Summing mixer for digital

  • Thread starter Thread starter joewolensky
  • Start date Start date
J

joewolensky

New member
Well basically I got so tired of mixing in the box, Ive decided to try sending the outputs off my delta 66 to my 12 channel mackie mixer.The only problem I am having is isloating the individual channels. I want it so that say vox are on output 1 ,bass 2, guitar 3, drums 4, just an example.Mixing on a comp just sounds so boxey, I really am likin what im hearing so far. :)
thanks
-joe
 
joewolensky said:
Mixing on a comp just sounds so boxey
Click to expand...
It's probably because you've got too many sounds building up in the 400-800Hz range.

Analog summing makes sense if you can route all the individual tracks out of individual outputs on the audio interface, which is not really possible for most home setups. Otherwise you'll still be doing submixes.
 
It's going to be pretty tough to do what you want to with the Delta 66, because it only has four outputs. I can see the possibility of taking mono voice, bass and guitar tracks, but drums are certainly going to need to be stereo or more. I'd suspect you'd want to put the guitars on a stereo buss as well.

As far as assigning the tracks to the outputs in your computer, that would be a matter of setting up the interface properly and then assigning the tracks in software. Then they should show up on the right outputs. That's it in theory anyway, I don't use a Delta 66 so I can't help out more specifically than that, sorry.
 
The mix bus in your software likely has much more headroom than your Mackie.

Most of the time people that complain about mixing ITB sounding bad are mixing too hot. DAW software has notoriously bad metering. Just because you don't see any red doesn't mean you're not clipping things.

If your mixes are peaking above -6 dbfs you're likely creating overs on fast transients. This applies to inserts/sends/plugs as well as the mix bus.
 
M.Brane said:
DAW software has notoriously bad metering. Just because you don't see any red doesn't mean you're not clipping things.
Click to expand...
Aaaannnd yet another old myth resurfaces.

G.
 
Well basically I just want to use the mixer to actually well mix the stuff, becuase I really hate mixing on a computer.Im doing my record at a full blown studio,but wish to make accoustic demos at home.Though I could mix on the computer, I would prefer the mixer .Just really simple stuff,plus ya This is more fun, I also might record off the outputs of the mixer to a tape machine,just for kicks :) .
 
M.Brane said:
that happens.
Click to expand...
I don't know how far you expect me to dive into the 14 or more mostly irrelevant pages of that thread to find whatever nugget you want me to see, but I used to work for a company that developed broadcast-level video and audio editing software, and I can tell you that from a development standpoint, getting digital metering and clip detection in DAWs correct is a relative piece of cake, and that there is no good reason for a quality piece of DAW software to get it wrong.

I can also tell you that as of 1999, Sonic Foundry, Avid, Discreet Logic and (I think?) Steinberg all got it right. And this was almost seven years ago. I can't say for certainty on any other brands, as these were the only brands I was familiar with being tested by our engineers (actually they also tested products from Media100 and Pinnacle, but I'm unfamiliar with or don't remember the results of those). But considering the three or four that I do recall are three or four of the biggest names in digital audio and video editing software, I'd have to say that yes, the whole digital software metering thing is, at best, information that has beeen obsolete since the retirement of the 486 CPU, or, at worst, myth that belongs in the same folder as UFOs and Intelligent Design.

This is not to say that there's anything wrong or bad about mixing in the real world vs. mixing in the box. Just that the argument that digital metering is so damn inaccurate as to cause people to regularly ride over the line has been just plain wrong for years now, yet every so often it keeps popping up like a drunken leprechaun in these forums.

G.
 
I'll agree with you on the point that meters inside DAW's aren't "inaccurate," the problem is a lot of people are looking in the wrong spots, or not looking at the bigger picture. People record a track damn near zero and then boost it with EQ, but then bring it back down with a compressor, so that the track meters don't show clipping. Unfortunately, the audio has already clipped at the EQ stage and has just been turned down by the compressor plug. Similarly, if you are metering post fader, as in some master faders, it's easy to think you haven't generated overs because the meter isn't showing red, but again, the mix buss could already be clipped and you just turned it down with the fader, or limiting plug. The audio is still clipped in the sense that there are flat spots where there shouldn't be, but your meters won't necessarily tell you that because the level is below 0.
Back to the original poster, it seems like you are more concerned with mixing outside the box for it's tactile feel than for the (potential) audio benefits. Have you considered a control surface? Behri (:eek: ) makes a really cheap one in the BCF2000 and CM Labs makes the Motor Mix. That will free you from having to use the mouse to do your mixing.
As far as analog summing with regards to quality goes, well IMO whatever benefits you'd receive would be (greatly) outweighed by the extra A/D/A cycles, having to create multiple submixs/stems, noise generated by the mackie (depending on which mackie you have, you may still be going through the preamps, even if you plug into the line in, which adds even more coloration)
 
reshp1 said:
I'll agree with you on the point that meters inside DAW's aren't "inaccurate," the problem is a lot of people are looking in the wrong spots, or not looking at the bigger picture. People record a track damn near zero and then boost it with EQ, but then bring it back down with a compressor, so that the track meters don't show clipping. Unfortunately, the audio has already clipped at the EQ stage and has just been turned down by the compressor plug. Similarly, if you are metering post fader, as in some master faders, it's easy to think you haven't generated overs because the meter isn't showing red, but again, the mix buss could already be clipped and you just turned it down with the fader, or limiting plug. The audio is still clipped in the sense that there are flat spots where there shouldn't be, but your meters won't necessarily tell you that because the level is below 0.
Click to expand...

That shouldn't be the case in a properly implemented 32-bit float system, meaning that there is no truncation between plugs.
 
mshilarious said:
That shouldn't be the case in a properly implemented 32-bit float system, meaning that there is no truncation between plugs.
Click to expand...

I'm not sure I follow, but I have heard the distortion when overloading one plug, but the track meter showed nothing because the next plug in the chain knocked down the signal. I guess it depends on the DAW software, my experience is specifically with protools LE.
 
32 bit floating point software can be overloaded into distortion just the same as any other software. Also, the digital metering in the software I use seems to me to be pretty accurate.

I think inexperienced recordists/mixers can tend to push the levels up too high, causing overloading. I've seen songs where every track was recorded as close to "0" as possible, and then the volume level pushed up from there. once you get a bunch of tracks like this mixed together it's going to overload the software. It will also overload any digital hardware downstream.

The whole "loudness wars" thing has had an extremely detrimental effect on the quality of recorded audio, in my opinion. It went from getting the master as hot as possible to now people recording every track as hot as possible. And the worst part is that many think now believe that is simply the way recording is done, business as usual.

Much better sounding tracks can be had if *peaks* never go above -10 to -6 in the recording process. Keep it above say around -26 to -20 on the low volume side and you'll have plenty to work with come mixing time.

The whole issue of metering, 32 bit float. etc, becomes almost completely moot if proper levels are observed when tracking and mixing. *No* system should overload if used correctly.
 
SouthSIDE Glen said:
I don't know how far you expect me to dive into the 14 or more mostly irrelevant pages of that thread to find whatever nugget you want me to see, but I used to work for a company that developed broadcast-level video and audio editing software, and I can tell you that from a development standpoint, getting digital metering and clip detection in DAWs correct is a relative piece of cake, and that there is no good reason for a quality piece of DAW software to get it wrong.

I can also tell you that as of 1999, Sonic Foundry, Avid, Discreet Logic and (I think?) Steinberg all got it right. And this was almost seven years ago. I can't say for certainty on any other brands, as these were the only brands I was familiar with being tested by our engineers (actually they also tested products from Media100 and Pinnacle, but I'm unfamiliar with or don't remember the results of those). But considering the three or four that I do recall are three or four of the biggest names in digital audio and video editing software, I'd have to say that yes, the whole digital software metering thing is, at best, information that has beeen obsolete since the retirement of the 486 CPU, or, at worst, myth that belongs in the same folder as UFOs and Intelligent Design.

This is not to say that there's anything wrong or bad about mixing in the real world vs. mixing in the box. Just that the argument that digital metering is so damn inaccurate as to cause people to regularly ride over the line has been just plain wrong for years now, yet every so often it keeps popping up like a drunken leprechaun in these forums.

G.
Click to expand...

As far as I know none of the apllications you mention (or any other for that matter) use oversampling meters. If you don't feel like reading through that thread I understand completely (I read much faster than I type, and really as far as you're concerned who the fuck am I?), but it's your loss IMHO. There's some really good discussion of this issue in there by people who know what they're talking about. :)

joewolensky:

I agree that mixing in the analog domain with real knobs, and faders is fun. It's also a lot more intuitive for some (like me) who come from the analog world. Just be aware that the sonic results of doing so with a mid to low-end board will likely be inferior to a good ITB mix. Unless you like the "color" (read=distortion) added by the extra conversions, and electronics in the signal path.
 
I read the whole thread, and the issue seemed to be intersample peaks that clipped without actual samples clipping. I understand the theory, but it seems to be an over-the-horizon situation. How often is that going to happen across the space of an entire song without a few sequential samples clipping?
 
SonicAlbert said:
32 bit floating point software can be overloaded into distortion just the same as any other software.
Click to expand...

I don't follow that. Practically speaking, I ran a test in Wavelab with 2 sequential compressors where the first had like a +24dB makeup gain (with almost no compression, I just used the compressors as a handy gain source), and the second -12dB. The resulting file was very badly clipped. Then I repeated, but with the master fader at -12dB. There was no audible distortion in the second file.
 
noisewreck said:
Analog summing makes sense if you can route all the individual tracks out of individual outputs on the audio interface, which is not really possible for most home setups. Otherwise you'll still be doing submixes.
Click to expand...

I would add that it also makes sense if you have high quality DA converters and a fairly high quality professional studio mixer or summing device. Unfortunately, this kind of setup carries with it a high price tag.
 
mshilarious said:
There was no audible distortion in the second file.
Click to expand...

You have to clip for quite a few consecutive samples before the clipping will be audible on a single track.

The cumulative effect of mild clipping on multiple tracks can turn a mix into crap pretty quick though.
 
This is why, in general I leave the master fader alone (at 0dB), and try to mix so that overall it doesn't peak past -3dB. I've noticed that my mixes sounds clearer when doing this.

Good points about having signals clip within a plugin while the channel isn't (presumable because say the fader has been turned down). This is why most plugins have output level controls. It's always a good idea to adjust this control so that the signal isn't "louder" than what the channel's fader is set at, in another words, if a fader is set at -5dB, you don't want your EQ or the like to output at -3dB. I say with current high-bit systems, there is no reason not to be on the safe side.
 
M.Brane said:
You have to clip for quite a few consecutive samples before the clipping will be audible on a single track.

The cumulative effect of mild clipping on multiple tracks can turn a mix into crap pretty quick though.
Click to expand...

This was on a finished mix with peaks at 0dB and RMS around -14dB. If +10dBRMS isn't a quite a few consecutive samples, I don't know what is :eek:

Just for grins, I repeated the experiment, this time with 4 consecutive plugs with +15dB output gain, and the master fader at -60. The resulting file nulls with the original.

Again, it depends on the mix engine of the DAW and the data handling between plugins.
 
Last edited:
You must log in or register to reply here.
Back
Top