The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
My Theory About Prosumer Audio Audio Interfaces
Old 21st June 2006
  #91
Quote:
Originally Posted by Lawrence
I have to disagree with the guy who's here saying "tracking as close to digital 0 or full scale as you can" gives better sound. In my experience that is simply not true.
Hey Lawarence....

Just a quick FYI.

"The guy" you are referring to above is the same "guy" who designed the The Distressor (AKA The Distressor's "daddy") and the "chief cook and bottle washer at Empirical Labs." He knows a lot more about audio than most of the good folks here at Gearsluts combined and his opinion deserves a ton of respect. His word is worth more than 50 anonymous posters like you or me.... it's all good just wanted to point out that his opinion is just as valid, maybe a little more so.

Old 21st June 2006
  #92
Quote:
Originally Posted by Dave Derr
Sheesh I guess I'm done here. If a converter's manual says to leave 10dB of headroom or it will clip, then they are doing something very strange. But Im pretty much done repeating myself over how to get all the bits of resolution out of your converter. That is simply how converters work. To get 20 bits of resolution, you must go all the way to full scale on a 20 bit converter. If people find some sound better at a lower level, thats fine, just realize that you ARE NOT GETTING THE RESOLUTION SPECIFIED BY THE CONVERTER. THAT'S AS UNDENIABLE AS THE LAW OF GRAVITY.

Ill also repeat that recording down 6 to 10dB below clipping will not ruin any normal recording made on a modern converter, but if the tonality changes going to full scale, there is something wrong somewhere. Perhaps some converter boxes are poorly designed and can't actually handle full scale cleanly, but I certainly wouldn't put my name on any converter like that!

Unlike analog tape, converters are digitally filtered with "non-dynamic" filters whose frequency and tonal response are constant at ALL AMPLITUDES. Lastly, I suppose its also possible that when a mastering engineer "normalizes" a low amplitude mix, the normalizing process changes the frequency response... but that is also REALLY SCARY TO ME!

Use your ears when you are recording, but trust me when I say that if a converter is working with -6dBFS (or even lower) peaks, it is not working at it's designated resolution. Like buckling up... IT"S THE LAW!
Hey Dave.

As I said above I am not knocking you or anything and I agree in principal with what you are saying.

It's just that it seems like in practice things are different. Just wondering if you have any explanation? Also we might (maybe? I don't know) be talking about different things here perhaps? Maybe the issue is not with the conversion process and the "better audio" is coming from trimming the DAW faders? No idea but it just seems strange that many people report better results with lower levels in the DAW.

Any ideas?
Old 21st June 2006
  #93
Lives for gear
 

Quote:
Originally Posted by Dave Derr
Sheesh I guess I'm done here. If a converter's manual says to leave 10dB of headroom or it will clip, then they are doing something very strange.
Interesting what you just did... you took what I said and added something I didn't say.

First of all this was not "a converter's manual" this was a "digital console's manual". I used the -10 reference even when using external converters. Tracking into the d8b with external converters at -2 (or tracks that hot already recorded elsewhere) also sounded harsher. I don't know why. Trimming pre-recorded digital tracks a few db before sending then into the desk sounded better.

Here's what I actually said...

Mackie (for the d8b) set the recommended peak level (per the user manual) at around -10 for tracking. I never said (nor did they ever say) anything about signals "clipping over -10", you made that leap all by yourself. Audio clips in the d8b at digital 0 like most other digital desk or daw.

However... (this is the entire point that you seem to be missing completely)

They knew (from testing I presume and from real world use) that for whatever reason, the console sounds better in a real world production / mixing environment (not the theoretical lab you talk about) with lower input levels. It does. I can verify that after 7-8 years of use. It doesn't "clip" when you go to -2. It also doesn't sound as good.

I really don't care _why_ that's true, but it is in fact true. I'm just glad I knew it so I can make it sound "better" when I had to use it. I made some great rrecordings on it doing just that.

Dave... my new found friend...,

Even though some people here keep telling you they hear the same thing in their daws, you keep talking about math and converter design theory and such things. You keep implying (with that theory) that we can't be hearing what we hear.

Quote:
Like buckling up... IT"S THE LAW!
It's entirely possible that you're theory is correct and something else is causing this ??? I'm not talking an immediately obvious difference between two solo'ed bass tracks recorded at different levels (although I have heard a tiny bit of that also on vocals) I'm talking about how it sounds summed out of the daw as a whole with plugs, groups, busses etc or even without any plugs or groups etc. It does "sound" better.

So you could be dead on right about converters (I actually think you are dead right) and still "tracking near 0" could still be a bad thing due to another reason that's not been identified here.

Or I could simply be imagining what I hear. See. I'm being fair. I'm not trying to prove you wrong. I'm just saying my experiences don't match up with your theory,... that's the "up near 0" recording is good theory, not the "how converters are designed" and what they do theory. I buy the latter 100%. Something IS going on though and I just don't know what, it may not have anything to do with converters at all. ??

The real question is (not an argument but a question that's relevant here) ...

Have you actually done it? Recorded and mixed multiple real world sessions with lower peak levels at the input and master bus stages and compared them to previous work where you "tried to get as close to digital 0 as possible"?

I've done the "close to 0 thing" and the "conservative level" thing.

It's all subjective I suppose but you can't _reasonably_ dismiss it until you actually try it. That's what you appear to be doing, saying that "...because of the way converters are designed and the math etc, etc, that can't be true...". Well, in my case, and many others, it is true. Sorry.

Until you actually do it you're talking theory... which is fine... we all do that. But I'm not talking about what _should_ happen according to theory, I'm talking about what _has_ happened during use. Multiple times over and over. Sorry. Should I just ignore that experience and not pass it on to other users who may find the same to be true? Or not.

You may have done that already and found it 'worse' for you than "shooting for 0". If so we'll just agree to disagree, so much in this business is subjective anyway. But again ... have you? You never did say.

Just try it for a week. Come back and tell me what you think. Even if we disagree we can stil be friends.

Lawrence
Old 21st June 2006
  #94
Lives for gear
 

i guess all it takes to really verify this is the ability to turn the volume nob a little higher on playback. shouldn't be too hard...

i have recently been experimently with this "lower tracking levels" deal. no firm conclusion with the lavry blue as A/D. i keep inadvertantly hitting higher just out of habit.

i want to try a little something this week trying to be really disciplined and only peaking at -10.

maybe some desks and converters are more forgiving of higher (or much lower) levels.
Old 21st June 2006
  #95
Lives for gear
 
Nordenstam's Avatar
 

An identical claim about digital audio and resolution was recently put forth by Roger Nichols in an Sound on Sound article. It raised a stir among the gurus in this REC forum thread.

Much respect for the distressors! But the idea of tracking near zero to gain resolution is simply not correct. It's sorta intuitive if looking at the waveform as a pixelated digital picture. It's not. It's a waveform. Tracking louder will only help capture the analogue noise in finer detail. (with a decent converter and real world gear).

Since I'm an anonymous poster(Andreas Nordenstam, Bergen, Norway) and it needs to be 50 posts like this to make an impression, I'll have to qoute Dan Lavry from the above linked thread. (math-haters may skip this!):

Quote:
The first statement (as posted) was about having 256 more accuracy with 24 bits (then 16 bits). That one is correct IN THEORY. Each additional bit is a factor of 2 improvement so with 8 bits you have 2*2*2*2*2*2*2*2 = 256. From an ear stand point, each bit is 6dB additional improvement, so 8 more bits will improve the dynamic range by 48dB.

But first, even in theory, note that the improvement is about fine detail BELOW the 96dB range offered by a 16 bit format. In other words, a perfect 16 bits yields 0.001526% accuracy so the additional bits will improve on that.

Second, we can talk about 24 bits all day long, but there is no converter that will yield real 24 bits. The lowest bits are noise. In fact, take a mic, any mic. Take a mic-pre, any mic pre. Set the mic pre gain to say 30-40dB. You now have enough noise to burry the top 5-6 bits with noise making them useless. Your real world statement becomes: My 20 bit AD is receiving enough noise to make it function as an 18 bit AD (or much less), so I have a 4 times improvement over a 16 bits machine, that is 12dB more accuracy.

Try this: run your typical recording chain without running/mic'ing the input source. Check the level of the signal noise floor captured on a digital meter. Turn down the input til the noise floor stops decreasing. This level of steady noise is the inherent capture floor of your converter.

Turn the input back up a tad to bring the signal noise above the converter noise. Record! If it peaks at -30, fine, you have 30dB of headroom(and 90dB of signal range with a 20bit/120dB converter). Turning the level up to make the peaks higher will only result in a better capture of the noise floor.


Best wishes,

Andreas N
Old 21st June 2006
  #96
Lives for gear
 

Quote:
Originally Posted by not_so_new
Hey Lawarence....

Just a quick FYI.

"The guy" you are referring to above is the same "guy" who designed the The Distressor (AKA The Distressor's "daddy") and the "chief cook and bottle washer at Empirical Labs." He knows a lot more about audio than most of the good folks here at Gearsluts combined and his opinion deserves a ton of respect. His word is worth more than 50 anonymous posters like you or me.... it's all good just wanted to point out that his opinion is just as valid, maybe a little more so.

First of all... I'm not an anonymous poster. Lawrence is my real first name. Being wary of identity theft and such I don't typically use my full name on net boards, though I do at one board I won't name here. Anyway, that's me. Lawrence (x).

Secondly, I don't think I "disrespected" him. I admire what he's done and the reason I called him "the guy" in my first post is because I was doing two things at once (at work) and didn't take the time to go back and locate his name. In retrospect I probably could have easily located and used "Dave" instead of calling him "the guy". I don't think he felt disrepected by that. I hope not.

I had read the posts a few minutes earlier. He has now, and has always had, my respect as a fellow professional, which he obviously was with his knowledge of converter design and such. Even though at the time I had no idea at the time who he "was" in the industry. However I still disagree, generally, with what he's saying.

I'm a regular guy (who's done nothing in the industry to speak of) who's been engineering for over 20 years. Sorry, but I don't agree with people just because of who they are. Especially if the subject is in direct conflict with my personal experiences.

As I said to Dave (at least twice I think) if he's recorded (or if he will record) a few sessions and do mixes using both methods and finds my position to be untrue we'll just agree to disagree. Otherwise he's dimissing my real world observations out of hand.

No big deal.

Lawrence
Old 21st June 2006
  #97
Lives for gear
 

The way I see it:

If we don't track at close to 0dBFS, we aren't maing full use of the available bits, and therefore getting less resolution.

However - 24 bits has so much more headroom than we actually need in reality, we can afford to waste a lot of resolution - it's not the problem.

As Dave said: "Perhaps some converter boxes are poorly designed and can't actually handle full scale cleanly, but I certainly wouldn't put my name on any converter like that!"

I think that is the real problem - cheap converters aren't as good as the amazingly expensive converters, therefore they fall far short of the theoretical expectations.

In theory, a pure synthesised sine wave should sound good right up to 0dBF - and with good converters (like my DAC-1) this seems to be the case. But with cheaper converters (like my M-Audio 2496) you can clearly hear the sound gaining harmonics as it gets closer to 0dBF. (Very different from the hard clipping once 0dbFS is exceeded).

I have no doubt that many of the A/D converters being used for tracking aren't perfect either, and therefore tend to alter the waveform in this way if they get too close to 0dBFS.

So I guess it's a balancing act between the lesser of the two evils - and it will depend on the gear that you have.
Old 22nd June 2006
  #98
Hey Lawrence.

It's all good, just want to clear up a point or two. Not flaming...

Quote:
Originally Posted by Lawrence
First of all... I'm not an anonymous poster. Lawrence is my real first name. Being wary of identity theft and such I don't typically use my full name on net boards, though I do at one board I won't name here. Anyway, that's me. Lawrence (x).
Anonymous in my reference was not that your name is or is not Lawrence or that my name is or is not Michael (it is by the way, glad to meet you). My point was that in the audio community we, you an I, or at least me, are not as well known or known at all for that matter as the Michael Wageners and the Dave Derr's of the world.

Does that mean everything that Michael or Dave speaks about are to be taken as gold? Hell no but they do have more credibility (and credits) to their names than the 15 year old kid with an M Box in the bedroom of their mommy and daddy's house. Yes there are 15 year old kids on GS giving advice, I will put a little more weight on Dave's words than a 15 year old.

This is not to say that you or I as professionals don't have valid opinions, we do and they are just a valid as David's, it just when you are talking to Michael Wagener or Dave Derr or Bob Olhsson etc. you know who you are talking to.

Quote:
Secondly, I don't think I "disrespected" him. I admire what he's done and the reason I called him "the guy" in my first post is because I was doing two things at once (at work) and didn't take the time to go back and locate his name. In retrospect I probably could have easily located and used "Dave" instead of calling him "the guy". I don't think he felt disrespected by that. I hope not.

I had read the posts a few minutes earlier. He has now, and has always had, my respect as a fellow professional, which he obviously was with his knowledge of converter design and such. Even though at the time I had no idea at the time who he "was" in the industry. However I still disagree, generally, with what he's saying.
No problem with that at all, to disagree with an open mind is to learn, learning is a good thing. My only point was that Dave is probably not just your average "guy" like me.... heh and while I probably am more in your court than his on this matter I would like to hear him out because I know he is not a 15 year old kid in his mom's house.

Quote:
I'm a regular guy (who's done nothing in the industry to speak of)
Nice to know we are in the same club!
Quote:
who's been engineering for over 20 years.
almost that here as well, I'm still with ya on this
Quote:
Sorry, but I don't agree with people just because of who they are. Especially if the subject is in direct conflict with my personal experiences.
Funny enough I am still with you here as well. I don't give a rats ass what Geoff Emerick himself has to say on a subject, if my personal experience is different I am not going to back down from what I believe. Good for us... as long as we keep an open mind so we might be able to learn something from someone else that is.

Quote:
As I said to Dave (at least twice I think) if he's recorded (or if he will record) a few sessions and do mixes using both methods and finds my position to be untrue we'll just agree to disagree. Otherwise he's dimissing my real world observations out of hand.
I agree, as I said I am probably more in your camp than his but I still like to pick his brain if you know what I mean, that is what I am here for after all..... heh

Quote:
No big deal.

Lawrence
It's all good man.... as you said not realy a big deal either way. Peace out!!
Old 22nd June 2006
  #99
Lives for gear
 

Man, we must be twins! Thanks for the kind words. Here is a better explanation of why recording at lower levels may be subjectively "better". Not my words.

Quote:
A doubling of dynamic range equates to 6db. Therefore, each bit in the word contributes 6db of dynamic range. A 16 bit word therefore has a 96db signal to noise ratio, and 24 bit word can express 144db of signal.

In the real world, the audio electronics in the converter provide a higher noise floor than a 24 bit word can represnt, so a good 24 bit converter will give, lets say conservatively, 110db of signal to noise.

This means that if you record your audio with peaks no higher than -14db under Full Scale, you'll still be experiencing a recording with 96db of dynamic range, which is the best any 16 bit CD has every accomplished. If you've heard a "good" CD in your life, then you can stand 96 db S/N, even in the face of perfection.

To make the point even more graphically - this all assumed that your source signal has a dynamic range in excess of 96db too. I would bet you a beer that it isn't even close. There's no tube mic that operates that cleanly. Your studio room has noise higher than that. All your hardware compressors and EQs operate with a much higher noise floor.

If you were very careful, and ended up having a source with 70db of dynamic range (congratulations!) you could record it with peaks at -26dbFS (-26 under full scale) and still have preserved every ounce of dynamic range.

So its obvious that hitting full scale isn't necessary at all - why not preserve some headroom just in case? Why not give your plug ins some headroom? Let's say you do make it just under full scale. No harm in doing that if you don't go over, right?

Well, what do you do when you want to EQ something +2db? Where does that 2db go? Into clipping of course, unless you lower the input level of the plug in, which is going to lose any hypothetical S/N benefit you had preserved anyway.

Even more importantly, when you record this hot, I've got to ask - what did you do to your preamps, and analog chain to get this level? Most converters are set so that 0dbVU = -18dbFS. This means that if you set your preamp so that the output level shows 0 on a Studer 24 track's VU meter, then if you switched the preamp's output into your converter without changing anything, the level would read -18 below full scale.

That means that if you're getting -6 below full scale on your converter, that you're +12db over the 0VU point! Many analog electronics can crap out here, but almost all will sound different at least. Some times it may be "better" but usually, its a small accumulation of distortion that builds into a waxy fog that makes people blame "digital recording" for its pristine playback of their slightly distorted, but "pretty on the meter," tracks.

If you record with levels around your 0 point, some thing like -18dbFS or -14dbFS, depending on how your converter is calibrated, you'll have your analog electronics in their sweet spot, headroom for plug ins and summing, an appropriate analog friendly level if you use analog inserts later in the process, and on a modern 24bit converter with 110db S/N, the ability to faithfully record signals with a dynamic range of over 90db.

And by all means, 0dbVu is no glass ceiling like 0dbFS is. Keeping levels around 0dbVU doesn't mean that peaks won't exceed that by 6 (or more) db. If they do, your ability to record 96db of S/N (if you even have it in the source, and you don't), just like the best CD you ever heard will be preserved, if peaks don't exceed -12dbFS! More if they do.

So all this means is that the noise is SO low in a good modern 24 bit converter that you can keep gobs of headroom for proper interfacing with analog gear, and still get the full 96+db of dynamic range, just 12 to 18 db lower on the meter. Your analog gear will thank you too.
Now again, I record at -6 in the daw and it sounds noticably better overall (at the end of the production and mixing) than getting up near 0 in the daw. I will experiment with using my converter's calibrated reference... which I assume is -18. I'll try it on a demo session or two and see what happens.

It's this last paragraph that perfectly describes more of what I hear since I brought my levels down from -1/-2 trying to "maximize the bits" some years ago to -6 peaks (lower average levels obviously)...

Quote:
So all of this results in a pristine, beatiful, airy, detailed 24 bit mix with peaks around -12dbFS? Cool! Just bump up the output of the last plug in on the master fader 11.9db (before dither, if used) at the last minute, and you've preserved all your dynamic range when the thing gets truncated to 16 bits. You've also achieved the same 'normalized' or 'loud' result that everyone wants, but you kept your gain staging correct through the process, and only got up the meter right at the end before printing.
Now that makes sense to me, partly because I've already experienced some of it. I couldn't explain it in detail but this guy did perfectly. Mix bus peaking at -12.

This mirrors Bob O's comments about the mix he received at -25, open ballsy and more analog like. Just don't get caught up thinking your individual tracks will somehow audibly degrade by not "filling up" the full resolution.

We're not talking about maximizing the performance of the converter by hitting close to full scale, we're talking about maximaizing the performance of the entire daw system and it's many components, including plugins, summing etc.

Lawrence
Old 22nd June 2006
  #100
Lives for gear
 

Quote:
To get 20 bits of resolution, you must go all the way to full scale on a 20 bit converter.
I think a more correct statement would be that you need to go all the way up to full scale and all the way down to -120 dBFS to use all 20 bits of resolution, and that would be on a theoretically perfect 20-bit converter.

However, to get the "best possible" resolution you only need to make sure that there's enough headroom above the loudest point that you don't clip and enough footroom (?) below the quietest part that you don't hear the quantization noise, and with most of the real-world signals we record with today's 24-bit converters there's a fairly wide range over which we can capture our signal with the most accuracy possible. There are certainly reasons not to record too close to either the noise floor or digital zero but if the converter is designed right maximum "resolution" isn't one of them.

-Duardo
Old 22nd June 2006
  #101
Lives for gear
 
kurt's Avatar
Couldn’t say it better.

Very good post Lawrence. Exactly my point of view too.
Old 22nd June 2006
  #102
Hey Lawrence, nice post.

Quote:
Originally Posted by Lawrence
We're not talking about maximizing the performance of the converter by hitting close to full scale, we're talking about maximaizing the performance of the entire daw system and it's many components, including plugins, summing etc.

Lawrence
This is EXACTLY what I was talking about above in one of my posts as well.

It is possible that Dave is completely correct (actually I am pretty sure he is) about bit depth with the converter. I think what we are experiencing over and over here is that we are not as concerned with the levels to the converter as we are with proper gain staging soup to nuts once the digital conversion has taken place.

To get the proper gain staging through the system things just seem to work better with lower levels on the input of the conversion process. Some math supports this some goes against but math be damned it just sounds better to me.

Now as we talked about above, I don't give a rats ass what anyone has to say on the subject, I know what my experience tells me. That sais by the same token there are some people that I value the opinion of pretty highly. People like Dave, Bob O and Slipperman for instance because they have credits to their names and I know they are not 15 year old kids telling me about audio.

When people like Bob and Slipperman make statements about lower levels into the converters.... well.... I listen. When their experience backs up my own then I have to assume I am on to something regardless of what the math says.

thumbsup
Old 22nd June 2006
  #103
Lives for gear
 
Albert's Avatar
 

Quote:
Originally Posted by eligit
i had really thought that people had started to come to the consensus that peaking at -5 on the A/D or even lower during tracking was recommended, not just "ok". then you do not have to pull the digital faders back nearly as much while mixing.
I would consider this consistent with what Dave Derr said. Peaks in the -10 to -5 area would be as close to full scale as practical in a lot of situations. You have to leave room for musicians or singers "headroom" as well. If they put out a sudden burst of level beyond the normal you need space for that.
Old 22nd June 2006
  #104
500 series nutjob
 
pan60's Avatar
 

Quote:
Originally Posted by Albert
I would consider this consistent with what Dave Derr said. Peaks in the -10 to -5 area would be as close to full scale as practical in a lot of situations. You have to leave room for musicians or singers "headroom" as well. If they put out a sudden burst of level beyond the normal you need space for that.
that was why i original started just going for the -10.
if i get closer to 0 it seems their will inevitably be a over and i just can not stand the digital distortion.
this discussion is causing me to look at this in more depth.
the digital world is new to me and the learning curve has been steep, guys like Cave and countless others have been a tremendous help, thanks.
i want the best possible sound, and then a bit more.
Old 22nd June 2006
  #105
The Distressor's "daddy"
 
Dave Derr's Avatar
 

Quote:
Originally Posted by Lawrence
We're not talking about maximizing the performance of the converter by hitting close to full scale, we're talking about maximaizing the performance of the entire daw system and it's many components, including plugins, summing etc.
Actually, thats all I was talking about: Maximizing Converters.

If a system is poorly designed and can't handle well recorded full scale signals, that's a whole other story!
Old 22nd June 2006
  #106
Lives for gear
 

I hear you Dave.

It seems that if you gainstage your analog gear properly you'll never actually approach 0dbfs in the daw where that maximum performance is. I thought you were suggesting that people recording into the daw do just that... attempt to approach 0db full scale when you said...

Quote:
Record right up to Full Scale if you can. The digital converter is best there. If you hear distortion or some degradation there, you need to troubleshoot
We all can do that by cranking up the inputs and/or outputs of the analog devices but then they are often working past their reasonable limits. The question is... do we want to do that or should we be doing that? I am in the camp that say's no. I think the "distortion or degradation" you might hear wouldn't be coming necessarily from the converter but from the wacky gainstaging to get the signal that hot into the converter in the first place.

As Kyle (the author of that text I posted) stated, this is equal in many cases to hitting an analog recorder at +12dbvu or more. No pro engineer would typically do that (run an analog signal that hot) _except_ to get daw meters running up real high. It's apparently totally unnecesary. Some will still disagree but I for one am convinced.

Why don't they just calibrate converters so that 0dbvu = 6dbfs (for example) so that normal analog gain staging would put the input signal levels closer to the max performance of the converter? I have no freakin idea. There must be a valid reason for that. It's WAY over my head. Way over...

Even though the (-15dbfs or -18dbfs) converter is not working at it's maximum "resolution" (as you correctly stated) the overall sound of the music (without the abnormally gain-staged analog gear, and with the increased headroom for plugs and on the mix buss) to me sounds much better. I didn't know why until I found that explanation from Kyle that I posted. It makes perfect sense.

Good talking to you... uh ... can I get a free Distressor? heh

Lawrence
Old 22nd June 2006
  #107
Lives for gear
 

Quote:
Originally Posted by Albert
I would consider this consistent with what Dave Derr said. Peaks in the -10 to -5 area would be as close to full scale as practical in a lot of situations. You have to leave room for musicians or singers "headroom" as well. If they put out a sudden burst of level beyond the normal you need space for that.
well in a live tracking situation this is 100% true.

it seems like the exact specifics might come more into play in a studio overdub situation where you are doing 10 takes in a row of a single part and after a few takes the dynamic range of the performance is reasonably predictable.

then, especially on something like a lead vocal...if you REALLY thought that trying to peak at -1 or -2 would get you better sound....it would be possible. then i guess what people are debating is....is this desireable.

and then you run into the further complication of 1)how does a track that was recorded hot or "cold" sound by itelf and 2)how does that recording level effect the way a (often digital) mix bus/plug ins etc handles the big picture when ALL tracks in a mix were tracked with a "as close to 0 as possible" method.

which then goes directly into ITB OTB issues. meaning that it is likely that clean signals that were recorded hot will not screw things up when mixed OTB whereas if you are planning on mixing ITB and have many tracks then more conservative levels might be the best all around plan.
Old 22nd June 2006
  #108
The Distressor's "daddy"
 
Dave Derr's Avatar
 

CONVERTERS CONSCHMERTERS

Quote:
Originally Posted by Lawrence
I hear you Dave.
Good talking to you... uh ... can I get a free Distressor? heh
ABSOLUTELY! The Distressor is free, your only cost is the power cable which is $1495. Its a high resolution power cable with a special filter end on it that only accepts "good" electrons. Of course, we can't ship a Distressor without a power cord...

OH... did I mention there's also a $100 handling fee?
Old 23rd June 2006
  #109
Quote:
Originally Posted by Lawrence


This mirrors Bob O's comments about the mix he received at -25, open ballsy and more analog like.

Lawrence

More open, ballsy and analog like than what? What's he comparing it to?

Assuming that there's a reference and the one with more headroom sounded better, the real question is whether or not that's the only variable. I doubt it, but I didn't see the original post.

I'm guessing that when people make generalizations, especially from the mastering engineer point of view their saying that the see sonic patters that coincide with peak levels. Correlation does not equal cuase.

That may be a reflection of the person who tracked and mixed the project and they may just be a better engineer. It maybe be the content that sounds more open and analog and that at a higher peak level it was the same.
Old 23rd June 2006
  #110
Quote:
Originally Posted by Mike Caffrey
More open, ballsy and analog like than what? What's he comparing it to?

Assuming that there's a reference and the one with more headroom sounded better, the real question is whether or not that's the only variable. I doubt it, but I didn't see the original post.

I'm guessing that when people make generalizations, especially from the mastering engineer point of view their saying that the see sonic patters that coincide with peak levels. Correlation does not equal cuase.

That may be a reflection of the person who tracked and mixed the project and they may just be a better engineer. It maybe be the content that sounds more open and analog and that at a higher peak level it was the same.
Hi Mike

Bob's post was actually on this very thread, a page back... see below.

Quote:
Originally Posted by Bob O
Instead of jawing about theory, I'd recommend that people do some actual testing for themselves.

My epiphany came when I set up a songwriter friend with a digi 001 system and he turned up a few weeks later with bounced mix files that peaked at -25 because he couldn't figure out how to get any higher levels.

I was working on a mastering project that had been mixed by a great engineer on an SSL 9k using Lavry Blue converters. My friend's demos using the 001 with just an old U-87, a 414 and a 57 sounded bigger, ballsier and more like analog than 90% of what you hear today including the high-end project I was working on.

It's important to never think of digital audio in analog terms. It's numbers. Period. Digital clipping doesn't sound that bad provided you never try to calculate any additional signal processing math using those numbers. At that point you'd have more resolution with a 12 bit signal than the mathematical gibberish that digital clipping produces.
Old 23rd June 2006
  #111
Quote:
Originally Posted by Lawrence
I have to disagree with the guy who's here saying "tracking as close to digital 0 or full scale as you can" gives better sound. In my experience that is simply not true.
Lawrence
I think you're directing that at Dave, but it could be directed at me as well.

One big thing that's being over looked is content.

The following math maybe a little of but I'm sure someone can paraphrase it and I'm sure you'll get my point anyway.

I think the consensus is 24 bits translates to 19 usable.

If you print your peaks at -6 you're at 13 which after some math translates in to something adequate, but not the best.

But, suppose we're talking about a kick or a snare or some kind of percussion? That transient could be peaking 10db higher than the average or the body of the sound, so subtract another 10db and you're at 3bits for your kick or snare.

Now the great engineer using ADATs who printed with peaks at 0DBFS has got a better resolution than mr conservative using HD or Apogees or whatever.

In our debate here, we really need to specify what content we're talking about when we argue about peak level. Hitting or crossing 0DBFS with a kick or snare is a totally different thing than hitting or crossing with a synth pad or a mix.

So if you're watching your meters say you're kick is peaking at -6, what you're mostly hearing is the sound that comes after which is being printed at -13.

Actually, I used bits and dBs interchangeably which is wrong, but the point is the perception of fidelity is going to come more from the body of the sound than the transient and that may be 10dB lower than the way you think of your track.


Now, what if you're printing wtih compression or limiting, especially limiting? Now you hear "distortion" when you print hot and get up near 0DBFS. Is that really what you're hearing or are you hearing the sound of a flat top/clipped wave caused by the limiter or clipping at the pre which is now audible because you've got better resoltion and can hear flaws that weren't apparent before?

Since we're not referring to an actual piece of audio, these questions have no right or wrong answers, but they need to be considered.
Old 23rd June 2006
  #112
In the end, I think the connection between dynamic range and bit resolution is completely irrelevant.

If it was, we would print a mix with a modern 3dB dynamic range to an 8bit converter and be fine and no one would debate that that's not true.

Sample rate is like the frame rate in film. The higher the rate, the smoother the reproduction - for example, movement in film, the roundness of a sine wave in audio.

Smoothness is not detail.

Bit rate is the digital word length. The longer the word, the more detailed the description of that slice of time.

You know the saying a picture is worth 1,000 words? I'd rather have my picture described with 24 words than 19 or 12 or 8 or whatever we get down to from leaving tons of headroom for a peak that's 3-4 samples long (1 10,000th of a second!) that happens 6-8 times in a 3.5 to 4 minute period.


The concept of targeting ODBFS for peaks does not mean that your levels are all smashed up within 1-2dB of 0DBFS for the entire song, it means you've got your average levels, the stuff that counts because you actually hear it, up to a level where it's getting more 1s and 0s to describe it's detail.

If we needed bits to describe the dynamic range we actually use we could all uses blackface ADATs for converters. If you want each sample to be the biggest, most robust, beautiful pictures going by 44,100 times per second bit rate does matter.

Personally, I think it matter more than sample rate, but it's far easier to make clocks run faster than it is to make CPUs that can deal with a 48bit word or 64bit word at 44,100 times per second
Old 23rd June 2006
  #113
Quote:
Originally Posted by not_so_new
Hi Mike

Bob's post was actually on this very thread, a page back... see below.
Thanks, I recognize it.

You see the flaw right?

Which is the more significant different the level the two mixes were printed at or the fact that the two mixes were of 100% different content?

Drawing an inference from that context is about as unscientific as you can get.

I know for 100% certainty that Bob knows more about all of this stuff than I do, but that choice of example doesn't make the point.
Old 23rd June 2006
  #114
Quote:
Originally Posted by Mike Caffrey
Thanks, I recognize it.

You see the flaw right?

Which is the more significant different the level the two mixes were printed at or the fact that the two mixes were of 100% different content?

Drawing an inference from that context is about as unscientific as you can get.

I know for 100% certainty that Bob knows more about all of this stuff than I do, but that choice of example doesn't make the point.
Hummmm... well I don't agree 100% with your take but that is cool.

Ben Harper has nothing to do with Elvis but I can tell there is a HUGH difference in the songs "Fight For Your Mind" and "Hound Dog." They sound different and I can hear that they sound different even if the songs have 100% different content.

I would not discount Bob's experience out of hand. The guy has forgotten more about audio than you and I have ever known and I doubt he would post what he did if he did not think it was relevant right?

Old 23rd June 2006
  #115
Quote:
Originally Posted by not_so_new
Hummmm... well I don't agree 100% with your take but that is cool.

Ben Harper has nothing to do with Elvis but I can tell there is a HUGH difference in the songs "Fight For Your Mind" and "Hound Dog." They sound different and I can hear that they sound different even if the songs have 100% different content.

I would not discount Bob's experience out of hand. The guy has forgotten more about audio than you and I have ever known and I doubt he would post what he did if he did not think it was relevant right?

I doubt it too, but he is human and if he made a mistake it could be limited to choosing a bad example to make a correct point.

I bet if that Digi001 mix was normalize it would still sound better than the SSL mix in question.
Old 23rd June 2006
  #116
Lives for gear
 

Mike,

I feel "converter resolution", while certainly true, is somewhat overstated in the context of production and mixing in a daw. I simply don't think in the context of a session and mix it has a greater effect on the overall sound of the final result as some other things that could happen when you track up high... _ excluding _ digital distortion by exceeding 0dbFS.

I doubt, given identical recordings of the same source, one peaking at -2 and one peaking at -12 anyone could even reliably pick which was which in a double blind test once the playback levels were matched. That's how good modern converters are. I don't think anyone could reliably pick the one with more "resolution" until tracking levels get much much lower than what I talked about.

My point (...again for the 45th time ... ) is that in the context of mixing in the daw the increased headroom (among other things) seems to make my mixes and many other peoples mixes sound better. It has absolutely nothing to do with converter resolution.

It has to do with how a native daw like mine might operate with much more headroom or how the mix buss might breath and be more open and airy peaking at -10 or -12 or how my analog gear might sound better being driven to 3 or 4dbVU instead of 9 or 14dbVU to get a signal hot enough to peak at -2 or 3dbFS no matter what the source audio is or where the "body" of a particular source's sound is.

Long sentence.... bad writing.

Maybe it just sounds better because of the increased headroom for plugins? Maybe it's a thing only true for 32-bit float and not for PTHD's audio path. Maybe I'm hearing things.

I'm not here disputing converter design theory and never was. I know there's more "resolution" at the perfect 0dbFS than -10dbFS. I'm saying that in my experience giving up that (for all practical purposes inaudible) gain in "resolution" and "detail" I get a very audible gain in the quality of my final mixes by lowering the levels going into the daw when tracking and by keeping a more conservative master bus level when mixing.

I keep being told "..but the converter sounds better up closer to 0dbFS.." which is entirely irrelevant to what I'm talking about. It has nothing at all to do with converters. I'll ask you the exact same question I asked Dave...

Have you intentionally tracked and mixed in your daw peaking at -10 or lower (peaking not the "body of the sound" you talk about), mixed into the master bus in your daw peaking at a conservative level (-10?) and compared it's subjective quality to your previous work where your levels were peaking much higher and recorded with more "resolution"?

Did you hear any significant lack of detail or degradation of the audio? What did that mix sound like? Overall quality-wise was it better than you've come to expect from your daw mixes? Worse? The same? What? Lose imaging or detail? What difference if any?

It's all subjective but we make subjective decisions about how we approach things everyday. When it sounds "bad" to us we stop or change, when it sounds "good" or "better" we try to repeat it. 5 people may have 5 different views about what's actually "good" or "better". I'm asking your opinion... but you gotta do it first. If you don't it's not a subjective opinion it's a guess.

People keep talking about converter resolution and I've been here (in vain) asking some of the converter scientists to just put the papers down, go record some tracks at much lower levels and see what happens. I'm curious to know if you'll come back and say something like ...

Quote:
"I tracked everything peaking at -10/-12 and mixed it. I heard a definite lack of detail in my mixes when I did that. My previous work sounds noticably better. That's the wrong approach. You shouldn't do that. My acoustic guitar definitely lacked detail. My mix seemed flatter and a little grainier..."
If that additional "converter resolution" is as important as I'm being told you should hear a degradation in quality of some sort because you know what you studio quality sounds like and what to expect from your system.

Like I told Dave, if that's what you're telling me right here right now then fine. I'll take your word for it, we'll agree to disagree, and we'll move on. No big deal.

If you haven't actually done it then ... ??? ... what exactly is the point of going back and forth with this?

It's really easy to do... just do it (if you haven't already, if you have already just tell me about it) and get back to me.

This is a polite discussion about a subject that can very easily be physically tested to some reasonable degree.

Okay?

Lawrence
Old 23rd June 2006
  #117
The Distressor's "daddy"
 
Dave Derr's Avatar
 

Quote:
Originally Posted by Mike Caffrey
The concept of targeting ODBFS for peaks does not mean that your levels are all smashed up within 1-2dB of 0DBFS for the entire song, it means you've got your average levels, the stuff that counts because you actually hear it, up to a level where it's getting more 1s and 0s to describe it's detail.
Well put, Mikey! Your rare transient peaks are what's up there at 0dBFS for the most part.

Of course with something like modern vocal tracks, they are tracked with so much compression that there's often only a 20-30dB dynamic range between the breaths and the peaks!

(Wow, as a manufacturer of compressors, did I actually say that? )

I had a blast out at Tape Op. Wish you and I had more time besides that last late night conversation... although I did like the subject matter.

There was a lot of new stuff brought to my attention at the Tuscon convention, especially late nights around the pool... with beer. I think one of the issues that interested me, was the center channel thing in 5.1. I heard about 20 major dudes say they thought the center channel is kind of a waste for them many times. I was curious if you were in on any of those conversations? Perhaps we should email or private message, or even use that phone thing, since this is an unrelated subject.
Old 23rd June 2006
  #118
I follow you a bit better now Lawrence, I'm going to leave the quote out and just respond.


Let me try paraphrasing our two points.

My point is that an AIFF file recorded in a way that optimizes the resolution played back unaltered will sound better than one that hasn't taken advantage of the full resolution.

I beleive that you point is a file recorded with X amount of headroom will sound better in the mix.

Micro vs Macro. In a way we're arguing two separate points.

I think Dave mentioned earlier that if you have to record lower than your optimal level to compensate for inadequate plugin headroom, that's a flaw with the plugin design.

Pratically speaking, if that's the case, for mixing ITB I'd have to say your approach is better.

For my situation, I don't use plugins, so it's not somthing I worry about. If we're talking about the fidelity of a raw file, I'm still for maximiaing the resolution.

As far as have I don't A/B test, none formally that I can remember. Have I had an opportunity to hear tracks recorded very low and hear them not sound as good? Yes, but it's been years since they were tracks that I cut that way.

Another problem with low volume tracks is sometimes you have to add gain which can add noise - like today on some tracks that someone else recorded. There were two guitars that had to be turned up at the channel line in to get them to balance with everything else, which was recorded properly and had proper gainstaging in the mix etc. The noise wasn't unbearable, but it definitely brought some out. I'm glad all of the tracks were'nt like that.


Since I started by paraphrasing, I'll paraphrase one more point. I beleive that the resolution of the average/RMS level is more important for perception of "quality" or fidelity. Have you recorded a track, say a synth pad, where the average level was -16 and then then same thing again where the average was -3 and auditioned the difference? What about at -26?

My choice of levels is to reflect the 10 dB differnece between peak and average in something like a kick or snare and the -6dB peak level that seems to be pretty popular. So -16dB is a realistic average level, and in the case of a kick or a snare, probably necessary.

I got interupted for a minute and sort of lost my point, but what I'm getting at is try recoring by ignoringg your peaks and setting your average levels way low or high, becuase I believe that's more significant than the peaks.

What do you do if a higher average sounds better, but now you're peaking all the time? That's a whole different story.
Old 23rd June 2006
  #119
Quote:
Originally Posted by Dave Derr
Well put, Mikey! Your rare transient peaks are what's up there at 0dBFS for the most part.

Of course with something like modern vocal tracks, they are tracked with so much compression that there's often only a 20-30dB dynamic range between the breaths and the peaks!

(Wow, as a manufacturer of compressors, did I actually say that? )

I had a blast out at Tape Op. Wish you and I had more time besides that last late night conversation... although I did like the subject matter.

There was a lot of new stuff brought to my attention at the Tuscon convention, especially late nights around the pool... with beer. I think one of the issues that interested me, was the center channel thing in 5.1. I heard about 20 major dudes say they thought the center channel is kind of a waste for them many times. I was curious if you were in on any of those conversations? Perhaps we should email or private message, or even use that phone thing, since this is an unrelated subject.
And if you're watching the little red box light in protools, that might not just be a rare transient peak, it may be the transtient peak from a rare extremely loud hit, or just the downbeat of the bridge or a mis-hit of a drum and that maybe there's a 20dB difference between that peak and the average.

Maybe, this is one of the significant differences between analog and digital - that we set levels in analog with VU meters, where as digital our metering is so much more accurate that we're retaining not jsut far more of the transient, but far far more of the transient. Then multiply it by more than 24 channels or even 48, beccuase in digital "we can" when it comes to tracks so we end up with more. We compress or limit and of course master differently as a result. In otherwords, I wnder if the differences in metering between analog and digital tracking affect the final signal.


The pool was cool!

I didn't see and of the 5.1 stuff, but I'm curious, were any of those 20 major guys under 50? How about under 40?

I think for the most part, 5.1 is a Senior Engineer's game as opposed to Staff Engineer or whatever most guys are. By "senior" I mean expereince and skill, which of course tends to correspond to age. I don't mean this as an age dig at all.

I can see not needing the center speaker, but I went over to Tony Visonti's room one day to hear the 5.1 Young Americans mixes and he did some cool stuff with the vocal in the center - more subtle than major.

I think somewhere right now there's a 12 year old kid struggling to get a decent sound out of his MBox who's going to be the next Brauer and win a mixing grammy 2036 for rocking the snot out of the center speaker.

When basketball first started there was no jumpshot. It was a later generation that really moved things forward.

If we're going to get into a 5.1 discussion, I say we crash Quad and go play in Brauer's room!
Old 23rd June 2006
  #120
Lives for gear
 

This is a very deep discussion, but I'll try to add SOMETHING... The difference between converting at full scale and not at full scale you could think is not important as long as the resolution is as high as it needs to be to cope with the noise floor level. There is a difference though. With the highest possible resolution you get the best possible description of the sound, both the signal and the noise. The more detailed the signal is, the more accurately it can be processed by the audio engine. It's the same with the noise. The better description of the noise the more effciently it can be targeted for digital processing, for instance in the sound shaping/dithering process. No matter the amount of noise the converter processes/adds, a high resolution means a more detailed gain level for each sample. Since the DAW is limited by computing accuracy recording at the highest possible resolution is always optimal. However, in practise when dealing with converters, they are not 100% linear in quality on all signal levels, especially not close to full scale, this varies between units and models, depending on the manufacturing process, the material quality of the components (power supply, converter chip, op amp...), room temperature, clock rate etc. Leaving a little safety headroom for this reason is a good thing to do.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
thejook / So much gear, so little time
15
Ziggy!! / So much gear, so little time
12
DeadPoet / So much gear, so little time
1

Forum Jump
Forum Jump