question about tracking too hot......

  • Thread starter Thread starter dastrick
  • Start date Start date
On a side note, you know how old people ramble on about subjects... I was wondering what we're going to do when we get old? Wait a minute, we are old! How'd that happen?
Yeah, I remember when moldy bread was a nickel, we got goiter becaise our salt was not iodized, and we had to dig out of 10 feet of snow with a teaspoon, and WE LIKED IT! ;)

I'll be the first to admit that, even though by today's standards I'm still pretty young, that I'm turning into a Grumpy Old Man in any ways. I'm trying to fight that every day. I admit that I don't always win that fight. At the same time, there comes a time when you realize that life is just too damn short to put up with Net2.0 bullshit for any longer than you have to.

G.
 
there are plenty of reasons not to map 0dBVU to -50dBFS. This is mostly because the A/D converter does not have a dynamic range of 144 dB even if it is 24 bit (there are no 32 bit converters). The dynamic range is typically around 100dB for the good converters. mapping 0dBU to -50dBFS would limit the dynamic range in the nominal range to about 50dB, which would make it less than the analog front end.
FIrst, I'm asking about 0VU, not 0dBu. Maybe that was just a typo. No matter, it still holds true that 100dB converter range would map 100 dB of range to the canvas no matter what. Considering the average analog signal will yield *at the very best* maybe 70dB of range, you've got 30dB to "spare" (so to speak) no matter what.

(As a side observation, what rule is there that says it even needs to map all the way up to 0dBFS? if it's a 100dB converter mapping to a 140dB canvas, why not map to the -20 to -120dBFS part of the canvas, calibrating 0VU to -50? I'm not suggesting that, I'm just saying that there are reasons to all this stuff, and that the numbers are not just picked out of a hat.)

The question remains, why the tight rage of -14 to -20dB (more or less) as *defined standards* for 0VU and not something else? You're heading down the right path, boz, but you're not quite there yet. Keep going. Think it through.

G.
 
Last edited:
Glen

Not trying to be insulting, but your post #132 isn’t revealing any earth-shattering info…or asking anything very complicated.
You’re setting up a scenario that focuses on a total newbie at the start with either a #1 very bad tracking SOP or a #2 more “safe” tracking SOP.

SO WHAT???

If this thread is all about how an inexperience newbie should proceed…then fine, let them follow the “safe” method without having to ever “think” about what/why they are doing…but like I already said, that smacks of a “preset” mentality (something you always tend to argue against ;) ).

All your points are 100% valid when only aimed at how a newbie should precede in order to be “safe” during tracking. But that doesn’t answer the broader question of why HOT BUT GOOD analog levels should ever be adjusted/turned down if the converter is not clipping…???

Easier mixing…sorry, not compelling enough. It takes about 10 minutes to adjust levels if needed…and you know what, if they are all on the same “hot side”…there no adjustment needed, only if you have some real hot and some real low.
You just set your monitors where they sound comfortable.

Newbie mixing mistakes….like I said, this thread isn’t just trying to answer Newbie Tracking 101 and how they can be “safe”.

Gold VS turds...ahhhh, that implies that hot levels always = turds…and THAT is certainly NOT the case.
 
This isn't the DHS. You guys should be able to get off just measuring the individual local test dots and connect the dots of the big picture by yourselves. Since that seems to be asking too much, let me paint some pictures for you.

I wish to pose four different potential real-world studio situations for consideration. In each, there is a studio operator operating a typical digital mixing station (computer-based or DAW, it doesn't matter). No theoretical laboratory - no oscilloscopes, no calibrated microphone hooked to audio frequency response analyzers, none of that fun stuff that one finds in an anechoic chamber or laboratory, but virtually never finds in an actual real-world studio, whether home or Big Box:

1.) The operator is an inexperienced newb with little-to-no knowledge of gain structure gaming other than the mythical knowledge that you should use all the bits you can without clipping, and he is given a project comprising 10 instrument/vocal tracks each of them tracked to the highest possible RMS values available without causing peak values to clip. Such values were attained by gaming the gain structure to push all equipment and converters hard under the theory of keeping the S/N as high as possible while at the same time avoiding clipping in the converter.

2.) The same rookie operator in #1 is given the same project, except he is given no preconceived notions about making things hot without cliping, and the tracked levels tend to be a bit lower, with tracked RMS values somewhere in the upper-teens to lower twenties negative RMS and peak values typically ranging several decibels below clipping. Such values were attained by paying attention to the VU readings on the analog side and - unless some kind of tube or tape saturation were purposely desired at one stage or another - the signal average was never pushed too hard past 0VU, and such nominal signals were kept that way when pushed through the converter.

3.) It's the same situation as #1, except you replace the operator with a seasoned, experienced engineer who will do what is necessary to make the mix sound as good as possible.

4.) The same experienced engineer is given the project from #2, with the more nominal track levels, and will do what is necessary to make the mix sound as good as possible.


Two questions: In which situation are we most likely to find the best sounding resulting mix?
and
Which situation will require the least amount of work and signal management on the part of the operator to do so?

Not theoretically; theory and test measurement are only part of the whole equation. In which of these real-life situations are those results most likely to happen?

G.

I would propose that 3 and 4 are identical, unless you are making an assumption that the analog front end is doing something to the sound in 3. From my experience, it is not any more or less work to mix a song that is tracked near 0dBFS. There are so many volume adjustments made in the mixing process that it is no hardy for me to tell a plugin to turn the signal down by 10dB than it is to tell it to turn it up by 3 dB. Or, even easier, just turn the master fader down by 10dB and let everything else go over 0dB. We're in floating point land, so it doesn't matter if it goes over.

There are some rare situations where a plugin or a system can't handle this, but even all the freeby vst plugins I've used can handle signals over 0dBFS without any problems. I believe that tdm systems my have an issue because the interface to the tdm plugins are 24 bit, so you do have to worry about headroom. I don't use TDM nor protools, so I don't care all that much about that problem.

There may be some plugins that really want the signal to be around -20dB. I'm making one right now myself. In order for it to work as intended, it has to assume certain things about the level of the signal coming in, and even the level of the audio coming out the speakers. In order for it to work correctly, it needs to assume that the playback volume is set to 85dBSPL. I think that's a ridiculous volume to have when mixing, but it's been deemed the standard, so that's what I'm going by. It'll be easy to change if I have to.

As far as the input level goes, that part will be easy. It's very easy to detect the average input level of a signal and apply the non-linear processing based on that. Any plugin that doesn't work correctly when when the signal is up around 0dBFS is a design fault of the plugin developer, not the user.
 
Not at all. A 100dB converter range would map 100 dB of range to the canvas no matter what. Considering the average analog signal will yield *at the very best* maybe 70dB of range, you've got 30dB to "spare" (so to speak) no matter what.

(As a side observation, what rule is there that says it even needs to map all the way up to 0dBFS? if it's a 16-bit converter mapping to a 24-bit canvas, why not map to the -20 to -120dBFS part of the canvas, calibrating 0VU to -50? I'm not suggesting that, I'm just saying that there are reasons to all this stuff, and that the numbers are not just picked out of a hat.)

The question remains, why the tight rage of -14 to -20dB (more or less) as *defined standards* for 0VU and not something else? You're heading down the right path, boz, but you're not quite there yet. Keep going. Think it through.

G.

I think we can ignore mapping to 32 bits for this, because that part is irrelevant. whether 0xEFFFFF maps to "0dBFS" in 32bit mode doesn't matter because the data is not stored in 32bits, it's stored in the raw 24 bits that comes from the converter. The reason 0dBFS is chosen as the maximum is because it physically can't do more than that, and it wouldn't make sense to choose anything but that for the max value. Plus, when we send that data back to the D/A converter, we have to know whether or not the data is overflowing.

An A/D converter doesn't have an infinite dynamic range, and it's dynamic range is not constant no matter what it expects the incoming signal to be. If we were working with audio signal levels with much higher voltage levels, it would be easier to design an A/D converter with a higher dynamic range.

Think of it this way, we can easily measure the difference between 10V and 100V, but it is hard to measure the difference between 0.00001V and 0.0001V.

You're heading down the right path, glen, but you're not quite there yet. Keep going. Think it through.
 
and how did this turn into talking about A/D converters anyways? We can assume that the converter itself is linear throughout it's dynamic range. It's the preamp we're conserned about having a "sweet spot."
 
In order for it to work correctly, it needs to assume that the playback volume is set to 85dBSPL. I think that's a ridiculous volume to have when mixing, but it's been deemed the standard, so that's what I'm going by. It'll be easy to change if I have to.

A weighted or C weighted?

I'm sure we can start a debate about that!!! :D


Yeah...I kinda' stick with 85 dBSPL (C weighted)...though I find myself kicking it up to 90 or even 95 on some occasions and often towards the end of longer sessions.
I've tried the "quiet level" test...but I can't really listen to a mix much lower than 75dBSPL for too long. I find myslef "leaning in" too often! :p
 
A weighted or C weighted?

I'm sure we can start a debate about that!!! :D


Yeah...I kinda' stick with 85 dbSPL (C weighted)...though I find myself kicking it up to 90 or even 95 on some occasions and often towards the end of longer sessions.
I've tried the "quiet level" test...but I can't really listen to a mix much lower than 75dBSPL for too long. I find myslef "leaning in" too often! :p

I chose A weighted, but I'll admit I haven't given it much thought yet because it's so far off from what I monitor. I guess I listen to my music pretty quiet. Metal at 85dB hurts my ears. It's a pretty constant pounding on my brain.
 
I guess I listen to my music pretty quiet. Metal at 85dB hurts my ears. It's a pretty constant pounding on my brain.


Well Metal...yeah, I guess I would turn it down too. :)

The 85dB SPL works for me in my room...though it's more like 80 with the peaks hitting the 85 mark.
I get a good response from my Mackie 824s at that setting.

PS

When I click on your music links...I get "IE cannot display the webpage"....???
Even if I just go to http://www.bozrecords.com it's the same error.
 
Not trying to be insulting, but your post #132 isn’t revealing any earth-shattering info…or asking anything very complicated.
That's not insulting at all, miro. That's what I've been trying to say all along. I just don't get why people seem to think that I'm trying to promote some new-age flim-flam or something. This is simple shit...or at least it should be.
If this thread is all about how an inexperience newbie should proceed…then fine
No, no no, you're still missing the point. This as about a systemic approach that makes it easier for EVERYBODY, not *just* the newb.

Ever hear the phrase, "the best mixes are usually the ones that mix themselves"? Virtually all the top engineers in this racket claim the day they had that epiphany as the day they shifted gears on the quality of their work. Now, nobody should take the phrase "mix themselves" that literally, and I surely don't want to be misunderstood as construing that one should track so incredibly specifically as to create self-mixing tracks. But tell me, which tracks are closer to ones that mix themselves, ones that all peak at 0dBFS or ones that leave some room for summing?

Whats the point of cranking the tracking if you're just going to have to turn it down again when mixing, as easy as that may seem to be to do? Doesn't it just make much more sense to keep it at an even keel all the way through the process? In that way the "sweet range" on the analog side does correspond to a "sweet range" on the digital side as well. Less worries, less things to manage, results that require less signal control to achieve and the levels just kind of work themselves out. It's simple, easy and (care I say) organic.
But that doesn’t answer the broader question of why HOT BUT GOOD analog levels should ever be adjusted/turned down if the converter is not clipping…???
I think it was you who asked the question yourself earlier, miro; how did those levels get so hot to begin with? There's no really good reason that I can think of or that anyone else has offered why the levels need to be that hot going into the converter to begin with.

And again, I reiterate: what's the point of sending the levels that high to the converter to begin with, when you know your just going to have to turn them down when it comes time to mix? Isn't it easier to just nip that potential situation in the bud at each track at tracking time than it is to give yourself a mixing project of ten or twenty or more hot tracks that you know you'll just have to turn down at mixing time anyway? It's not unlike letting garbage pile up into a ten foot pile that you have to clean up instead of the much easier task of just throwing away each little piece when you come acoss it.

And yes, this does all make things much "safer" and easier for the newb. It also helps him/her understand the whole confusing mess as one simple integrated signal path. But that's not all. For the pro working a $180/hr room, every little 10 minutes saved is a savings of $30. Over the course of a 10 album CD, that's a good $300 saved. Maybe chump change for Madonna, but for people like you and me that's a lot of bread.
if they are all on the same “hot side”…there no adjustment needed
until you start summing them together and the 2mix starts clipping all over the place. And no, as someone suggested before, just throttling the master bus is not always the solution. Depending upon the mixing desk or DAW, clipping in the mix will not always be alleviated by pulling the master bus down.
that implies that hot levels always = turds…and THAT is certainly NOT the case.
I never said always. But the chances are greater that hot levels will yield more stank at the end for the simple reason that they require more attention and management; i.e. they give a greater chance that an easy oversight or mistake on the part of the operator will be negatively audible than levels that are pre-tamed.

And as far as final mix levels, just ask any mastering engineer here what they would rather receive, an unnecessarily hot mix or a properly mixed on with more "natural" (for lakc of a better word) levels as giving them a better chance for making a better master.

It's a whole bunch of little stuff here and there, but it all adds up right, and adds up to a sum that does make a difference. And since it all involves a very simple and basic way of viewing the entire signal chain that works for all scales of studio size and everyone from the newb to the pro, how can one possibly go wrong with it?

Note that I NEVER said that this is the only way that one can look at it and do it. Just that it is an easy and valid way that, when you understand it, makes a lot of sense and helps both solve and avoid a lot of problems that one might otherwise have. It works. And it works well.

And GOD I am tired of talking about it. :(

G.
 
I'm not going to lie, the only valid argument I've seen on why to track at lower levels is for convenience only. Not a single actual sound quality issue is addressed, other than saying the words "sweet spot" but I haven't seen any evidance other than "trust me" that this sweet spot even exists. I can't think of the last time I saw a plugin clip. The red light shows, but rarely do I find a plugin that actually clips. The plugin developer has to go out of his way to make sure it clips in order for that to happen.

If it's purely convenience, then I don't think it's anything more than newbie recording 101 sop.
 
.though I find myself kicking it up to 90 or even 95 on some occasions and often towards the end of longer sessions.
Word of caution. If you listen above 85 for long periods of time, one of these days you may wake up with this ringing in your ears that won't go away, ever.:(
 
OK Glen…maybe we are all talking extremes for effect. ;)

First off…I don’t know how we got to equating “hot” levels with only/always hitting 0dBFS?
Heck…I think I stated a couple of times that at most, I’m hitting around the +8 maybe +6 on occasion…and you know, if anything, I’ve had to turn UP levels in my DAW on some tracks rather than EVER turn them down!!! :D

Why so “hot”…?
Well, I don’t think that’s REAL hot…and it’s pretty much what my analog front end sends to the converter without me ever “pushing” the output just to get hotter levels into the DAW, thoguh traking to tape may have something to do with it, as I tend to go for a hot tape signal.
I run my front end gear well within it’s working range, though certainly more toward the hot side than the cold side of its so-called “sweet spot”.

AFA as the summing issues…my DAW never clips, even if I BURY the tracks into the red.
I guess for those folks whose DAW will clip…well, they have to adjust for it I guess.
Furthermore…since I mix OTB from the DAW (as a lot of folks do)….that’s not even a consideration for me.

When I record my stereo mix back into the DAW, I let the peaks get up to -2dBFS…and I’m not running anything across my analog stereo mix buss.
Then for the “mastering” stage I do some global EQ and comp/limiting in the digital domain to push the mix level up a bit. I can do that without any issues within my DAW, and I’ve not run into any problems. Only time I noticed anything is when trying to nuke tracks with some crazy straight-line limiting in order to have VERY LOUD final CD levels. But I don’t do that….I only apply some mild “lift” in order to be near the ballpark of commercial levels.


But if this entire thread is focusing only on the people who track everything at-2, -1 or 0 dBFS all the time….well then proceed with your “safe” mantra. :)
 
So now what...

I'll just butt in on this conversation to thank everyone for arguing so much.

It's made me understand gain staging a lot more than I used to. It's not that complicated when you get right down to it. But important!

So I use a Presonus Firebox. I haven't gotten around to buying an external preamp yet, so for now I'm using the internal preamps on the unit.

When I track my acoustic I typically can get the acoustic recorded at around -20 on the DAW meters. That's the loudest I can get before I start introducing noise from the Firebox. I can turn it up a little more and deal with a bit of hiss (which I can deal with fairly well with a bit of gating and EQ). Or keep it to a lower level where I can't audibly hear it at normal listening levels.

So once I've done that, and then recorded the rest of the parts for a given song at around the same levels. What kinds of things can to do for the acoustic to then bring it up out of the mix better? I find that consistently my acoustic tracks don't stand out enough from the mix.

I know that a better preamp will help here, so no need to give me that worthy advice. I'm more interested in techniques to use the equipment that I have now.

Thanks!!
 
Word of caution. If you listen above 85 for long periods of time, one of these days you may wake up with this ringing in your ears that won't go away, ever.:(

"On occasion" for me usually lasts for a minute or two when I want to focus on something in the mix. The 85dB SPL is approved by most for extended listening.
My console has markings for the control room monitors...so I always put the knob back where it should be, and I've checked the SPLs from time to time as I keep my trust Radio Shack SPL meter right there near me.

AFA the ringing in my ears...if I run the music loud...it drowns it out! ;) :D
 
AFA as the OP's original question....yes, he is indirectly/unknowingly talking about the analog side of things...but I think his question WAS actually focused on the DIGITAL operating range.
The only reason for that (and I would have made the same mistake) is because if the only meters he's got are the ones on his DAW, then dBFS is going to be the only reference scale that he has. Naturally, if you don't have analog VU meters in your "studio", you're not going to reference them. You're going to reference whatever is at hand.

Again, just because he was referencing the levels in his DAW, doesn't mean the question was about what happens in his DAW. The question was about TRACKING...TRACKING...TRACKING.

The question was about TRACKING LEVELS.
 
he is NOT some kind of audio dolt.

Thanks Miroslav. It's really sad when people have only insults but express strong opinions anyway. Especially after admitting, "I know less than all of them."

Actually...the initial post was purely talking about the digital side if I recall.

Exactly. If someone gets "bad sound" when they set their digital levels near zero, then something is wrong either there or earlier in the chain. It's really that simple. A professional audio setup should be able to accommodate levels cleanly right up to the point of hard clipping. I agree that it's not necessary to record right up against Digital Zero, especially in a 24-bit system. But there's nothing wrong with recording at those levels either.

The whole notion of a "sweet spot" is incorrect and misguided IMO, and that is what I challenge and hope to clarify.

--Ethan
 
I can't think of the last time I saw a plugin clip. The red light shows, but rarely do I find a plugin that actually clips. The plugin developer has to go out of his way to make sure it clips in order for that to happen.

Yes, and in my AES Audio Myths video I offered irrefutable proof in the form of a null test. I wonder how many here have bothered to watch the video. It explains - and proves! - a lot of this stuff.

the only valid argument I've seen on why to track at lower levels is for convenience only. Not a single actual sound quality issue is addressed, other than saying the words "sweet spot" but I haven't seen any evidance other than "trust me" that this sweet spot even exists.

Here's a bit more to ponder on the concept of a sweet spot:

All large signals contain smaller signals. If you record a clean guitar at, say -10 dB on your DAW meters, that -10 signal contains the fundamental frequency, plus many harmonics at various lower levels. (Actually, the fundamental of an electric guitar is often softer than the harmonics, but that's a different discussion.) But even with a single frequency, the portions of each wave near zero volts are at a softer level! If a device "favors" some levels more than others, that manifests as basic distortion.

--Ethan
 
Back
Top