Analog Summing vs. Digital Summing

Posted: June 6, 2011 in Analog vs. Digital

ANALOG SUMMING vs. DIGITAL SUMMING

By Justin Fisher

With all of this “Analog Summing” craze going on right now, I’ve been doing a lot of reading lately on the sonic differences of summing signals together in the box (ITB) or out of the box (OTB) – Pro Tools in my case. Many engineers SAY there are astonishing differences in imaging, low end reproduction, depth, etc, etc. The tests I’ve heard online either showed almost NO difference at all, OR they were managed in such a way that I didn’t trust the final results.  SO, I decided to do a test for myself. Below you will find links to the test that I performed as well as a small journal of the steps I took. I attempted to make this as scientific as possible without going to impractical. I decided to choose 4 different songs to highlight the effects of the mixer in different situations.  There are four different styles available for you to download.  Electro/Pop Rock, Metal, Folk, and Southern Rock.  All of the bands showcased in this test are groups that I personally recorded and mixed at SmithLee Productions (except for the Electro/Pop Rock which I only mixed) and they all were kind enough to give me permission for you to download full quality versions of their tracks.  Please, take care not to distribute these songs, and if you like the artist, visit their websites and help them out for their generosity.

The bands and songs are as follows:

BAND – SONG – STYLE

1. Bears on Fire -“Girls Aren’t Robots” – Electro/Pop Rock
2. Black Fast – “Chaos Orb” – Metal
3. Mike Kendrick and the Blue Eyed Sons – “Old Time Love” – Folk
4. Lida Una – “One Last Victim” – Southern Rock

In doing this test, a friend of mine and fellow engineer brought up a good point. We were talking on the phone about how to deal with a certain routing problem for this test when he told me he performed the same test himself and came to an insightful conclusion.  He said, there aren’t going to be that many differences between the two until you actually mix THROUGH a summing mixer from the beginning.  To do a test after the fact won’t result in many changes.  The real magic happens when you are making your mix decisions based on the fact that you are going through a mixer in the first place.  That will change your though process and in turn the way you mix.  This statement really made sense to me and probably explains why I didn’t hear a lot of differences in the tests previous engineers posted on the internet.  They are putting the mixer on after something as already been mixes specifically for ITB. It would be like finishing a mix and then putting a compressor on the 2 bus after you are done.  You should be mixing through it from the start.

Immediately below the test specifications are links to the files. You will download a ZIP file that contains three versions of the same song.  All files are 24bit 44.1kHz WAV. The versions included are

1. Original Pro Tools mix summed completely in the box

2. Original Pro Tools mix summed in the box with the 2-Bus mix going out of the 192, to the summing mixer as a stereo mix, and back into PT.

3. Same mix summed externally via analog summing mixer.

The songs were all mixed by me in Pro Tools HD3 version 9 via a 192 I/O.

The summing mixer used was a Speck Electronics X.Sum mixer which is a very clean and quiet summing mixer courtesy of Barry Hufker.

So without further ado, here are the files! It’s obviously up to you how you perform this listening test, but I suggest loading them into a DAW and renaming them randomly, have someone else switch the files for you while they are playing, or make sure you aren’t looking at what’s what while you are listening. Otherwise the test is pointless since you will most definitely be biased one way or the other.  I won’t inject any of my own thoughts here to keep from skewing the results.

1. Bears on Fire – “Girls Aren’t Robots” – Electro/Pop Rock

2. Black Fast – “Chaos Orb” – Metal

3. Mike Kendrick and the Blue Eyed Sons – “Old Time Love” – Folk

4. Lida Una – “One Last Victim” – Southern Rock

I hope you enjoy the test and at this point I would like to list the steps I took to ensure that the testing remained accurate.  There will always be someone that says, “well you didn’t do this” or “you did this wrong”.  If that’s the case, then you should obviously be trying it yourself.

SUMMING TEST STEPS

1. Found sessions appropriate to cover a multitude of genres and textures, but that also fit my personal style.

2. Made a new session coy of each of the four songs just for this test, each with its own individual routing options.

3. Bounced a new PT only mix for each of the songs.  Some of these songs were mixed by going out to an external 2bus compressor such as the Amek 9098CL or API 527s.  It doesn’t matter to you that much because you’ve never heard these songs before, so you wouldn’t know if they sounded any different with the compressors engaged.  I did replace these compressors with a similar 2 bus compressor on the pro tools master.

4. Got the Speck X.Sum mixer set up and ready to go.

5. Checked all TRS to TT cables to make sure they all worked and were phase coherent with each other. Generously used DeOxIt to clean all contacts.

6. I had a few less cables than I had hoped and not lots of extra hours to make new cables so I consolidated some of my sums to come up with this:

– Kick – Mono

– Snare – Mono

– Drums – Stereo

– Bass – Mono

– Guitars – Stereo

– Keys – Stereo

– Vocals – Stereo

– Effects – Stereo

7 .Hooked up the mixer to my patchbay and sent signal out to the summing mixer to check that each channel is level matched to 0.2db. (knobs on the speck were kind of hard to get levels to match any better.

8. set up my outputs in Pro Tools to send to the Speck.  Made sure all outputs were at the same level of -20dBvu.

9. I also did a mix where it was summed internally in Pro Tools but the 2 mix from Pro Tools went to the Speck box.

10. Once all 3 versions of the 4 songs where bounced, I brought them in to a new Pro Tools session and made sure levels were matched between each version of the song.

Advertisements
Comments
  1. Tr says:

    Hey glitchfactor,

    I came across this page goggling summing–thanks for taking the time to do this!

    Listening to the tracks…
    At first, I put them into a DAW and switched between but obviously it was bias and hoped summing made a difference. So, instead, I played the three versions of the songs in media player and turned off my computer screen not knowing what order they were playing in.

    Things that I listened for…
    So, why did I hope summing made a difference? Well, I felt like my mixes in PT were a bit narrow in the stereo field and felt like it was always uptight/choked and bit harsh. In that respect, these were the qualities I listened for when trying to identify which was the PT internally summed mix.
    In regards to analog summing, I hoped to identify a more natural, relaxed separation of the instruments. I was on the look out for an increase in stereo width and how the overall mix of instruments fit together better.

    Using the Bears on Fire track as a general consensus…
    The most prominent instrument that made a difference between PT internal vs Analog summed was the piano key. The Internal mix felt a bit veiled and lacked detail, however, in the Analog version, it had clarity and separation (however, I feel like maybe the level matching of the keys was louder in the Analog–could be why I felt like it was ‘better’). However, since I second-guessed that factor, I was convinced that Analog summing made a difference when the double-tracked guitars came in. It actually increased (slight) in stereo width vs the PT internal. There was also a more relaxed feeling listening to the chorus in the Analog sum.
    I did not feel like the 2-bus sum was too far off from the all Analog sum but it was easier to listen to than the PT internal.

    Even for the Southern Rock track I picked Analog sum twice and 2-bus once as the ones I preferred.
    The width and spatial factor were the things I noticed and enjoyed more. For some reason, the PT internal just sounded choked to me.

    Oh, what I noticed w/ the Bears on Fire track was the clarity and detail of the delays on the vocal. The Analog summed was more detailed and better heard–smoother I’d say.

    *However, the Metal track was one I didn’t really listen to because I felt the differences were not as noticable

    I’d like to get into summing, but, I gotta upgrade my converter first and probably need a 16-channel DA. I listened to these tracks through my Digi 003 rack, so, if I heard differences through the 003 DA, I guess it isn’t THAT bad lol

    • glitchfactor says:

      Thanks for reading! I appreciate the comment and I hope it helped you out!

      -Justin

      • Stefan W. says:

        Really great comparison. The Speck makes the tracks come alive a bit but the differences were more subtle than I would have expected. The Speck seems to be quite transparent but still adds that little “analog” feel to the mix (wider stereo image, warmer sound and added punch) which sets it apart from ITB mixes. I listened through headphones through a good headphone amplifier.

        I’m currently testing a Nicerizer 16 and I think it has a more obvious punch effect on drums that I have been missing in ITB mixes.

      • glitchfactor says:

        Thanks for listening Stefan. I wish I had a Nicerizer to compare! The Nicerizer and the Speck are kind of on the opposite ends of the spectrum as far as summing in concerned and I’d love to hear it. My taste is probably towards the Nicerizer since I’m looking for warmth and gel, but I was honestly surprised at how much a “transparent” summing mixer helped in that regard.

  2. Mårten H says:

    Hi, great comparison! Thanks for the hard work. I am considering an analog summing mixer vs some plugin (like Slate VCC) right now.

    I only listened to Bears On Fire so far, and the biggest surprise was the apparent lack of bass response in both the analog variants. Am I the only one hearing this? The rest of the instruments sounded “punchier” in the analog mix, but I felt part of the bottom end was lacking, especially in the chorus.

    • Stefanw says:

      The lacking lowend in analog summed Bears On Fire is probably caused by trying to make OTB- and ITB-mixes comparable. I don’t think any analog summing mixer will make the bass response weaker. My experience is the opposite. Analog summing adds punch to the lowend in a mix in a way that is very hard to get ITB.

      I highly recommend both VCC and analog summing mixers. I use both on every mix. VCC is of course the cheaper place to start. Analog summing requires quality converters, quality cables and a mixer. An analog attenuator like A-Designs ATTY may also be needed unless your AD-converters have analog level controls. This way you can push the output stage of the summing box and reduce the level again when going back ITB.

      I can only speak for myself, but I’m pretty sure that once you go OTB you will not want to go ITB again. But this doesn’t of course exclude using plugins.

    • glitchfactor says:

      Yes, this was something I noticed, and was kind of a strange anomaly. But keep in mind. This is not exactly a scientific test. Just setting up a bunch of outputs from your convertor, running them through a summing mixer and pulling those tracks back in won’t always result in a complete 1:1 leveled mix…..especially with the Speck mixer I used which has volume pots. That was a dilemma I had. Some things just didn’t quite sound right as passed through the mixer, but I didn’t want to change the mix because that wouldn’t be a direct comparison.

      Anyway, the point is, there is not perfect way of doing this test. This was just me doing it for myself and hoping that others could maybe gain something from it. I set out to see if a summing mixer would improve the sound ONLY by keeping my mix the same…and in my opinion, it did. Where it gets crazy, is the fact that working with a summing mixer from the BEGINNING of a mix, rather than putting it in the chain after it’s done will COMPLETELY and totally change your workflow and thus, your mix. I think this is the crux of why it’s such a heated topic right now.

      But again, thank you for noticing that, and no you aren’t the only one 😉

  3. G.L. says:

    Valueable comparison ! , no doubt in my ears, that the file -Girls Arent Robots – Analog Summing – sounds so much better overall than the other two mixes by all means.
    Thank you

  4. Patrick Bakker says:

    Thnx, for the very helpful test!
    Now I know I don’t need to spend $$$ for minor differences that apparently I am not capable of hearing, lol. I’ve also found another summing mixer shootout and had the same experience (http://toddburkeisarecordengineer.blogspot.com).
    That leaves my crusade of enhancing my ITB mixes to get back to perfecting my skills…

  5. DW says:

    Thanks so much for this test.

    In one of the songs, the analog stuff made it a little muddy, you could fix this in a real mix though, I’m glad you didn’t change anything for the test, so you can really hear what the summing/analog is doing.

    On the other three songs I clearly preferred the X-Sum mix downs. I just hear more space and separation. Some of the instruments even sound better, like the hi-hats. The ITB mix downs sound flat, congested, so familiar to me as that’s the sound I’ve always struggled with on my long serious ITB mix sessions over the years. It’s a wall I hit that I could never get past–now I know how!

    The 2-buss analog sounds halfway between, not as wide and separated as the analog summing, but still with a nice extra fatness or grit or whatever that the ITB is lacking, a little more muscle. I could see this being a great go-to for ITB mixes when you don’t have the time, budget, or energy for full on analog mixing. Like if you need something done fast, and recallable, throw some analog on the 2-buss and get a little extra mojo there sounds like a great method vs. 100% ITB.

    So I’m convinced I need to build a simple summing mixer. I’m going to roughly follow the NY Dave / DIY Recording Equipment plan, and end up with something similar to a Folcrom for way less money (DIY). Just need to stay motivated… Thanks again for doing this.

    • glitchfactor says:

      Hey no problem! I’m glad you got something out of it! In the next few weeks, I’m going to do a similar thing. Read the latest post on my blog, and you’ll see what it’s all about!

      _Justin

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s