When mixing on a computer, many prefer to sum the tracks in the analog domain, which can result in more depth of field and a wider, more defined stereo image. There are several analog summing boxes on the market, but they can be very expensive. Naiant is now making a very affordable passive analog summing cable that could be a cool alternative. Please comment on this post if you have any experience with it.
An Inexpensive Analog Summing Option
September 19th, 2007 by TW · 16 Comments
Tags: Mixing
16 responses so far ↓
1 Anders // Sep 20, 2007 at 4:04 am
Has anyone seen any experiments or tests if you can tell the difference between digital and analog summing what so ever? Like a blind test listening to mix A and B and say which sounds better. I have mixed both analog and digital and I’m not sure if there is a real measurable difference in the output file or if the difference is “just” psychological.
2 TW // Sep 20, 2007 at 9:46 am
Hi, check out this thread at the electrical forum.
3 Anders // Sep 20, 2007 at 1:58 pm
I’m a big fan of taking the sound outside the DAW, running it through some amp, compressor or any old cool gear and rerecording it. I believe that the unlinear behaviour of analogue stuff can alter the sound in positive ways. I am however skeptical to that the actual summing of signals is that more different in a DAW than in a summing box. In a DAW its just adding two or more numbers, right? What does the summing box do instead?
But I’m really interested in finding out! Could any of you who have access to a summing box do some tests and post files somewhere? Maybe like this:
- Use the summing box to record a mix of some tracks, call the file “Summing box mix”
- Then record each track by itself just running through the summing box. No summing, just going through the same D/A, cables, transformers and A/D. Panned mono tracks should be recorded in stereo I guess?
- Sum these rerecorded tracks with the DAW summing. Call this file “DAW summing mix”.
It would be really interesting to hear the difference!
4 Anders // Sep 20, 2007 at 2:00 pm
Sorry, thanks for the link! Interesting reading!
5 TW // Sep 20, 2007 at 2:00 pm
There are actually dozens of these tests out there on the internet. Google it.
6 Anders // Sep 21, 2007 at 9:16 am
OK, I found this one: http://vintageking.com/site/files/sumshoot.htm, but since they only compare straight DAW and straight analog summing, it doesn’t really say that much I think. On the other hand, my test method won’t work either since it will add (uncorrelated) noise each time you run a single channel through the summing mixer. Must be +12dB for 16 channels or something?
I will do some tests with my crappy Behringer mixer. Please don’t laugh.
7 TW // Sep 21, 2007 at 9:22 am
I’m confused about what you are trying to test. I thought you wanted to compare DAW summing with analog summing. That vintage king does exactly that…it pairs several analog summing boxes against a protools bounced mix.
What test method are you proposing?
I don’t think using the Behringer mixer is a good idea…it may improve the stereo imaging, but it will add some terrible sounding artifacts.
8 Anders // Sep 21, 2007 at 12:00 pm
DAW summing is (theoretically) perfect: 1+1 = 2. Analog summing is imperfect: 1 + 1 = 2.001, but it sounds better.
I believe that analog summing sounds good (or at least different) not because the actual summing is different, but because you run the audio through D/A, cables, resistors, transformers , valves, and A/D possibly introducing good sounding compression, distorsion and noise. I wonder if running each separate track alone through a summing box and the summing the rerecorded tracks in the DAW creates the same audible result. I want to test ONLY the summing, not the D/A -> valves -> A/D chain.
9 TW // Sep 21, 2007 at 12:40 pm
Well, there are a lot of problems with that experiment.
Are you saying that you want to compare the analog summing with DAW summing with each track run individually through the summing device?
In that case, instead of adding noise _once_ at the summing stage, you’re doing it 16-32 times, and then summing digitally. What’s the point?
10 Anders // Sep 23, 2007 at 6:04 pm
Yes I realized that (see post Sep 21).
This is getting silly, but my point is that I don’t believe the benefits in “clarity” or “stereo image” from analog summing mixers come from the actual mathematical summing (adding signal levels) but from the fact that the tracks pass through a lot of analog components adding good-sounding compression and distortion. I believe you could get the same audible result from running each track in your mix through a single channel strip of good quality.
11 TW // Sep 23, 2007 at 6:34 pm
Sorry man, that is totally incorrect.
12 Anders // Sep 24, 2007 at 12:21 pm
OK show me
Seriously, you don’t have to, but I would really like to see a test where you isolate the summing.
13 TW // Sep 24, 2007 at 12:26 pm
There’s no such thing as “isolating the summing”. Your test removes the analog summing from the equation. This is a dead end.
14 VelvetoneStudios // Oct 29, 2007 at 5:47 pm
Yeah, individual “summing” is technically impossible. You’re talking about sending tracks out to be processed by analog devices and bringing them back to a session. I do this a good bit, but this is, by definition, not summing. (you can’t combine 1 source with nothing!) The theory behind analog summing is all of the seperate sources being fed into a summing buss, just like on a console. Things happen to the actual combined audio in the circuitry of the summming buss, while in the digital domain the combined streams of 1’s and0’s are digitally captured as another set of 1’s and 0’s
15 Anders // Nov 6, 2007 at 6:54 am
Hi Valvetone! I understand (hopefully!) that you can’t sum a single channel. I’m interested in a test where you eliminate the DA/AD conversion and as much of the extra analog components (tubes etc) as possible and just benchmark an analog summing bus vs. adding floating point numbers in a CPU.
I can clearly hear that using analog summing mixers add something good to the mix, I’m just not sure that it depends on the actual summing.
What kind of analog processing of tracks do you do (I’m curious)? I have a TL Audio 2051 preamp that I send stuff through, and also a customized 10W Hagström bass amplifier that sounds great.
Here is a picture:
http://www.haskinshagstroms.com/John/img/sub_sub_section/965.jpg
16 Anders // Jan 17, 2008 at 10:19 am
Hi, I found a Digidesign white paper that I founnd interesting:
Gannon Kashiwa - The Pro Tools 48-bit Mixer
http://akmedia.digidesign.com/support/docs/48_Bit_Mixer_26688.pdf
Leave a Comment