Do Your In-The-Box Mixes Bottleneck??

undermind

Member
I just got done reading an interview with Terry Howard in the recent June issue of Electronic Musician. He's spent nearly 20 years as Ray Charles' engineer, getting three grammy's along the way.. His recording method of choice now is Sonar. He's got a Pro Tools HD system he "has to have for the labels". But he says that regardless of whether it's Sonar, PT, Nuendo, the mixes will bottleneck on the mixdowns. He says the stereo image starts to collapse when mixing more than 32 channels. So he rough mixes in the box; gets his edits, stereo balance, etc and then takes it to a major studio on a major board. At that point his DAW basically acts as a tape player.

So how many of US do this?? I'm just starting this thread because I'm curious how much of a common practice this is..

I recorded a complete album a couple of years ago that was heavy with panning and volume movement. I greatly prefer mixing and doing my edits in the box. But I have to say I found the same phenomenon on my final mix. There really isn't great stereo imaging, and each instrument really doesn't have it's own space. I thought this had to do with my poor mic and pre and converters, which I've addressed, but now I'm wondering how much was the fact that everything was done in the box.
 
undermind said:
So how many of US do this?? I'm just starting this thread because I'm curious how much of a common practice this is.
The higher the budget, the more common it is, I think.

But I also think that the converters on the way in and on the way out, along with quality digital clock control have a whole lot to say about that also. Though it is part of it, it's not just the summing algorithm that's at issue there, especialy when you start piling dozens of channels on top of each other.

G.
 
Technically speaking, I don't see how what Terry Howard said is possible. :confused: Too bad he didn't offer a technical explaination for what's happening. Other than what his ears percieved. Not that I don't trust Terry, he has made some great recordings.

With that said, I think it may have more to do with comfort and pure control during the mixing process. You have to admit, hardware knobs and buttons feel so much better and powerful. And Terry spent many-a-year tracking and mixing behind an analog setup.
 
Glen mentioned a big part of it: the summing of signals.



Part of the power in a large format production console is its summing amp. Essentially, it does what it says...it's a central location for all your routed signals.

The big problem is that digital algorhythms can't get up to speed with an analog summing amp. The most I have to my credit is just hearing the difference. When I've listened to a mix done at home and the same done on an SSL, the SSL just sat right with me.

However, be that the case, DAWs are a growing thing. So how do you get passed having to own or rent a large console?

Well the big thing now is buying an outboard summing amp. This is kind of like the makeshift cure for that. Lucky for engineers that are fans of proven name brands, more companies are meeting the demands of the DAW engineer. So you'll see API & Neve summing amps out there for those who are fans of the sound.

I'm currently working up a half a million dollar studio that's essentially DAW based, but we're going with the Mixdream summing amp by SPL simply because it's a powerful option, but also flexible. It's not really colored in any definitive way, which is perfect for a large assortment of music.


Whereas we could of gone with a neve 16X2 summing amp, but then you're limited to just the Neve transformer sound....

which oddly enough is not fitting for all styles of music.
 
LeeRosario said:
Glen mentioned a big part of it: the summing of signals.



Part of the power in a large format production console is its summing amp. Essentially, it does what it says...it's a central location for all your routed signals.

The big problem is that digital algorhythms can't get up to speed with an analog summing amp. The most I have to my credit is just hearing the difference. When I've listened to a mix done at home and the same done on an SSL, the SSL just sat right with me.

However, be that the case, DAWs are a growing thing. So how do you get passed having to own or rent a large console?

Well the big thing now is buying an outboard summing amp. This is kind of like the makeshift cure for that. Lucky for engineers that are fans of proven name brands, more companies are meeting the demands of the DAW engineer. So you'll see API & Neve summing amps out there for those who are fans of the sound.

I'm currently working up a half a million dollar studio that's essentially DAW based, but we're going with the Mixdream summing amp by SPL simply because it's a powerful option, but also flexible. It's not really colored in any definitive way, which is perfect for a large assortment of music.


Whereas we could of gone with a neve 16X2 summing amp, but then you're limited to just the Neve transformer sound....

which oddly enough is not fitting for all styles of music.

So, in that case do you do the track levels and automation "in the box". But route each track through a seperate output of your DAW and have all them summed externally? Interesting.... I might have to try this! You could use a mixer for this, instead of a dedicated summing amp? Seeing as a mixer can sum signals, but have even more power with mixer EQ and gain if needed. But I'd probaly set all to unity.
 
SouthSIDE Glen said:
But I also think that the converters on the way in and on the way out, along with quality digital clock control have a whole lot to say about that also. Though it is part of it, it's not just the summing algorithm that's at issue there, especialy when you start piling dozens of channels on top of each other.

I think we can assume Terry Howard is using top of the line everything. So if it is happening to him at 32 tracks, it could be more obvious for others using lesser gear at fewer tracks.

My own experience has been the same, I really don't care for how an all ITB mix sounds. At least of my own music. I've done it, I went all digital for a while and hated it.

Now I pretty much do the same thing as described in the first post. I blow everything out to my mixer and mix analog there. You have to fight noise a bit like in the old days, but overall I find I like the sound a lot better. And it's a lot easier to get a good sound.

The whole thing of stacking dozens of tracks and dozens upon dozens of plugs does degrade the audio in my opinion. It's okay up to a certain point, but then the sound just starts to fall apart.
 
yeah this is common. I notice this phenomenom when summing more than 8 tracks. :eek: reason why??? who knows.

Lee, half a mill and you went with a mixdream? Did you just feel you didn't need a console?? If I started over with 50,000, I'd probbly get a neotek or an oram desk. what's the rest of the setup like?? where'd all that dough go to?? where's the place located?
 
FALKEN said:
yeah this is common. I notice this phenomenom when summing more than 8 tracks. :eek: reason why??? who knows.

Lee, half a mill and you went with a mixdream? Did you just feel you didn't need a console?? If I started over with 50,000, I'd probbly get a neotek or an oram desk. what's the rest of the setup like?? where'd all that dough go to?? where's the place located?
I think he wroking at.....
 
I do 90% of my mixing in the box with anywhere from 20-40 tracks per song and have never noticed anything like he's describing that hasn't been my own fault.

Lets face it, mixing 32 tracks together is damn hard! The more I do it, the more I realize it's all in how good the source material sounds. You need to take mucho time in getting everything tight and streamlined so it sounds good in those big mixes. The stereo image issue has always been due to poor EQ use for me. Once I figured out exactly what I needed and how to get it, my mixes got a much deeper stereo field. Also, don't forget about automation. I usually have every track automated out the ass. And you know what? It makes a huge difference in the end. Very slight volume changes help a lot when you've got five tracks panned left and right each coming and going every 2 seconds.
 
tarnationsauce2 said:
could use a mixer for this, instead of a dedicated summing amp? .


You could, but just as long as you know that there are major differences between summing amps.

So, in that case do you do the track levels and automation "in the box". But route each track through a seperate output of your DAW and have all them summed externally?


Exactly. That would be a little towards the end of your signal flow. But you see, some people are like "whoa dude....16 CHANNELS?! WHY?". Logically manywould think you'd only need 2 for stereo. And honestly that works.

But if you want to route out stereo stems (Drums, vocals, guitars, etc), then at least you have the channels to do it.
 
Last edited:
FALKEN said:
Lee, half a mill and you went with a mixdream? Did you just feel you didn't need a console?? If I started over with 50,000, I'd probbly get a neotek or an oram desk. what's the rest of the setup like?? where'd all that dough go to?? where's the place located?



It's going to be in Melbourne/Palm Bay, but really serving the Brevard area. We're still in the developmental stages. Just barely got finished with the equipment list.


But yeah, the console was an option, but remember, the half a mill also has to go into the construction of the studio, which to a certain extent is more important than the gear itself. I was fighting for a Neve 88R at first, but there where too many other factors going against that. A console like that will easly blow our budget. Plus I'm a name brand/MVP kind of guy. I was really trained on Neve, API, SSL and a little bit of Amek (you don't find those boards anymore), so it's really more about sound and familiarity to me.

But also, I had to look at the long term. The juice it takes to power an SSL9000J is rediculous. I takes not one, not two, but *three* 210volt plugs to power this thing. That's like running 3 washing machines 24/7. For a studio like us, that's going to put us out in the first month.

So in the end I only really had 150k for gear. But beleive me, if you saw the complete list, I think it would give a better scope of the idea behind this studio. I worked closely on that with Marko of DDG. He was formely working at Audio One down in Miami, but he opted to start his own firm.

The descion didn't come light though, I had to work that out with thier knowledge of the industry, but also good business sense. It's called the DAW Design Group (for the sake of networking). They're kind of like a Russ Burger design group, except on a lower scale. I had wanted to keep this a surprise, but I figured a little promo here and there is alright.

http://citywalkstudios.com/

This link is one of the studios they help set up a couple of years ago, except the studio owner himself took care of construction, so it's not a full reflection of what DDG does.


I'll send you the gear list we have as of now via PM if you'd like.
 
Last edited:
I agree wholeheartedly with Al and Lee that a good analog summing amp beats the majority of digital summing algorithms hands down, especially as the number of tracks being summed increases.

The point I was trying to make in my original post, and that I think got a bit mis-interpreted or lost on the fog of thread, was that the advantages of good analog summing can easily be lost or negated if the control over the digital domain and the conversions from digital to analog are not up to snuff as well. As to just how even or uneven that trade-off may be is open for debate, but I believe it to be no less than signifigant for a couple of reasons.

First, much of the "imaging focus" problems in the digital domain happen before the process even gets to the summing amp. Average to below average A/D conversion combined with a lack of master clocking causes much imaging "fuzz" to begin with.

This fuzz will certainly be piled on even more by a poor summing algorithm, of that there is no question. But this fuzz will also be compounded in perception even by the best analog amp simply by virtue of the piling on of the tracking fuzz itself.

Add to that the fact that one has to take yet another 2 conversion steps to analog sum one's digital tracking (D/A to go out to the summing amp, and another A/D to come back to the digital master again), and one better have some top-shelf converters to make that extra-long signal chain worthwhile. Send your stuff out through a Soundblaster to go to a summing amp and then back in through a Soundblaster, and any advantage provided by the analog sum will be be like taking two steps backwards to go one step forward. OK, the Soundbalster example is a bit extreme, but only to make the point that the converters had better be great.

Al, I agree that Terry Howard is probably using top-shelf converters with a quality master clock, which makes it fine for him - and for you, BTW, since I know you are so equipped as well. I was just pointing out that that's exactly why it's OK for him, but for those currently using off-the-rack conversion and flying without synch, that the equation is not so simple, that the ol' "weak link in the signal chain" considerations have to be accounted for as well.

G.
 
Wow. I didn't get an email update that there had been any responses.. I missed quite a bit.

tarnationsauce2 said:
Technically speaking, I don't see how what Terry Howard said is possible. :confused: Too bad he didn't offer a technical explaination for what's happening. Other than what his ears percieved. Not that I don't trust Terry, he has made some great recordings.
tarnation, I compacted quite a bit in regard to the technical discussion from the article. He says that after 32 tracks it starts to sound like it's compressing itself, because you're throwing so much data down a pipe. Before you know it you have to start panning hard L and R to get a stereo image. He goes on to talk about the early days of multitrack recording, before they put buffer amps in each channel. Then he talks about digital, adding noise with each channel and dithering and so on.
The technical discussion is actually somewhat non-technical..

FALKEN said:
yeah this is common. I notice this phenomenom when summing more than 8 tracks. reason why??? who knows.

Sonic Albert said:
I think we can assume Terry Howard is using top of the line everything. So if it is happening to him at 32 tracks, it could be more obvious for others using lesser gear at fewer tracks.

Yep. I have always noticed this to a certain extent, and at much less than 32 tracks.. If I could get away with 32 tracks with no problems, I wouldn't complain!!
 
undermind said:
He says that after 32 tracks it starts to sound like it's compressing itself, because you're throwing so much data down a pipe.

Does anybody really think that adding 32 24 bit numbers is particularly taxing for a modern computer?

He goes on to talk about the early days of multitrack recording, before they put buffer amps in each channel. Then he talks about digital, adding noise with each channel and dithering and so on.

Digital adds noise with each channel? I believe he is really talking about analog there. That's where a large number of channels become problematic in terms of noise. In fact large format consoles tend to be designed to sum in subgroups, such that there isn't a single summing amp.

Look, obviously there is some phenomenon with analog summing that causes people to prefer it, and that's OK. But I don't see any sense in blaming digital for perceived imperfections, because that implies that there is something that will be fixed, and maybe it isn't broken, maybe it just works too well.
 
I have used an exteranal mixer with my DAW for many years, and never liked the way the computer summed the individual channels. I also prefer outboard gear over the effects provided by the software.
 
Back
Top