For all the ITB mixers a MUST read

[csiaudio]

http://www.gearslutz.com/board/so-much- ... mixes.html

[James Steele]

I'll read this someday. Perhaps in the future if I have the money for big buck consoles and external summing amps, blah, blah, blah, I'll read this. For my money, the extra tiny boost in difference in the sound can't be justified since the average consumer has a pretty much tin ear anyway and they'll be listening to it on a crap speakers with lots of ambient noise. I've heard ITB mixes that sound darn good. My friend Jim Watson mixes ITB, and has mixed some of my stuff and the mixes have drawn raves from the "average" listener.

For all I know the poster over there may be right, but for now this doesn't apply to the word that I must live in.

[Phil O]

The OP says,
"Running a Digital mix right to the top of the scale is like running your SSL mix buss where the VU meters are slammed all the way to the right and you are constantly hitting it at +25. No one will get a good sounding running the desk like that. You won’t get a good sounding mix in digital either."

That's not exactly true, and yet that seems to be what his argument is based on. Although I agree that you shouldn't run a digital mix right up to full scale, the reasons are not the same as for analog. This is an old argument that doesn't hold water. The nonlinearities that exist on an analog mixer as you approach full scale don't exist in the digital domain. The digital concerns as you approach full scale in a DAW don't exist on an analog mixer. Yada, yada, yada. You got the idea.

It is quite possible that it's those nonlinearities that color the sound in a pleasant way and result in what some believe is a better sound.

Bottom line. There are some great mixes out there that have been done ITB. It's a matter of learning that it ain't analog, and some of the analog tricks you learned won't work. The people who understand this have gotten beyond the rhetoric and just make great mixes.

Phil

[jlaudon]

FWIW, earlier this year I did a project that was mastered by Bob Katz. I mixed a number of the song ITB with DP6. Another very experienced engineer mixed some other songs with very high end outboard stuff using Pyramix (not native, but the card based version), and the top end Adam speakers in a 'proper' studio (he's the Adam dealer here).

I pumped Bob for info about my mixes (what is lacking, etc), and he said they sounded great - done in my simple acoustic treated bedroom in my house (well, I have Adams too, but the very bottom end ones). For the other guy's mixes, he had to do some eq and other tweaks to the overall sound.

It's all so subjective, IMO.

[michkhol]

The OP says,
"Running a Digital mix right to the top of the scale is like running your SSL mix buss where the VU meters are slammed all the way to the right and you are constantly hitting it at +25. No one will get a good sounding running the desk like that. You won’t get a good sounding mix in digital either."

That's not exactly true, and yet that seems to be what his argument is based on.

Actually it is completely false and disorienting. That guy teaches students but after all right words he says a complete BS which reveals the true level of his (in)competency in digital processing. Digital is mathematically perfect. Floating point processing has the headroom any analog console could only dream about. The digital noise (a.k.a rounding errors) is determined only by the sample size and even at 16 bits it is lower than the console noise floor. You can mix ITB as hot as you like, the problem begins only in the analog domain when it hits your DAC. Just make sure your final output does not exceed 0dBFS and you will be fine.
One caveat: if you print your stems in the integer (16 or 24 or whatever) format, the rule is the same.

[Kubi]

The thread is actually quite useful (and has nothing to do with extolling the virtues of analog gear, James.) It's merely about gain management, which many people working ITB still approach as if they were using analog gear (that is, trying to maximize levels at every step. Remember the "normalize" threads?)

And michkhol, while you are correct that floating point has virtually endless headroom, in reality there are a number of plug-ins that for a variety of reasons don't, i.e. because they are 'modeling' analog behavior. In the real-life situation of a mix, you may be going out to analog processing and back in, or you may overlook a plug that does overdrive, so limitations of i.e. analog outputs or plug-ins that do distort when driven hard can come into play.

All the OP of the gearslutz thread is advocating is leaving plenty of headroom when in digital, since for the same reasons that you have almost infinite headroom in floating point, you also have no reason to hit "the tape" hard - there's plenty of resolution to leave yourself 12-18dB headroom.

And I for one think that is excellent advice. First thing I do on every audio or VI instrument track is instantiate a Trim Plug and take the whole thing down by 6-12dB.

[Phil O]

...Actually it is completely false and disorienting...

Agreed. I was trying to be polite. I got into a rant a little while back with another poster and it got out of hand, so I'm exercising restraint. Yeah, he is a stupid SOB isn't he?

Phil

[Phil O]

All the OP of the gearslutz thread is advocating is leaving plenty of headroom when in digital, since for the same reasons that you have almost infinite headroom in floating point, you also have no reason to hit "the tape" hard - there's plenty of resolution to leave yourself 12-18dB headroom.

I think the point is that although it may be all that he's advocating, it's not all that he said, and much of what he said is false and misleading.

BTW, hitting the tape hard was one of those analog techniques that many try to reproduce with digital plug-ins. In some cases it was considered desirable.

Phil

[Kubi]

Not so sure he actually is wrong - he refers to ProTools, which is not floating point, but 48bit fixed, AFAIK. (PTLE is floating point, but not HD). So it seems quite possible in a 48bit fixed system to overdrive the master buss by too much cumulative level, no?

Be that as it may, trimming levels down across the board is solid advice, for all the reasons listed above.

[Kubi]

(And of course even in a floating point system there is an absolute peak level, in theory. But in practice I'm aware that DP etc. aren't laid out in a way that absolutely maxes out at 0dB and waste all the floating point resolution below that level...)

[Kubi]

BTW, hitting the tape hard was one of those analog techniques that many try to reproduce with digital plug-ins. In some cases it was considered desirable.

Phil

Absolutely! But you want to do that on purpose, not by accident...

[Phil O]

Everything I do is on purpose. Well, almost everything. OK, so I screw up once in a while.

Phil

[Kubi]

Of course I wasn't referring to you personally... it's obvious you know how to handle your gain staging!

[HCMarkus]

Excepting Thanksgiving, I stage my gains fairly evenly over the Holidays.

Back on topic, while I largely agree with michkhol, kubi presents a rational and compelling reason to keep levels in line: Plug Ins, which are generally operating pre-fader, may have a non-linear impact on sound quality/character depdending on input levels. Plus, if your meters are all over the map, the potential for error on output (mix, bounce, etc) is likely increased.

I liked Phil's response the best tho.

[michkhol]

In fact, trimming has nothing to do with digital or analog. The whole purpose of trimming has been to put the signal in the area of maximum resolution of faders and meters. Traditionally that maximum is at 0VU and it corresponds to the optimal signal level in an analog console. If you use floating point, you can map that 0VU to any digital level. If you use integer (like in ProTools HD), the -20dBFS is the safe level for all musical material. Using the Bob Katz' K-System is very illustrative (even without SPL calibration) and shows how relative the digital levels are. Another source of confusion is that digital meters are usually peak meters and analog meters are RMS. So if you map your 0VU to any particular digital level, make sure you use the RMS meter. All that said works ITB. To get it out of the box is a different matter. For that you have to remember that there is no floating point DACs, so there is the hard limit of 0dBFS. If you want to do it "by the book", you output a 1K sine wave at -20dBFS from your DAW to your DAC and set your analog VU meter after the DAC to 0. If you want to be even more pedantic, then you grab your SPL meter and make sure it shows 83dB for each channel by adjusting the monitor output level. This way you will have a fully calibrated system.