Passive summing w/ makeup gain vs line level amp saturation?

I asked Eric this exact question in his first Q&A video.
He said that it's possible that the difference is not the summing and the make-up gain, but just the signal running through the preamp like you said.
He also said that he will do test of using the line-in of the preamps with no summing and compare, but he never posted anything about it since.

Really just depends on the topology of the line amps vs the mic pre. Some mic pres when in line level bypass the input transformer for instance.

Overall, seems like a bit more work than worth it. I'd just use the silver bullet and call it a day.


Or if you want to experiment with something similar to what Eric is doing, grab some SB2 summing mixers from DIYRE. They are only $50 each, do 16x2, and take about 30 minutes to build. Then you can use the silver bullet for makeup gain or whatever mic pre you have and build a cascading setup just like Eric just no switches.

www.diyrecordingequipment.com/products/sb2-16x2-passive-summing-mixer

Well, for one, if he's doing it like that without a virtual earth summing network, each input to the summing network causes signal loss and the need for more gain.  Most modern consoles use virtual earth summing to get around this problem.

Loss before gain induces Noise Figure (NF) equal to the amount of signal loss plus the amplifier self noise.   You lose that dynamic range.  If he's needing to gain up 35dB as makeup, then you can expect that the noise floor is 35-40dB higher in his rig.

Summing resistor values need to be chosen that are not so low that the total parallel resistance of each resistor and each source impedance are too low for any single source to drive, but not so high that excessive signal loss occurs.  this can greatly affect the output levels and how much gain is needed vs. crosstalk and NF.

Passive summing doesn't sound like anything because it's just a resistor and some wire.

It's hard to separate out gain from the circuitry, because the more gain you ask from a circuit generally the more distortion it's going to produce. And, separately, distortion is also going to rise with absolute level.

So all things being equal if you're at +8 dBu you'll have more distortion than at +4. And, if you're doing 35 dB of gain you'll have more than at 10 dB gain.

Soooooo...both?

There's many ways to gain stage to accommodate make up gain without using summing boxes.  I'm a HUGE advocate of multiple analog stages before your final print - not just slapping one box on the 2 mix output of your DAW.  There's a bunch of stuff I wrote about that in the Silver Bullet manual, and actually, it's designed into the workflow of the box.  Tracking, re-Amping, bouncing, Busses, mix Buss, mastering.  I use em all over.

I saw this video weeks ago while I could not sleep and asked myself if he could not get similar results with much lesser effort.
And yes the SB also came to my mind...

Get yourself a CAPI SumBus and forget about it. Or, plus one it with the SumBus going into a Silver Bullet. Works for me and took a long, long journey to get there (including the Burl Vancouver, which is a great unit, IMHO).

Whatever is your jam, go for it. That's the thing. Find what works for you and your flow.

Its been a while since I watched that video, but I don't think he's using mic amps for makeup gain like I think you think he is?

Technically he probably could've called that video "I built a passive summing mixer" 

Passive mixer is just resistors and bus bar. There's gain loss so an amplifier stage is needed at the end of the line. It's an oversimplification but technically the summing bus of an old Neve is that with some 1272s for gain...

His thing wasn't cascading amp sections into each other, they're discrete sections.

Drums would be resistor network, an amp and get dumped to another resistor network. Rinse and repeat for vocals, instruments etc with one last discrete amplifier at the end of the line, the main resistor bus as the final 2 mix.

What his thing does vs a standard summing mixer is offer more control over gain staging in the analog domain. And yes, changing the amp topology will absolutely change the sound. 

There's no difference using a mic preamp for gain vs a 1272...Gain is gain.

And yeah, this and more is in a console. If you're using a large format console with virtual earth summing for best noise each subgroup might be summed in each bucket, 6 or 8 at a time, before dumping into the main bus. Or even 6:4:2 in some very large desks. This gives a noise advantage over all-into-one because of the way the noise adds from bucket to bucket.

One thing is that all amps have a preferred input impedance for best noise performance. The 1272 in a Neve uses an input transformer from the bus for this reason. Depending on how you set up the resistor network you may or may not be giving the "right" impedance for good noise.

As an easier solution to do what he did, I'd be strongly tempted to sum in the box to stereo stems, sending out of the DAW at, say, -40 dBFS or something into your preamps. Sum in the DAW, introduce an artificial loss, and makeup in analog. Then maybe a passive 4:1 mixer or something in the analog domain - or use hardware inserts and do this once more for the stereo bus.

You could do this in a bounce situation for a real nice comparison between your final mix with and without the cascade treatment by busing in the DAW, using a level-controlled hardware send from each bus, and printing the final sum separate from the 2track mixdown. Then any fader automation you have on your groups faithfully carries through automatically.