Is DP's summing wack?

[qo]

Wow, guys, thanks for all the thoughtful replies and useful links! This has been a learning experience for me, for sure.

I'm going to try a test this weekend. I'm sure this has already been thought of, but maybe not this particular implementation. What do you think of this?

1. Take a white noise file (and maybe also a pink noise file) and copy them.
2. Invert the copies.
3. Create pairs of tracks in DP; one for the original file and one for the copy
4. Start with one pair, just to make sure we get complete cancellation

5. If so, duplicate this setup to 30 tracks and adjust the faders on each pair of tracks to different levels (that is, each pair has the same level, but different pairs have different levels).

OK, so now we should still have complete cancellation, right?

6. Adjust the Master fader (still we should have complete cancellation).

If DP's summing is at least mathematically perfect then the above test should pass, correct? And, if the above passes, then I'll certainly be convinced. And, yes, if Bob Katz sees no problems with DP's summing, then I'm really on shaky ground suggesting that there might be

[stickwolf]

Yeah, that sounds like a perfect test. I'll be shocked if anything goes against what would be expected.

[alphajerk]

heavy accumulation of noise? not with any decent analog system. let's leave mackie and behringer out of this., eh? The only time i ever experience analog noise is when I patch in some older equipment (ie: eventide instant flanger WHOOSH). my console is very quiet, as are all pro desks. ITB guys who try OTB do tend to change their tune. For under $5k you can get a fantastic pro analog console used these days.

relatively speaking... im NOT talking about mackie/behringer... no matter how quiet your desk is, it is NOT as quiet as digital blackness period. not that i even care to argue the fact. i started on analog... but there is more creativity ITB mixing for me and i have a lot more fun ITB than OTB. and name a PRO board you can get for $5k that doesnt need to be fully recapped and restored? oh wait, you said UNDER $5k. if i found an API, neve, or even some quad8 id take it apart and sell the modules for more than $5k and make a boatload.

[danny]

relatively speaking... im NOT talking about mackie/behringer... no matter how quiet your desk is, it is NOT as quiet as digital blackness period. not that i even care to argue the fact. i started on analog... but there is more creativity ITB mixing for me and i have a lot more fun ITB than OTB. and name a PRO board you can get for $5k that doesnt need to be fully recapped and restored? oh wait, you said UNDER $5k. if i found an API, neve, or even some quad8 id take it apart and sell the modules for more than $5k and make a boatload.

I would gladly address all your points in detail, but you seem to want a duel, not a discussion. I decline.

[alphajerk]

there is no duel... but name one PRO board under $5k. i dont care about any of your points otherwise.

[zandurian]

You guys are making me crazy!!! I've got an early MCI board that is just MAGIC but have mixed ITB for years because of full recall ability (which really is a HUGE advantage in many ways). After reading all this though I'm tempted to run the MCI faders at unity and feed them individually while still doing the level and effects automation from DP (using the MCI faders only for panning). Of course it will also be tempting to grab some vintage eq at that point, but I know mixes would sound GREAT done that way.
I do respect people who can do a mix and then just say "That's it, it's done", but for us "tweekers" out there, the idea of going back to endlessly patching in outboard gear (and never having as many processors as needed) and then, once it's all unplugged, not being able to go back and touch it up a little - it's not feasable.
Does anyone here really doubt that analog mixing sounds better though? I try to deny it because I don't want to mix the old fashioned way ever again (w/o full recall) mainly because I could never get all the changes right and it's just too much damn work. But in my heart of hearts, I know I would like the result if adopting the above mentioned technique.
But then mixing would be real time only. Man, I'm so spoiled with bouncing to disk....

[alphajerk]

i doubt analog mixing sounds "better"... better is a subjective term anyway. most blind tests i have seen most people do prefer the ITB mix by a large margin [most run like 90% picking the ITB mix], and when commenting on it think its the analog mix to boot!?!

i think the analog summing is better mantra has been repeated so many times that people assume the better sounding mix in blind tests are analog and are suprised to find out otherwise.

in the end its ALL the user... its akin to blaming a cheap mexican strat for sounding like •••• when its really the guitar player who sounds like ••••. let me make my first reference... i have a friend who is producing a band for a major label... now let me say, i dont see summing as ITB vs. OTB... its REALLY digital vs analog summing because even digital boards are simply computers with a HUI. now mick guzaski [sp?], mixing digitally beat out andy wallaces mix, mixed analog. both guys are some of the top mixers in the industry. did digital win that round and we can say its BETTER?!? not really... it was just the mix that captured what the producer wanted. its NOT the tools.

i have used MCI boards... ones mans magic is another mans distortion i guess. not what i would call "pro"... but then again, i hate the term "pro" anyway in reference to gear. a PRO is someone who doesnt blame his tools. this thread might have tempted you to try to sum through it, but why stop with its summing? run the mix through it if you are going to go to all that trouble, add EQ's, comps OTB. i mean What the heck?!? though i seriously hear you on the 'tweaker' comment.... i LOVE chunks in DP [the ONLY feature keeping me on DP at this point] because i can literally mix and sequence an album from start to finish instead of having 10-15 project files having to open and close each one for an album.


here is another little analogy i heard yesterday. someone is shooting a movie here in town. they just stopped production on it because the DP felt they needed some particular camera to proceed with the movie and they cant locate one yet. i heard they fired the guy and i hear [might possibly be true, might not] that spike jonez took over and claimed 'i dont need that camera to make a movie... i will use whatever i have at hand and get it done'... thats being PRO.

so while people piss and moan about some supposed summing issue in DP [and other DAWs] others out there are just getting it done.


i have an album i an wrapping up the tracking on now and i got to mix it... i have a now dying dualGHZ on the upgrade to OSX/4.6 that i can either mix in 4.6 and not use hardly any plugins on and mix what i have, go back to OS9/3.1 and mix with plugins but w/o ADC, or go to another studio and mix on an analog board, or find a way to upgrade to a quadG5 and find some FW interface [yeah, im not PRO about pissing and moaning about apple fully getting rid of the PCI slots in favor of PCIe exclusively, it pisses me off highly right now as i need to upgrade but didnt really want to go FW for a audio interface at this point].

i think about what i want to do in the mix and know my automation is going to be pretty extensive, so the analog option is probably out, OS9/3.11 is running favorably right now... but it could very well be mixed on an analog board at another studio [more because i want to use some of their vintage compressors on the mix than analog summing cause im looking for LOTS of color]....

but regardless of which way i end up choosing, its gonna get done... fans will buy it and listen to it AND THEY WONT GIVE A RATS ASS ABOUT WHAT IT WAS MIXED ON!

[chrispick]

Geez, alphajerk, take a deep breath, man. Kick your feet up; chill.

This thread is going overboard.

Original poster, qo, simply offered to run a few summing tests on his own time and dime, mostly to quell his own curiosity. No one's trying to shove any production philosophies down anyone else's throat or anything like that. I don't think this thread was ever intended to devolve into an ITB vs OTB debate (which is a tired, pedantic exercise anyway). It certainly needn't drift farther into a who's-a-pro / who's-not-a-pro quagmire.

I, for one, have learned some things from this thread -- and qo hasn't even posted any results yet.

In the end, I know I have room for improvement when it comes to mixing. Whether or not I'm a "pro" or not, I can't answer; I do get paid to do it. I'd just like to feel greater confidence that my tools do not need to improve (and given the Bob Katz quote, I'm definitely leaning that direction).

[qo]

Yeah, I don't have much bias one way or the other with respect to how music is produced and so agree with Chris' suggestion that we try to remain neutral and objective here. I started this thread based on an initial subjective impression that could very well be inaccurate. That is, because the Ghost sounded different, I equated that to better on first listen with a specific set of tracks (a section of a tune that contains a sea of layered vox parts). And, this was all intended as an investigation into DP's summing (or, as mentioned in my original post, any other aspect of DP that might contribute to a difference).

If ITB (or OTB) was clearly superior, then things would be different. But, based on contributions to this thread, and some of the links posted here, and other threads I've read, I'd say neither has an upper hand and so will probably just use whatever methods seem more efficient and/or conducive to getting a particular track finished. This also includes monitoring, which the Ghost excels at (since the Apogee's don't have zero-latency monitoring capability unless used in Advanced Routing Mode and that limits you to 48kHz). So, I'll still be using the Ghost when tracking simply because I've had several complaints about latency from clients.

Nonetheless, I still plan the white/pink noise test this weekend which should lend a bit of objectivity.

[Fibes]

Nonetheless, I still plan the white/pink noise test this weekend which should lend a bit of objectivity.

We do have to keep in mind that tones, sweeps and all sorts of things are merely testing how it hadles THAT. While it can produce some insight into weaknesses the only REAL test is with actual audio. There are plenty of devices that "spec out" perfectly with normal testing procedures but when faced with audio completely crap the bed.

Since i have backed off my mixes/tracks using the -6 concept things are much better on my end. YMMV.

[Don T]

Hello:
OK, so I have to say something about that supposed digital blackness. I started a thread about noise but there is not much interest. Have you done a noise floor test on your A/D? Our new motu 24I/O's are running -87db s/n. That is not digital blackness, infact it's not much better than my old G16 Fostex MTR (dolby C). I don't find either ITB or OTB better, just different. I use what ever will create the sound I'm after or improve the ones I already have. Generally these days, I tend to stay ITB until it fails me. That being said, for the high dollar projects I usually start on analog. Almost every project ends up a hybrid of some sort.

[stickwolf]

Digital blackness is just "0" nothing, silent. Your noise from your converter is the noise of whatever the converter is getting from its circuitry. Theoretical digital silence is totally silent.

[zandurian]

most blind tests i have seen most people do prefer the ITB mix by a large margin [most run like 90% picking the ITB mix], and when commenting on it think its the analog mix to boot!?!

Wow, that really surprises me!!! I would like to read that report - where is it? I guess if ITB is kicking butt by 90 percent in blind tests then all of us analog lovers better shut up

[zandurian]

i have used MCI boards... ones mans magic is another mans distortion i guess. not what i would call "pro"...

You mean I paid $40,000.00 for a 40 channel, automated, plasma metered NON-PRO BOARD (used - 1993) ???!!! Dang! I bet that Columbia studios didn't know either when they bought it brand new!!! Double dang!!! We're all so stupid.
(Don't worry, I'm just having fun with this - that is the actual used price paid though. Can you imagine how many crappy sounding plug-ins I could have bought for 40 grand??? All of them!!!! Again, just having fun - I actually love my UAD card)

[alphajerk]

yeah but that was 93... although most of that is probably in the automation package really. bag the plugins.... you could of bought a hummer and all the gas for it would be like the incessant computer upgrades every week it seems



zandurian, there is no "report"... it was merely some tests that people have been doing and others have been participating in. its amusing really. honestly the difference between the two were fairly small and probably 98% of the population wouldnt know the difference between either, so how is that for ya? 90% of 2% of the population... of which only 30-50 or so people would "guess" each time.


don T, sounds like your convertors are broken with less than 16bits of dynamic range to them. do you have a link to that thread? how did you test them?


i dont know how much pink or white noise is goign to tell you. there are also impulse tests to consider as well. system recovery etc.