question about tracking too hot......

  • Thread starter Thread starter dastrick
  • Start date Start date
Glen,
Hard to know what your main point here is. If you're trying to convert me to the value of standardised pro level voltages, impedances, balanced signals etc between hardware then you're preaching to the converted. Save yourself the trouble.

Tim
 
OMG, I have never seen so much over-thinking about something as simple as recording a microphone preamp that has one gain knob! :eek:

I just checked the input level software switch on my Delta 66 and it does indeed change the gain as you'd expect. Set it to +4 and you need to send a hotter signal than when it's set to -10. Mystery solved. The Delta 66 works as advertised.

Also, a converter is not a car engine! Running a car engine at 9,999 RPM is not a good idea because it stresses the parts and uses a lot of gas. It's not the same as sending -1 dB into a converter. A converter doesn't care what signal you send it. You can even send 10 dB higher than Digital Zero and it will not wear out any faster.

But on some of these subjects I feel like I (and others) must be talking in Pandoran, because some simple concepts that should be pretty easy and obvious to a guy of your education just don't seem to be coming across.

It's not me, it's you. :D

This means that if one is sending a +4dBu signal to the 66 that happens to be RMSing anywhere near 0VU (we won't argue a few dB here or there), which is common in analog chains with good gain structure, that leaves one with a possible crest factor of only 4dB before clipping in the 66's converter.

RMS and crest factor has nothing to do with any of this. All that matters with digital recording is peak level, and keeping peaks below Digital Zero.

Now your argument, if I understand it right, is "What's the big deal? I just throttle the output busses on my 1202 by some 12dB or more, so that the input to the 66 is basically the same voltage level as if I were feeding it a -10dBV signal or less instead, and I have plenty of room before I clip that way." And you're right, that will work. But it's an unnecessary extraneous step that does nothing musical, but only manages the signal, and is made necessary only because of the halfway design of the 66. If you had a true +4dBu interface, you'd never have to do that and the signal levels would flow much more naturally.

The above is totally wrong. I don't throttle anything back. There are no extra steps. A preamp having one gain knob goes directly into the Delta input using the Mackie 1202's direct out. Did you even read the article I linked to in Post #83? Here it is again:

Using a Mixer with a DAW

There is an operational synergy there that is not immediately apparent by looking at the technical specs of just the individual links in the chain alone.

Say what? We're talking about electrical signals, not ladies sharing stories at the gym. :D

--Ethan
 
Also, a converter is not a car engine! Running a car engine at 9,999 RPM is not a good idea because it stresses the parts and uses a lot of gas. It's not the same as sending -1 dB into a converter. A converter doesn't care what signal you send it. You can even send 10 dB higher than Digital Zero and it will not wear out any faster.
I'm not trying to beat a dead horse here, but for the 14,000th time, I'm (okay, "we" - many of us are) not talking about converters. I made that clear several times just my that last post and even made the assumption that the converters were "perfect" -- This is about the preamp before the converters - which are most definitely NOT "perfect" up to the failure point.

(Apologies for the outburst - I don't know how many times how many people can say the same thing and half the readers don't notice it - Not aimed at any particular person or persons - But this has gone in so many directions and I have no idea what sparked something like "don't track too hot" into the Spanish Inquisition - No one expected that)
 
Glen,
Hard to know what your main point here is. If you're trying to convert me to the value of standardised pro level voltages, impedances, balanced signals etc between hardware then you're preaching to the converted. Save yourself the trouble.
Nope, That's not my point at all.

What I'm trying to explain is just too high concept for half the crowd. The other half, OTOH have benefited very well from it. But some folks get so hung up on specifications and localized technical details and gear slutting that they have a hard time looking at the whole process as an overall systemic approach. It winds up sounding like another world, even though it's not.

Hey, whatever works for any given individual, I guess. The fact is, Tim, we are NOT in conflict on most points here, (from what I can tell), we're just looking at the same situation through different glasses and seeing identical things from different perspectives.

I don't know how to resolve that, I'm not sure if there is such a resolution. Frankly I don't know how else to explain it than I already have a couple of times. I really don't feel like repeating myself yet again. I think it's time to move on in that regard.

G.
 
RMS and crest factor has nothing to do with any of this. All that matters with digital recording is peak level, and keeping peaks below Digital Zero.
Forget it Ethan. You just don't understand what we're saying, and never will.

To continue is pointless.

G.
 
for the 14,000th time, I'm (okay, "we" - many of us are) not talking about converters ... This is about the preamp before the converters

Same thing. The analogy with a car engine is not valid because electronics don't wear out faster based on what levels they pass. (Within reason of course.)

Any decent preamp will be clean up to the point of gross distortion. If it's not clean at all levels, then it's not a decent preamp IMO. I suppose if it has mediocre transformers then the transformers could get grungy at high levels before the electronics clip. But who wants a preamp like that?

--Ethan
 
Forget it Ethan. You just don't understand what we're saying, and never will. To continue is pointless.

I agree it's pointless, but only because you continue to say stuff like that. Rather than, for example, explain why RMS and crest factors are relevant with digital recording. This is the second or third time you stated "You don't get it" in lieu of a technical explanation.

--Ethan
 
I agree it's pointless, but only because you continue to say stuff like that. Rather than, for example, explain why RMS and crest factors are relevant with digital recording. This is the second or third time you stated "You don't get it" in lieu of a technical explanation.

--Ethan
I already gave the explanation more than once. It was your comment after that that totally ignored them, because you just can't get off the Rainman-like mantra of "digital levels don't matter as long as you don't clip. Time for Wopner."

My god man, everybody here already stipulated to that a hundred posts ago. Nobody's arguing that. We're going above and beyond that but you refuse to follow.

Against my better instincts, I'll humor you one last time. If your'e sending a signal into a converter at an RMS anywhere near +4dBu (0VU), and that converter clips at +8dBu, unless that signal has a crest factor of about 4dB or less, you're going to clip.

Even if the RMS were as low as -2dBu (-6dBVU) (which you'll find to be probably just a bit on the low side of typical in most studios), that a crest factor of 10dB will cause clipping. That's not much.

This is a perfect example of why you can't ignore the analog side or the analog signal when talking about digital levels and why the converter exchange rate is key. Yes, you're right, as long as you don't clip (and as long as you don't push the signal so ridiculously low that it bumps into the digital floor, but that's highly unlikely as long as one doesn't have a blood alchohol level of of .45), the digital level doesn't matter. But the level and nature of the analog signal going into the converter DOES matter, as does the converter's calibration, whether you want to ignore them or not. They are the key factors that determine just how you need to throttle the upstream gain stages to modify the signal coming into the converter.

Now, can we just please close this thread? We both have myth buster seminars to write or to send to magazines with no editorial staff.

G.
 
Putting aside all the small details folks are now debating…the one thing that’s still peculiar to me is how/why so many folks would (apparently) manage to end up hitting their converters so hot and so close to the point of clipping….???

The only time I’ve managed to do that is when I was either running some sort of signal tests or just adjusting my levels, etc…however, during actual recording or transfers from my tape deck…I have never once gotten that close to 0 dBFS. On my absolute “hottest” tracks, I may see the occasional peek pop up and “lick” the red zone of my digital meters (which still leaves a few dB to spare). But even when I’m running generally “hot” analog signals…I’m only hitting around -8dBFS at the converter…which is why I hardly ever pay much attention to my converter meters, and instead focus on what my analog front end is doing.
My converters take up to +22dBu …so I doubt they will ever shit the bed before my analog front end does.

But that brings me back to my earlier comments…that the possible cause of many over-juiced tracks these days are probably the all-in-one boxes that make it hard to separate out what your pre is doing VS the built-in EQ and/or comp…not to mention the actual converter.
With a separate front end and/or all separate front end pieces (pre/EQ/comp)…you are forced to pay attention to the signal at each step, and while some folks may still screw up their gain staging…it’s harder not to notice when any one piece is being pushed too hard.
If you get the analog front end right…I can’t see how the converter would ever be pushed to it’s limits…???
 
Putting aside all the small details folks are now debating…the one thing that’s still peculiar to me is how/why so many folks would (apparently) manage to end up hitting their converters so hot and so close to the point of clipping….???
Well, converters like the one in the 66 that clip at 4dB over 0VU certainly can be one reason ;).

But I think the main contributing factors are the pitiful lack of VU meters in the average home signal chain, and the myth that still exists even in some user manuals that you need to record as hot as you can - i.e. use all the bits - in order to get the most out of your gear. Now there's a REAL myth that needs REAL busting.

G.
 
If your'e sending a signal into a converter at an RMS anywhere near +4dBu (0VU), and that converter clips at +8dBu, unless that signal has a crest factor of about 4dB or less, you're going to clip.

So the obvious solution is "Don't do that." :D

Again, RMS and crest factor have nothing to do with setting levels in a DAW. If you believe otherwise, please explain being as detailed as possible.

Earlier in this thread someone (sorry, too lazy to go find it) pointed out that all-in-one external sound cards with built-in preamps, like the PreSonus stuff, have only the one gain knob. That is equivalent to my setup with a Mackie 1202 feeding a Delta 66 through the Mackie's direct outs. The only circuitry between the microphone and sound card is a single preamp with a single gain control.

I will ask you a very simple question, and hopefully you'll have a very simple answer: If you were me, and had the gear I have as described in my Using a Mixer with a DAW article, please define the "problem" you believe I have. Then explain what you would change to solve it.

-Ethan
 
Putting aside all the small details folks are now debating…the one thing that’s still peculiar to me is how/why so many folks would (apparently) manage to end up hitting their converters so hot and so close to the point of clipping….???
Well, as I pointed out in my case, the outputs of my K2600 are pretty hot, so it is not that difficult for me to output at levels that saturate my Aardvark. On the other hand, I have a Roland Juno-106, that can't even approach those levels. With the Juno-106, at unity on the Aardvark, I'll be lucky if it peaks around -18dBFS before the synth itself starts to distort.
 
Putting aside all the small details folks are now debating…the one thing that’s still peculiar to me is how/why so many folks would (apparently) manage to end up hitting their converters so hot and so close to the point of clipping….????
Mainly because people have been (wrongly) told that running your levels that hot is the correct way to do it. (by people like Ethan, oddly enough)
 
I suppose if it has mediocre transformers then the transformers could get grungy at high levels before the electronics clip. But who wants a preamp like that?
Anyone who buys a Neve, API, or any other classic preamp or their clones to get that color. Your philosophical disagreement with the design doesn't disqualify them from being "real" preamps.

Again, everything in the world works the way you say it does, except for the stuff that you decide to ignore because it doesn't work that way at all.


Again, RMS and crest factor have nothing to do with setting levels in a DAW.
-Ethan
This entire discussion has nothing to do with the levels in your DAW. It's about setting the levels in the analog world. This is the part that you keep ignoring.
 
Again, RMS and crest factor have nothing to do with setting levels in a DAW. If you believe otherwise, please explain being as detailed as possible.
Now you're just purposely trolling. I'm not giving you the answer a FOURTH time, sir. Three strikes and you're out. Head to the dugout.
Earlier in this thread someone (sorry, too lazy to go find it) pointed out that all-in-one external sound cards with built-in preamps, like the PreSonus stuff, have only the one gain knob.
That was me.
That is equivalent to my setup with a Mackie 1202 feeding a Delta 66 through the Mackie's direct outs.
Weeeellll, not quite, but I get your point in that your treating the channel trims on the 1202 as the input gains to your Delta. OK, we'll run with that.
I will ask you a very simple question, and hopefully you'll have a very simple answer: If you were me, and had the gear I have as described in my Using a Mixer with a DAW article, please define the "problem" you believe I have. Then explain what you would change to solve it.
The simple answer is I would do the exact same thing you're doing.

The only "problem" (your word, not mine) you have is not in how you're using your gear, but in your assumption that what you have as a typical signal path. It's not. Your 1202 is a mixer, yet not being used as one. 90% of your circuitry there is unused, and much of what most studios have to qualify as a "signal" path is more sophisticated than that.

You are basing your position on your little setup there. It works OK for you, but it does not scale across all situations as a proper understanding of signal chain and gain structuring. I'm trying to offer a way of looking at gain structure that applies equally well to all configurations, from The Record Plant control room to the college student in the dorm room, and is completely applicable and scalable and *useful* when that college student graduates and moves into the real world of recording.

Your position also ignores the fact that while digital signal recording level is irrelevant as long as you don't clip and you don't bottom out - apparently the only thing we agree upon - track level *does* have an effect on the ease of creating a proper mix from those recorded tracks without having to reverse every choice one made in tracking level by changing those choices in mixing. What's the point in pushing gain when recording when you just have to pull it back again when mixing? It's just plain silly.

If one pays attention to 0VU as a guide post or reference - which is what is *always* done on the analog side by any engineer worth their salt - and pays equal attention to it as a reference on the digital side (using the converter conversion factor as the translation from the dBu to the dBFS scale), the mixes tend to mix themselves faster, easier and better with less gyrations needed on the part of the operator/engineer. In short, do it right and the levels just kind of work themselves out with a minimum of work and a minimum of bit-shifting or other processing required.

In other words, in that specific regard, it's just like mixing in analog, which is the intention of the overall design of an A/D system. One should not have to learn different rules of play for mixing just because of the type of gear or medium with which they are working. Nor should the simple set of rules they are taught with their small home system be any different than the simple set of rules they learn for a more sophisticated setup.

This knowledge is like gold to the newb. And for the pro, it both makes the difference between a good mix and a better mix, and cuts down on the amount of studio time needed to do so.

You like flashing your articles around so much, I suggest you check out my mini-app called "Metering and Gain Structure" on my website, available by clicking on my signature. It covers pretty much the entire subject, from theory to application, from microphone to HDR, and is as applicable for your signal chain as it is for a half-million dollar control room, and everything in-between.

G.
 
track level *does* have an effect on the ease of creating a proper mix from those recorded tracks

This is the crux of our disagreement. You have no evidence to back that up, and all logic and common sense says it's not true. You agree that the goal is to not record so hot it distorts, or so low you get added noise from the circuits or recording medium. So what exactly do you think varies at different levels between those extremes? Please be very specific. "Easier to blend" doesn't count, especially because it's entirely anecdotal and can't be proven or backed up with logic. "Because I say so" doesn't count either, nor does "I already explained this five times."

You like flashing your articles around so much, I suggest you check out my mini-app called "Metering and Gain Structure" on my website, available by clicking on my signature. It covers pretty much the entire subject, from theory to application, from microphone to HDR, and is as applicable for your signal chain as it is for a half-million dollar control room, and everything in-between.

That's a very nice treatise. But it doesn't address the subject of this thread, which is tracking too hot. That is the "myth" that needs busting, and all this other stuff is superfluous IMO. Yes, of course gain staging matters, but there's a very wide range of acceptable levels. For mic'd sources, the noise on a given track is usually dictated by the ambient noise level in the room. Many sources are lucky to have 70 dB total dynamic range, whereas most gear is way over 100 dB. That's 30 or more dB of allowable slop. So while understanding the theory of gain staging is of course important to avoid noise and distortion, I'm not convinced it has some sort of magical affect on the ease of mixing tracks. That is the point you need to prove. Either using a logical explanation, or sample wave files comprising a proper controlled test.

--Ethan
 
This entire discussion has nothing to do with the levels in your DAW. It's about setting the levels in the analog world. This is the part that you keep ignoring.

Not so. Here's the original post that clearly defines what this discussion is about:

Let's say that I've tracked to -12dbFS, then in my DAW I increase the gain by 10db (not moving the fader, but processing the track to increase the gain). Is that the same as tracking at -2dbFS?

--Ethan
 
I'm not sure if this fits here but...

Is there anything to be said for the logarithmic nature of a fader? I was under the assumption that mixing with your faders closer to unity would allow for more increments of gain and a more precise adjustment of level. If you have to reduce the gain on every recorded track to prevent overloading your stereo bus then your not optimizing the use of your faders. Level adjustment is much more coarse at the lower end of the fader. Maybe this doesn't transfer well to digital because you can just type in a number...?

Just a thought
 
Back
Top