can be left open or shorted without affecting the gain of other channels
3. Bus capacitance can seriously affect stability

As others have pointed out, there is nothing to choose between them from the noise point of view.

Cheers

Ian

I would say try summing for yourself and then decide. For me, even with an identical outboard 2 bus chain, it sounds better with things spread out and summed analog with multiple d/a than summed itb feeding a single d/a.

But that's what is fun about DIY, you can adjust things to your own liking 

I come from console land. If I want Analog summing, I will use a desk.  summing boxes do not do it for me, their value is more hype then substance.

ruffrecords said:

There are differences between passive mixing (with gain make up) and virtual earth (active) mixing. For passive mixing:

1. Interaction between channel controls depends on the driving source impedance and the number of channels summed.
2, Crosstalk between buses driven from a common channel also depends on the driving source impedance and the number of channels summed.
3. To maintain a constant gain, unused inputs need to be shorted.

For virtual earth (VE) mixing:

1. Interaction between controls and crosstalk depends principally on the VE amplifier open loop gain (which is frequency dependent.
2. Inputs can be left open or shorted without affecting the gain of other channels
3. Bus capacitance can seriously affect stability

As others have pointed out, there is nothing to choose between them from the noise point of view.

Cheers

Ian

Click to expand...

Perhaps I'm repeating myself but there are other variants. One uses the transamp (Cohen?) topology for the sum amp, which doesn't suffer from loop gain margin deficit when summing a large number if stems.

Another a  little esoteric but for decades I've messed around with substituting current sources for the input resistors in a typical VE topology. In practice (at least practice for 30 years ago) the current sources are not completely noiseless, so you don't get the full benefit of dramatically dropping the effective noise gain of a VE sum bus, but you can get a perhaps 20+ dB reduction in bus noise and the less worried about reduced distortion and reduced phase shift.  Operating a VE bus amp (or equivalent make up gain stage ) will show negative effects from consuming the loop gain margin by operating at elevated noise gain. (Some modern uber op amps show promise for old VE topology)

I really really really wish there was something to this "outside the box" analog summing but to the best of my knowledge there is no "there" there... Digital summing is lossless, in fact the resultant sum has more dynamic range than the stems, albeit more than you can practically use. 

I could improve upon my old current source summing design using quieter modern op amps for the current sources, but agree that this OTB trend is mostly a fashion statement, so we shouldn't waste real money on improving the actual performance which isn't actually driving the popularity (and it will still never be better than digital summing).

Just my $0.02 from decades of following this. Studios are notorious for following different paths to get their personal sound so it's all good even if not really better (or silly).

JR

PS: I covered this in an article I published in a recording magazine back in 1980.

JohnRoberts said:

Perhaps I'm repeating myself but there are other variants. One uses the transamp (Cohen?) topology for the sum amp, which doesn't suffer from loop gain margin deficit when summing a large number if stems.
JR

PS: I covered this in an article I published in a recording magazine back in 1980.

Click to expand...

I think the transamp is an oft missed opportunity. I am not aware of any current op amp, audio or otherwise, that is a transamp. Seems like a good opportunity for a new DOA. Transamp benefits are numerous and not just limited to VE mixing.  Maintaining a near constant loop gain means it is no longer necessary to use the brute force sub audio dominant pole technique to ensure stability that is employed in most op amps. Open loop gain can be maintained throughput the audio spectrum which means the benefits or NFB are applied equally to all audio frequencies.

Cheers

Ian

ruffrecords said:

JohnRoberts said:

Perhaps I'm repeating myself but there are other variants. One uses the transamp (Cohen?) topology for the sum amp, which doesn't suffer from loop gain margin deficit when summing a large number if stems.
JR

PS: I covered this in an article I published in a recording magazine back in 1980.

Click to expand...

I think the transamp is an oft missed opportunity. I am not aware of any current op amp, audio or otherwise, that is a transamp. Seems like a good opportunity for a new DOA. Transamp benefits are numerous and not just limited to VE mixing.  Maintaining a near constant loop gain means it is no longer necessary to use the brute force sub audio dominant pole technique to ensure stability that is employed in most op amps. Open loop gain can be maintained throughput the audio spectrum which means the benefits or NFB are applied equally to all audio frequencies.

Cheers

Ian

Click to expand...

The transamp/cohen topology is widely used for mic preamps (even IC ones). 

I think I've seen similar topology used in some ICs but not for low noise applications, more for higher frequency wide bandwidth applications.

Console summing amps are a very low volume application so unlikely to attract dedicated silicon.

I suspect you could take a mic preamp chip like a THAT 1510-1512. Ground the inputs, and cap couple balanced VE sum bus resistors into the two gain ports (RG1 and RG2).

But analog consoles are so last century..... 8) 8)  time to move on.

JR 

The interesting thing is that while summing digitally would seem to be simple, in practice the different software systems approach this differently and get different audible results. When summing a large number of channels in Protools something does happen if levels are too high, and Protools does not cope. One workaround is to put an attenuator plugin on every channel and pull it down 20 dB. Mixes do sound better. I got rubbished on the Protools panel for suggesting this! But one reason for the OTB experience is that it works, and one of the reasons is that the summed levels are lower. They have got the maths wrong. As you would know John, if you design a multichannel mixer, then the output summing amp needs to work at - gain for each individual input to allow for the summed level. Then follow with some makeup gain.

radardoug said:

The interesting thing is that while summing digitally would seem to be simple, in practice the different software systems approach this differently and get different audible results. When summing a large number of channels in Protools something does happen if levels are too high, and Protools does not cope. One workaround is to put an attenuator plugin on every channel and pull it down 20 dB. Mixes do sound better. I got rubbished on the Protools panel for suggesting this! But one reason for the OTB experience is that it works, and one of the reasons is that the summed levels are lower. They have got the maths wrong. As you would know John, if you design a multichannel mixer, then the output summing amp needs to work at - gain for each individual input to allow for the summed level. Then follow with some makeup gain.

Click to expand...

A common disconnect between digital mixing and analog, is that nominal 0VU signals on analog meters are 20+dB below clipping or saturation.  OdB(FS) in the digital domain is clipping, so adding together a bunch of hot digital signals will overload.

Most modest channel count consoles can handle stems from the channels to the mix bus without padding, thanks to the 20 dB of headroom, but one console of mine had over 100 stems feeding the L/R bus so I put a -10dB trim on that bus master.

IIRC Greg Mackie made a big deal about running his buses 6dB cooler, but IMO fixing something that isn't really a problem. He was good about doing things like that and turning them into marketing hooks he could advertise. It doesn't matter what your bus gain is if you clip the post fader gain stage in the channel before you hit the bus... 

But like i said analog consoles are old news.

JR

JohnRoberts said:

The transamp/cohen topology is widely used for mic preamps (even IC ones). 
JR

Click to expand...

I have been unable to find any. Can you point me to a couple?

Cheers

Ian

ruffrecords said:

JohnRoberts said:

The transamp/cohen topology is widely used for mic preamps (even IC ones). 
JR

Click to expand...

I have been unable to find any. Can you point me to a couple?

Cheers



Ian

Click to expand...

I already did...

look at this data sheet http://www.thatcorp.com/datashts/THAT_1510-1512_Datasheet.pdf

The Transamp was pretty remarkable for the 70's/80's but these modern THAT chips would kick it's ass all around the block.

Just like the Transamp you need to cap couple into the emitters for a sum amp application.

JR