Keyplayer wrote on Sat, 23 April 2005 21:04 |
With all the debate over the supeior ease of automation in the DAW vrs that of most mixing consoles, I was wondering if anybody was actually using their DAW like a tape deck/editor and mixing from their consoles to a mixdown deck or even back to a stereo or 6 stem tracks on their DAW? I'm pretty sure those of you with access to Neve's, API's, SSL's etc are doing just that. But for those running in the "Mid-Line" (I.E. DM2K, R-100, Soundcraft Ghost etc.) are you doing this or letting the DAW do all the work and having your desk just act as a router? |
ryst wrote on Fri, 29 April 2005 13:32 |
I have a question. If I can't afford large format mixing console at this time but want to start mixing OTB, would a small format console from A&H, Soundcraft, or Mackie be worth buying? Would I hear a difference (assuming I know what I am doing) between ITB and OTB with a console in the $1000 range? Right now I use DP with plug ins... Any advice would be great. |
Quote: |
DAW & Desks: Is ANYBODY actually still mixing on their desk? |
wireline wrote on Fri, 29 April 2005 20:54 |
But I am curious of the operative definition of 'using a board': are we talking about submixing in DAW then stemming, or are we talking about the huge format boards? I would think it makes a difference... |
gwailoh wrote on Sun, 01 May 2005 15:52 |
...Have any of you experimented with the "FATSO" from Empirical Labs inserted across the stereo bus at mixdown? ... |
compasspnt wrote on Sun, 01 May 2005 16:19 |
In that case you would still be summing inside the computer, then just inserting an external analogue simulation device across the stereo buss. That's not at all the same as coming out track-for-track (or even with some submixes in stereo) into an analogue console, and summing there. In fact, with the console scenario, you could still use your "FATSO" if you wanted to. |
Keyplayer wrote on Thu, 05 May 2005 11:52 |
Is there something way more complicated about doing automated mixes on a desk? It seems like once everybody started using DAWs, they're doing their automation inside and just past the stems through the desk for "analog sweetening." Why is that? |
Tomas Danko wrote on Sat, 23 April 2005 16:13 |
For whatever harm my DAC's are doing to the audio, my console is more than making up for it in mojo. Sincerely, Tomas Danko |
Jules wrote on Wed, 04 May 2005 22:35 |
I want an SSL AWS 900 !!!! |
gwailoh wrote on Sat, 07 May 2005 13:54 |
Also it would be great to know how people were approaching their digital summing: out the digital outputs to a MasterLink or other device? Back in to two new tracks on the DAW? Internal bounce? With SRC or not? Etc. |
J.J. Blair wrote on Sat, 07 May 2005 12:29 |
Bob, I thought there was some issue with digital headroom and/or ability to sum the full signals w/o degredation in real time when summing in the box? I'm a digital idiot, so I'm just asking. I don't know how any of it works. I just know which things sound better to me. |
bobkatz wrote on Sat, 07 May 2005 14:57 | ||
I think the key word here is "mojo". Digital summing is technically nearly perfect. Analog summing is technically far from perfect. It MUST BE, by DEFINITION, the imperfections of the analog summing and the additional analog circuits that the signal goes through that are so attractive to many of us. I can live with that And let us remember that an API sounds different from a Dangerous Two Buss from a Soundcraft Ghost from an SSL 4000 (E or G?). So to generalize too much about the sound of "analog summing" is also a dangerous concept. BK |
J.J. Blair wrote on Sat, 07 May 2005 20:31 |
Actually, as ludicrous as it sounds, I've never run out of headroom on my console, due to the custom buss section I had made by Steve Firlotte. My master fader in PT has run out of headroom, but I don't think it was the summing. Once I backed the master fader down, the distortion stopped. I thought the inadequacies of digital summing were what made the Dangerous 2 Buss, etc. so appealing? |
Quote: |
How is it that the crosstalk is a pleasant quality? I'm confused. |
bobkatz wrote on Mon, 09 May 2005 02:42 |
There is one and only one reason for the existence of these analog "summing" boxes, and that is coloration. |
bobkatz wrote on Sat, 07 May 2005 16:42 |
Any way you measure it, digital summing is as close to perfect as you can get. The reasons you may prefer the analog mixer include distortion, noise, crosstalk, selective high frequency versus low frequency crosstalk, high frequency rolloff at 20 kHz, you name it. All the "reputable" DAWs have tons of summing headroom (more than your analog board, to be exact!), zero crosstalk (which may be the reason why some people dislike it), zero leakage between channels, flat frequency response (which may be another reason why people dislike it), and it's not subject to the nonlinearities of converters. Add that up and.... well, that's what you get. BK |
Deep White wrote on Sun, 08 May 2005 16:10 | ||
If that's true, I'd be a bit disappointed. |
bobkatz wrote on Sat, 07 May 2005 16:42 |
One detail though of the subjective experiences which some proponents of analog summing report is narrowing or diminution of the digitally summed stereo image compared to analog summing. If people's ears are not deceiving them, it seems more likely that the explanation for this particular difference would lie in defects in the DAW summing software rather than in distortion, leakage or other artifacts introduced by analog. |
bobkatz wrote on Mon, 09 May 2005 00:56 |
How can increased apparent image width be caused by anything but DISTORTION since you just ADDED something to the previous chain. |
gwailoh wrote on Sun, 08 May 2005 16:54 |
How can increased apparent image width be caused by anything but DISTORTION since you just ADDED something to the previous chain. |
bobkatz wrote on Mon, 09 May 2005 07:49 |
Why be disappointed? Your tests appear to be fairly scientific in the sense that they eliminate extra variables, and so by process of elimination it seems to indicate that we like the analog summing for what it adds, not for what it "fixes". |
Deep White wrote on Mon, 09 May 2005 01:22 |
p.s.: When mixing ITB with Nuendo, I never drive it near 0dB, since I can always drive it with Vari-MU's input gain. I don't know if this is the key that save me from bad digital summing mixes. |
bobkatz wrote on Mon, 09 May 2005 08:18 |
Actually, the key to how to avoid "bad" digital summing mixes is to LEARN what it is that the analog stuff did automatically and apply that to your digital mix. Learn about the Haas effect. Learn how delays, crossed delays, early reflections, and phase manipulation, even the addition of noise and distortion, can be used to create a digital mix that is as dimensional (or MORE) than any analog mix you used to make. BK |
gwailoh wrote on Sun, 08 May 2005 22:18 |
It seems to me though that if analog summed stereo images were consistently superior, the logic could be the reverse of what you suggest. E.g, it's not that analog summing is adding something, but that digital summing contains digital errors which analog summing eliminates. I'm not saying that I believe this to be true, but it does seem to be a logical possibility, and I think it's consistent with the anectodes which someone posted in another similar thread re how removing plugins from the ITB mix improved the results. |
Tim Gilles wrote on Mon, 09 May 2005 08:30 | ||
Once upon a time there was an Emperor who held a wonderful parade every Saturday.... |
Nathan Eldred wrote on Mon, 09 May 2005 15:51 |
Everytime I come across the comparison, there isn't a comparison. When I hear different engineer's work. What they used to do, and what they do now. Analog wins every time. ITB, can sound 'okay', but it requires good outboard gear and incredible conversion (again just to get to a B+...and that's just the guys with the best skills). The same guy doing the same thing on a high end analog console, it becomes an A+. Give yourself a letter grade advantage without doing anything extra. Just my $.00002 & HO. Maybe this is obvious to a lot of people....but is average sound acceptable, because convenience and cost take precedence? |
bobkatz wrote on Mon, 09 May 2005 16:28 |
All other things being equal, Nathan? That is: NO PLUGINS on the digital side, and NO outboard on the analog side? Try that first. Start with excellent A/D and D/A converters, do a digital mix. Feed lots of good outboard with aux sends and analog inserts, stay away from (too many) digital compressors, and you'll have a mix that's both transparent and fat, dimensional and beautiful. It can be done. If you have heard ONE In the Box mix that sounds superb, that is a simple "exoneration" (as if there needs to be) of the all digital mix. I've heard (and done) my share of them. You do have to learn new skills. A noiseless mixer doesn't mask or hide problems. A distortionless mixer doesn't add any sounds of its own. |
maxim wrote on Wed, 11 May 2005 09:41 |
arys wrote: "I don't think we are doubting that working with an analog console is better than totally ITB. " i think bob k just intimated that it might not necessarily be the case the question is not whether you can do a better mix on a daw, but whether a better mix can be done on one not the same thing |
bobkatz wrote on Wed, 11 May 2005 09:41 |
and "jazz" and "classical" tends to sound better with an all digital mix. Oh boy, am I going to get creamed for that generalization... all I am trying to do is describe in complex words what only takes 10 seconds to realize by ear BK |
Juan Covas wrote on Tue, 10 May 2005 14:22 |
Yes I still using as much analogue as posible.Still a warmer and fatter sound.I've been using both for the past 15yr and is the perfect combination.Trident pre rules..... |
bobkatz wrote on Wed, 11 May 2005 09:12 | ||
I don't like to generalize, but "rock and roll" tends to sound better with an analog mix or lots of analog outboard, and "jazz" and "classical" tends to sound better with an all digital mix. Oh boy, am I going to get creamed for that generalization... all I am trying to do is describe in complex words what only takes 10 seconds to realize by ear snip... BK |
Nathan Eldred wrote on Tue, 10 May 2005 03:10 | ||
I have done the tests, with all things being equal. I've heard other people's attempts (that's what I said above). I was saying that analog outboard somewhat helps ITB, plugs destroy it. Analog outboard on a console (operator being equal) is godlike. It's not a skill set deficiency, it's a tool deficiency. Maybe it's distortion, maybe it's cross talk, maybe it's because we like the sound of electrons flowing through a wire. Either way, through an analog console the music is sweeter/wider/deeper/extended/more real/more emotional intuitive to my ears, brain, and spirit. This is one case where empirical experience has won over scientific dogma on a daily basis, for me. And it's why I've put my money where my mouth is by hiring a tech to continually maintain a 2" & 1/4" deck, an analog console, and a respectable mix of vintage and new analog outboard. FWIW I don't keep it completely in the analog realm, the multitrack gets bumped to the computer and spit back out to the console and tape deck. Running a commercial studio for my client base without a partial involvement of the computer would be impossible. But this is not for sonic reasons. |
compasspnt wrote on Thu, 12 May 2005 06:03 |
My firm belief is that if users of Protools, and the other DAW systems, would do the following, then MANY of the "digital" or "in the box" audio problems would vanish: ?STOP RUNNING YOUR SIGNAL SO HOT! Do not use the built in peak meters as you would use a VU meter. If you will keep your input levels lower, your sound will improve. You are not really gaining anything by trying to squeeze out that last little bit. ?If you can, find a way to also meter every input with a good old fashioned analogue VU meter...whether it's with an analogue console, a tape machine, a dedicated "meter box," whatever. Let these meter indications rule, while of course cross referencing the peak DAW meters as well. And use MUSIC as your general guide for I/O levels. Setting up with a 1k tone from inside Protools for your reference level to an analogue console following will not give accurate results on all types of program material. |
compasspnt wrote on Thu, 12 May 2005 14:03 |
•If you can, find a way to also meter every input with a good old fashioned analogue VU meter...whether it's with an analogue console, a tape machine, a dedicated "meter box," whatever. Let these meter indications rule, while of course cross referencing the peak DAW meters as well. And use MUSIC as your general guide for I/O levels. Setting up with a 1k tone from inside Protools for your reference level to an analogue console following will not give accurate results on all types of program material. |
Paul Frindle wrote on Thu, 12 May 2005 04:06 |
Every time this comes up I am left with exactly the same totally frustrated feeling. The reason people do not get best results from ITB mixes and digital processing in general is that the whole cultural environment of metering, level control, overload and production styles within the digital domain is based on SAMPLE VALUE and not SIGNAL However many times I re-itterate this very important fact it seems impossible for people to grasp exactly what it means and what the gravity of ignoring it actually is in repect of their audio results. And this is NOT even the user's fault, they cannot be expected to grasp it because they are totally buried in systems that are wholly based on sample value misconceptions and always display values which are NOT signal For a really fair analysis, this is not primarily a user problem - it is an equipment problem that the user must make himself aware of if he is to avoid it. People who hear differences are not wrong - the equipment is lying to you - it is ecouraging you to produce illegal results that you are not made aware of. IMHO & LE this is the sole reason underpinning ALL the arguments about ITB mixing, sample rates, 'resolution' - you name it. |
compasspnt wrote on Thu, 12 May 2005 07:03 |
My firm belief is that if users of Protools, and the other DAW systems, would do the following, then MANY of the "digital" or "in the box" audio problems would vanish: •STOP RUNNING YOUR SIGNAL SO HOT! Do not use the built in peak meters as you would use a VU meter. If you will keep your input levels lower, your sound will improve. You are not really gaining anything by trying to squeeze out that last little bit. |
Paul Frindle wrote on Thu, 12 May 2005 10:38 | ||
You are right in principle in everything you say. But a VU meter won't help you because it's too slow and will allow peaks to go unnoticed that could cause problems in the digital domain. They were fine when one used them with experience in the analogue domain, where healthy a degree of overload margin was implicit. |
compasspnt wrote on Thu, 12 May 2005 20:21 You are right in principle in everything you say. But a VU meter won't help you because it's too slow and will allow peaks to go unnoticed that could cause problems in the digital domain. They were fine when one used them with experience in the analogue domain, where healthy a degree of overload margin was implicit.[/quote |
Exactly Paul...that's why I included: "...while of course cross referencing the peak DAW meters as well." Thanks! |
Paul Frindle wrote on Thu, 12 May 2005 17:54 |
So - go back and get your fav test mix back up on your W/S, re-mix the whole thing making sure that at every place in all chains (including between all plug-ins) never gets bigger than -6dBr. Make sure your final output after any limiting etc also never peaks beyond -6dBr. Now do the comparison between this ITB mix and a similar OTB mix. You might have a big surprise |
blairl wrote on Fri, 13 May 2005 05:19 | ||
I am extremely intrigued by your comments. I am trying to reconcile your recommendation with a statement made by digidesign engineering. They claim that it is impossible to clip the internal mix bus. With the maximum number of tracks possible in Pro Tools, all with coherent signals at +6 db on each fader, it is impossible to clip the internal mix bus. Yes it is possible to clip the output, but not the internal mix bus. To avoid clipping the output, they say you can simply lower the gain on the master fader however much needed to avoid the problem. Your statement and digidesign engineering's statement may be completely unrelated, so forgive me if I am linking them together inappropriately. |
Eric Vincent wrote on Fri, 13 May 2005 05:54 | ||||
My initial take is that you are indeed linking them together inappropriately. Mr. Frindle has attempted to inform you on how to attain the optimal sonic quality from your DAW. The Digidesign Answerbase which you quote from, on the other hand, attempts to inform you on how their kit will potentially handle abuse of it. Measuring the performance of the kit by how well it performs under abusive conditions is not really very helpful, nor informing. It's kind of like asking, "If I crash my car head on into an oncoming truck at 65mph, what are my chances of surviving?" Mr. Frindle is attempting to inform you how to avoid the head-on collision. Digidesign is attempting to inform you of your survival chances. Make sense? |
Quote: |
I am extremely intrigued by your comments. I am trying to reconcile your recommendation with a statement made by digidesign engineering. They claim that it is impossible to clip the internal mix bus. With the maximum number of tracks possible in Pro Tools, all with coherent signals at +6 db on each fader, it is impossible to clip the internal mix bus. Yes it is possible to clip the output, but not the internal mix bus. To avoid clipping the output, they say you can simply lower the gain on the master fader however much needed to avoid the problem. Your statement and digidesign engineering's statement may be completely unrelated, so forgive me if I am linking them together inappropriately. Can you comment? |
Quote: |
Your experiment recommends never peaking above -6dbr, even after any final limiting. Are you saying it's impossible to ever bring a final mix up to 0dbr without adversely affecting the sound quality? |
Quote: |
If I put a limiter on a master fader in Pro Tools, the digital summing has already occurred at 48 bits then been dithered to 24 bits before it even hits the limiter plug-in. If I were to sum my mix, never peaking above -6dbr at any stage before hitting the limiter plug-in, then bringing the final level up to 0dbr using the gain on the limiter, would this negate your experiment? |
Quote: |
How could a DAW application be designed to eliminate the problem you point out? Would some kind of an internal reconstruction filter after every track and process be required? Is the problem apparent only in DAW's or does it show up in any digital mixer? |
gwailoh wrote on Fri, 13 May 2005 20:01 |
...Can level be safely added during mastering, or do the same issues apply there when processing the stereo mix one provides to the mastering engineer? |
gwailoh wrote on Sat, 14 May 2005 02:01 |
Paul, many thanks for helping us (me anwyay) to understand this issue better. Now, supposing one were to follow your recommendations re staying -6dBr down, and so on, mixing ITB. The mix is great, everyone's happy with it. But, you want it to be competitive, that is, loud, in the clubs and on the radio. What to do? Can level be safely added during mastering, or do the same issues apply there when processing the stereo mix one provides to the mastering engineer? (Would it be better to provide digital stems to master with?) Again, many thanks. |
Bob Olhsson wrote on Sat, 14 May 2005 17:37 |
Paul, a couple questions: 1. does the digi white noise generator produce an illegal signal? I can't get pink noise recordings, even that I eq., to do anything comparable using hi-pass or low-pass filters. OK, I just bounced the digi white noise to disk and see the same thing in both Pre Tools LE and Samplitude. 2. I notice huge variations between different software filters. Why is this, what is the mechanism? |
Eric Bridenbaker wrote on Sat, 14 May 2005 19:19 |
Thanks, Paul for clarifying this issue. Recently I've found that running the mixes at lower overall levels has yielded far better results, a certin clarity and punchy transparency that is quite audible, compared to a "hotter" mixing approach. Even though an analog VU meter is too slow to catch these overs, I find it to be a helpful tool, to have some form of metering OTB. I have a question about ITB meters though: Is it not possible to get better metering of these anomalies ITB? There is a plugin by Elemental Audio calle Inspector, for example which is supposed to show clipping instances. http://www.elementalaudio.com/products/inspector/index.html Is something like this of any use in getting a visual on this stuff ITB? Cheers, Eric |
Eric Bridenbaker wrote on Sat, 14 May 2005 19:19 |
Thanks, Paul for clarifying this issue. Recently I've found that running the mixes at lower overall levels has yielded far better results, a certin clarity and punchy transparency that is quite audible, compared to a "hotter" mixing approach. Even though an analog VU meter is too slow to catch these overs, I find it to be a helpful tool, to have some form of metering OTB. I have a question about ITB meters though: Is it not possible to get better metering of these anomalies ITB? There is a plugin by Elemental Audio calle Inspector, for example which is supposed to show clipping instances. http://www.elementalaudio.com/products/inspector/index.html Is something like this of any use in getting a visual on this stuff ITB? Cheers, Eric |
zetterstroem wrote on Sun, 15 May 2005 09:25 | ||
i don't think it will work..... it needs to emulate the reconstructio process in the d/a as this : http://www.tcelectronic.com/Default.asp?Id=3322 it's called gibb's effect... we talked about it here: http://recforums.prosoundweb.com/index.php/t/4140/6691/?SQ=1 8552ac44754d36523d0c0da1d75d3b5 i glad someone is debating it again.... i think we all really need to rethink our approach to digital sound..... |
compasspnt wrote on Mon, 16 May 2005 00:03 |
Very nicely stated, Paul. Thank you for the awesome detail. Now everyone turn it all down a bit... |
Dan Pinder wrote on Mon, 16 May 2005 00:50 |
As I struggle to grasp what's going on in Paul's explanations, I wonder if this is in any way related to the phenomenon that the Trillium Lane Labs TL|MasterMeter seeks to illuminate? Or is this purely something that happens independent of the DAC, as is implied by the unexpected metering results you get in the experiment? |
zed wrote on Mon, 16 May 2005 05:38 |
Paul, I have been running some tests in my protools system that have been nothing but inconclusive. Basically what i've done is to grab a session with 20 tracks, mix them as loud as possible without peaking individual or master tracks, bounce, and repeat the above but with trim plugins at -6dB inserted in every track. (they all have some plugins that are time constant) When comparing both stereo mixdowns the difference between them is zero, they basically cancel each other when phased inversely. I'm sure that i'm not understanding the difference between mixing at high levels or low levels since they are giving me similar results! please comment what i'm doing wrong! Best Regards Zed |
zed wrote on Mon, 16 May 2005 12:38 |
Paul, I have been running some tests in my protools system that have been nothing but inconclusive. Basically what i've done is to grab a session with 20 tracks, mix them as loud as possible without peaking individual or master tracks, bounce, and repeat the above but with trim plugins at -6dB inserted in every track. (they all have some plugins that are time constant) When comparing both stereo mixdowns the difference between them is zero, they basically cancel each other when phased inversely. I'm sure that i'm not understanding the difference between mixing at high levels or low levels since they are giving me similar results! please comment what i'm doing wrong! Best Regards Zed |
zed wrote on Mon, 16 May 2005 10:59 |
The recordings were all made with apogee converters not going any further than -5dB on input! |
zed wrote on Mon, 16 May 2005 17:59 |
Well, the mix i've been using is a fairly standard pop song with no vocals, only drums, bass guitar and electric guitars, the plugins used are EQ and delay, no compression (does compression behaves differently when changing -6dB on signal input even with the same value applied to threshold?*). I repeated the test, this time i didn't even care if the mix sounded good at all, i tried to get the summing of tracks as hot as possible (on the mixdown with no trim plugin on every track the master fader even peaked once!). Still, when checking the -6dB mixdown with the hot one the only difference between them was that evil peak on the hot mixdown, other than that, they were mathematically identical! It's truth that i'm not using compression and the added HF on some of the tracks is rather moderate, but shouldn't this be enough to start getting some differences when comparing these mixdowns? The recordings were all made with apogee converters not going any further than -5dB on input! Paul, with this tests i'm not trying to prove you wrong, believe me on this, i'm only trying to understand your statement trough practice, for all our best interest it is very important for us to understand our tools as best as we can! *- i did try this same experience using compression, in this case the difference was very noticeable, although it seemed more like a compression problem... Best Regards Zed |
Bob Olhsson wrote on Mon, 16 May 2005 21:44 |
What's the title? I'm also curious about what happens when it gets lossy coded. |
amwintx wrote on Mon, 16 May 2005 13:09 |
At the risk of posting a reply to the original topic .... I am surprised no one has mentioned the original reason I went to the time and expense of getting an analog mix setup working. It is fun to mix on a real board! I like to turn off the computer screen and simply listen to the speakers and pull up the faders/set eq/patch in outboard etc. I think it also leads me down different path creatively sometimes. Now back to regularly scheduled the ITB/OTB debate...... |
Eric Vincent wrote on Mon, 16 May 2005 01:44 | ||
That means (among other things): Reductive EQing. Yo? |
Eric Bridenbaker wrote on Sat, 14 May 2005 11:19 |
Thanks, Paul for clarifying this issue. Recently I've found that running the mixes at lower overall levels has yielded far better results, a certin clarity and punchy transparency that is quite audible, compared to a "hotter" mixing approach. Even though an analog VU meter is too slow to catch these overs, I find it to be a helpful tool, to have some form of metering OTB. I have a question about ITB meters though: Is it not possible to get better metering of these anomalies ITB? There is a plugin by Elemental Audio calle Inspector, for example which is supposed to show clipping instances. http://www.elementalaudio.com/products/inspector/index.html Is something like this of any use in getting a visual on this stuff ITB? Cheers, Eric |
bobkatz wrote on Mon, 16 May 2005 17:27 |
As Paul pointed out, there are some expensive oversampled peak meters that can measure if you are getting into trouble... |
bobkatz wrote on Tue, 17 May 2005 00:27 | ||||
Well, I'm sorry to say that sometimes an EQ reduction can produce overlevels! It sounds counterintuitive, but those intersample peaks are caused by filtering, regardless of whether it's a dip or a boost. Gerzon wrote a paper on this problem. So don't count on your "subtractive EQ" to keep the level down if the input level is hot. As Paul pointed out, there are some expensive oversampled peak meters that can measure if you are getting into trouble, but it's a lot safer just to peak to, say, -3 dBFS (max, lower being better) on the final mix and be done with it. BK |
Paul Frindle wrote on Tue, 17 May 2005 09:04 | ||||||
Yes indeed this is so - anything that 're-arranges' the phase and response can get you HIGHER peaks - in fact up 6dB higher for LF roll-off of an already totally clipped signal - the need for headroom isn't just intersample peaks, it's your whole freedom of artistic expression within your mix that's at risk. For instance - if the mastering engineer decides to roll-off some excess LF from your programme he could well end up having to REDUCE it's level (and all-important loudness) to accomodate the extra peak values!! In the experiment I described you can try this too by taking out the HF filtering and putting in LF filtering instead - and watch the levels rise Another interesting one is to get a squarewave from the genny at say 200Hz -6dB and then insert a filter set to cut off below 100Hz - the peak level will rise 6dB (or more if the filter is a high order). Ok this is an extreme example (not very musical) but if this was your programme at full level then it would have to be represented at 6dB (or more) lower than the original in order to avoid actual clipping!! Food for thought for those that like to clip the kick drum a teensy bit for that little bit extra attack and presence. And of course the same thing would apply at higher freqs (such as vocals) if you let it saturate a bit for effect then EQ it a bit to roll-of some of the 'raspy edges'. This is not new stuff - and the same things happen in analogue processing as well - it's just that analogue systems have signal operating levels 10's of dB below signal clipping - and analogue tape recording methods were more tolerant cos they could accomodate significant overload at LF (as much as 10dB) and saturated softly at HF (i.e. produced that fuzzy splashy 'air' in the presence of HF overload - rather than hard and ear-grating clips). |
Extreme Mixing wrote on Tue, 17 May 2005 11:30 |
Make loud copies for the producer, the artist, and yourself, for reference listening, because we all know that you can't send clients out the door with music that's 12db lower than everything else they compare it to. Steve |
zed wrote on Tue, 17 May 2005 20:54 |
Paul, I know you have extensive experience in this matter, when do you think that digital dynamics (plugins) will start to compete with their analog counterparts? Best Regards Zed |
Extreme Mixing wrote on Tue, 17 May 2005 17:30 |
I think you have to look at it this way, Eric. Any signal processing that you do in a mix requires some digital headroom to get the math done correctly. So leave a little room. |
Quote: |
You wouldn't want to use subtractive eqing because you HAVE TO in order to keep from clipping. Print at lower levels so you can do what you want in the mix. |
Quote: |
And give it to mastering the same way, so their options are open, as well. |
David Schober wrote on Tue, 17 May 2005 10:20 | ||
I'm with you there Steve. However my only fear about making a hot rough is that sometimes people can fall in love with a rough like that. When turning a rough mix for a sales conference or A&R meeting I'd be a fool to turn in a soft, but good sounding mix, knowing that other mixes will be smashed and louder than mine. However, my worry is that someone, the artist, record exec, or whoever, has tin ears, used to hearing smashed crap may prefer the rough over the master. Despite the fact that after a proper mastering the mix would sound better, it's hard to make people belive that will be true. Any thoughts, comments? |
Eric Vincent wrote on Tue, 17 May 2005 11:46 |
OK. But let's for a moment assume that my levels are not clipping, that I've kept my inupts and levels at a conservative position. In that scenario, doesn't one get less risk of clipping by EQing subtractively, rather than additively? |
bobkatz wrote on Fri, 20 May 2005 00:19 | ||
All other things being equal, I would tend to agree. But Paul (who is more expert on this than I) might intervene and say that if a high pass filter is considered "subtractive" then beware of the intersample artifacts of the high pass. Let's see if Paul catches this post. |
Eric Vincent wrote on Fri, 20 May 2005 01:20 | ||||
Notwithstanding Paul's views on the subject, I have my own observations of the effects of ITB high-pass filtering: Beware. I have switched settings on EQ plugins from a high-pass to a low shelf cut many times, because the high-pass setting introduced "something" undesirable, whereas the low-shelf cut alternative did not. Antectdotal, yes, but antectdotes sometimes serve a purpose. |
Paul Frindle wrote on Fri, 20 May 2005 10:57 | ||||||
The answer to this depends entirely on the actual signal content. Eq can cause 2 types of 'clipping error' - boosting Freq ranges increases level and changes phase so level clips are obviously more likely. Cutting EQ reduces gain (which intuitively seems safer) however it also changes the phase - so overs are still possible from cutting something. This situation is worsened by processing already clipped signals (even at LF) as the max sample value is totally reliant on the phase of the harmonics. You can test this by getting an oscillator plugin set to say 200Hz squarewave at -6dBr then filtering below 100Hz - the sample values will almost double depending on the steepness of the filter. |
innesireinar wrote on Fri, 20 May 2005 08:43 | ||
I work with one or two excellent mixing engineers who are doing just that, mixing in the box but routing to outboard compressors and reverbs and other gear. Sounds great to me... the key is they know what to do and how to get the most dimensionality and depth with their mixing technique. The problem is that to do a compressor plugin right requires tremendous CPU power, more than the manufacturers of plugins believe you are willing to tolerate with today's CPU's. In due time all digital compressors will be 2 to 4 times oversampled internally. One way around this is to track and work at 96 kHz, the sound of non-linear plugs like compressors will be much better, purer, more analog-like. I'll leave it up to you whether you think at that point they are the sonic equal of their analog counterparts. In that case, if you're working at 96 kHz, and not trying to push those comps beyond their mortal means, that is, using them subtly, 1-2 dB GR, to slightly "fatten" or pump" up a sound, or subtly control an overdynamic instrument, I believe that a number of the plugs, for example, like the Waves Ren Comp, could do a reasonable job on background things and who knows, maybe even lead vocals! It's only when you're trying to "punch" and "pump" and the rest---that's where the rubber meets the road and I still believe that good outboard gear adds that sheen, perhaps because of the unique distortion of certain analog compressors, which has not yet been emulated in digital-land. Or it might be something to do with the time constants. I certainly love my Weiss digital compressor and I can even make it work aggressively. I was able to add some "attitude" to a rock mix just today. Working at 96K and with the Weiss, I personally was very happy with the sound I got today punching up a rock project. So digital comps have come a long way. You'd have to show me a compressor plugin that at 96 kHz is as versatile and good-sounding as the Weiss...it may exist, as I say, I do not have any comps outside of the excellent Waves stuff. I'm not that current with the sound quality of the best digital compressor plugins, but of the ones I have access to, I still believe there are still some sonic tricks which cannot yet match (to my ears) the sound of some analog compressors. And I do love my Cranesong Trakkers, which I still think have a few tricks (including that "analog sound", the additional fattening which those rich harmonics can do) that I think even my Weiss cannot do. |
compasspnt wrote on Sat, 21 May 2005 00:38 |
Very interesting, Paul. It appears obvious that this odd effect holds true universally, both analogue and digital. Would it then not make sense for a HP or LP filter design to include "automatic" reduction of the nearby "offending" freq range, and that said reduction would be somewhat "hidden" to the operator, in that the labelled choice would not "mention" it? Or have you already done such a thing? Terry |
Bob Olhsson wrote on Sat, 21 May 2005 17:52 |
I think we're dancing around the main point which is that it's much easier to clip audio "in the box" than most people realize and that clipping has much worse consequences than most people realize. I go back to my own (and many others') experience in the mid '60s with the difference between working in a tube studio vs. working in an early solid state studio. We set up our rooms exactly as we had with the old boards and suddenly everything sounded terrible. We were told "you must be overloading your mike preamps" so we built mike pads. Unfortunately the pads changed what the mikes sounded like and the noise level went up. When we objected, manufacturers told us, "you must just like "tube distortion" better than solid state which is "much cleaner." It turned out that they were lying with their statistics. Our old tube gear clipped around +45 while the new solid state gear clipped at +18. This "warm euphoric tube distortion" was actually all about having 20+ more dB of additional headroom in every active stage of the old tube consoles! The myth of "warming" persists. |
innesireinar wrote on Fri, 20 May 2005 08:43 | ||
The problem is that to do a compressor plugin right requires tremendous CPU power, more than the manufacturers of plugins believe you are willing to tolerate with today's CPU's. In due time all digital compressors will be 2 to 4 times oversampled internally. One way around this is to track and work at 96 kHz, the sound of non-linear plugs like compressors will be much better, purer, more analog-like. I'll leave it up to you whether you think at that point they are the sonic equal of their analog counterparts. BK |
bobkatz wrote on Sat, 21 May 2005 01:33[/quote |
Is there same differences by using analog eq in insert instead of eq plugins? For example how is better using an original API 550b in a PT insert slot instead of an URS API replication? Thanx |
innesireinar wrote on Sun, 22 May 2005 03:41 |
Bob, Having said that, I've stated from your words that the goal could be the exact opposite: mixing ITB by using as many good analog outboard (inserted) as we can for processing. Expecially for dynamics. And assuming that digital processings ITB or OTB are pratically the same since a good algorithm works as fine on an outbord DSP as on an accel card (reverbs), the best solution could be having a good DAW, many good AD-DA and tons of good analog gears for eq and dynamic. Better than the scenario I've described at the beginning of this replay. I'm right? This thread has become very interesting. ranieri senni |
Paul Frindle wrote on Sat, 21 May 2005 17:55 |
... This removed a whole artistic dimension from the engineers of the time - and of course they found themselves feeling uncomfortable... |
innesireinar wrote on Sun, 22 May 2005 11:43 |
Thank you Bob. It's easy to understand that working at 96 it's better, but why at 48? Is there a particolar reason why 48 is better than 44.1? |
innesireinar wrote on Sun, 22 May 2005 11:43 |
I would like to know your point of view how are good rev plugins compared to hi-end outboard rev. Can rev plugs like Altiverb, Rewibe, TL space and also hi level no-convolution rev compete to, let's say, a System 6000? Thanx |
Bob Olhsson wrote on Sun, 22 May 2005 15:19 | ||
|
Quote: |
The problem with solid state gear was that the usable dynamic range was way way less. Yes it compressed when you pushed it but in fact we generally weren't pushing it. While test tones across every input compressed, dynamic impulses hitting one stage didn't because tube console had one gigantic power supply for the whole shooting match. |
bobkatz wrote on Sun, 22 May 2005 16:51 | ||
With A REAL GREAT A/D the difference between 44 and 48 is extremely small, perhaps inaudible. This is because of the quality of the low pass filtering and other aspects of the design. An exceptional low-pass filter would have thousandths of a dB of ripple in the passband or less, be calculated at very high precision and dithered cleanly to 24 bits. You only get that in a "roll your own A/D" and the number of manufacturers actually rolling their own filters may be counted on a couple of fingers of one hand! <snip> BK |
Bob Olhsson wrote on Sun, 22 May 2005 20:26 |
What I'm trying to say is that tube consoles had an astounding dynamic range for something like a kick drum. You could just pull the fader back on that channel without padding the preamp input. |
Paul Frindle wrote on Sun, 22 May 2005 18:47 |
24bit digital audio has 144dB or so total dynamic range - so you can easily provide more than 30dB of this kind of headroom before the digital signal noise becomes significant in respect of the DAC SNR. |
compasspnt wrote on Sun, 22 May 2005 18:17 |
...If this level is ok, then most people are digitally recording about 12 dB "too hot" most of the time (assuming a proper, non-sample based meter). This lower level would certainly negate many of the digital/plugin overload problem! |
compasspnt wrote on Sun, 22 May 2005 19:17 | ||
Paul, Would you then say we could record (in a good quality DAW with good converters) at, say, 12 dB below "red light?" Would there be any other trade off penalty (something like "using all the bits" which some people talk about)? If this level is ok, then most people are digitally recording about 12 dB "too hot" most of the time (assuming a proper, non-sample based meter). This lower level would certainly negate many of the digital/plugin overload problem! |
compasspnt wrote on Mon, 23 May 2005 00:17 | ||
Paul, Would you then say we could record (in a good quality DAW with good converters) at, say, 12 dB below "red light?" Would there be any other trade off penalty (something like "using all the bits" which some people talk about)? If this level is ok, then most people are digitally recording about 12 dB "too hot" most of the time (assuming a proper, non-sample based meter). This lower level would certainly negate many of the digital/plugin overload problem! |
Bob Olhsson wrote on Sun, 22 May 2005 12:26 |
What I'm trying to say is that tube consoles had an astounding dynamic range for something like a kick drum. You could just pull the fader back on that channel without padding the preamp input. I haven't found outboard tube preamps or small tube mixers to have this "effortless" quality and they indeed do compress. I can only assume the limitation is the power supply. The "vintage" tube sound people hear on old recordings was not very compressed at all. I didn't completely appreciate this until I used Deane Jensen's prototype servo mike pre around 1985 which also had a whopping dynamic range. |
Paul Frindle wrote on Sat, 21 May 2005 18:09 |
As people might imagine - increasing sample rates can never redress these problems - that's an open ended quest of diminishing returns which has potentially open ended costs for the user - that is doing nothing more than 'pasting over' far more simple issues that essentially cost nothing to fix - other than a bit of honesty and a change of attitudes within the industry. |
blairl wrote on Mon, 23 May 2005 18:23 | ||
Hi Paul, Would it be possible for DAW manufacturers to implement oversampled meters by default in their systems? I can imagine it will take processing power away from other functions in the DAW, but I have no idea how much. Would oversampled meters take too much processing power? If everyone had oversampled meters to begin with which accurately represent signal rather than samples, we might avoid many of the exisiting problems. I realize that a quick workaround is to just back off on levels, but I don't think the majority of users are being educated, or are even willing to do such a thing. |
Paul Frindle wrote on Mon, 23 May 2005 06:14 | ||||
Yes in essence this is the point. But it isn't a good idea to lose too much gain in the input stages of the recording ADC since you will lose SNR. It is perfectly permissable to record in the first instance at relative high levels (peaking around -3dBFS) because an illegal signal should not come out of an AD converter (as we have said here). The place where you need to make the gain loss to get headroom and avoid unreported overs is in the DIGITAL domain - right at the start of your mixing channel. In this way you preserve the converter's SNR during recording and optimise SNR and headroom for the whole system - making maximum use of the 144dB SNR the digital domain offers. The only provisor is that you will need to run good quality plugs that are not noisy and function correctly at lower reference levels. |
Ronny wrote on Mon, 23 May 2005 20:22 |
... Snare peaked at -48dB, kick at -51dB, toms and over heads were between -45 and -52dB and the guitars were around -38dB, peak not RMS... ...I can only say that there is even more footroom at 24 bit than I thought. |
David Schober wrote on Tue, 24 May 2005 15:29 |
Paul, I want to add my thanks for your contribution to this thread. This is a real eye opener for me. A couple of questions: In regard to your experiment of white noise and filters, I also tried a HP filter at around 30hz and found similar results. A net increase in level when employed. The question is, where in the audio spectrum is this happening? Globally i.e. all frequencies or does it tend to hover at certain ones? Here's why I ask: This thread got me to thinking about my own mixes and that lately I've been getting a consistent result of more bottom on my mixes than I expected. The reason it caught my attention is that I had placed a GML on the master with a slight bump up top and a filter at around 30 Hz. So...I tried your experiment with the GML. I set the white noise at -3 dB. The first thing that happened is that when I called up the GML was that the level jumped up to -1.4. When I engaged the HP @ 30hz it moved up another .5 dB to -1.1! If I hit the bypass button or disabled the plug it went back to -3. If I left the eq in, (not in bypass) but bypassed the individual eq sections it still didn't make a change, remaining at -1.6. Any idea why this happens? Is there something about the GML plug that adds gain, although I haven't noticed it when mixing? I know the GML isn't your baby and you're no doubt reticent to make negative comments about the GML plug. If GM is around he could answer this, but I suspect he's still got his head in studio construction. Any thoughts? Thanks again for your time. |
Luke Fellingham wrote on Tue, 24 May 2005 18:23 |
Paul, Thankyou for your input into this thread, I have found it very educational. It has raised a couple of quistions for me. I guess this might be a bit basic but will sample rate conversion show up illegal digital signals in the same way as high and low pass filters? I have recently completed a project which I had to convert from 48k to 44.1k. Before conversion it peaked at -0.3db, afterwards there were (occasional) peaks at 0db. If I were to redo this, is it just a case of backing off the limiter/lowering the level until the peak value of the sample rate converted version is the same as that of the original? Having read this thread I'm am getting the meassage about leaving a safe amount of headroom especially before conversion back to analogue. However, working as I do in a native DAW with a floating point mixer and mainly floating point plugins, will going above 0db in the signal path matter providing the level is back down again by the time it hits the output? To be specific, I have the Sony Oxford eq plugin for Powercore which uses fixed point maths, when the clip light comes on I can generally hear it and lower the levels accordingly. With the native plugins I cannot hear the sound deteriorate however high the level goes provided it comes back down before output. Have I grasped this correctly or will this way of looking at things potentially make my sound suffer? Luke Fellingham |
David Schober wrote on Tue, 24 May 2005 19:16 |
Thanks for the response Paul. One other oddity, when running white noise and no eq (or bypassed) the colored bar level indicator is nice and stable...just as one would expect with any steady tone. But, engage the GML and watch the level dance. I don't know why nor can I imagine why the insertion of a high quality eq plug would cause level instability. I tried several other eqs, Oxford, Filterbank, Waves Ren and normal, plus Focusrite. All of them would do this dance when the filters were engaged, but only the GML would do it just sitting there, supposedly idle. Again, I know this isn't your beast, but maybe you or someone else has an idea or GM might pass by. The question I'd have is this. If I see the peak meters dancing like this with white noise, am I to conclude that this dancing motion is going on all the time on the channels I'm mixing? Cheers |
David Schober wrote on Wed, 25 May 2005 13:16 |
Paul thanks so much! What a great response! This is the kind of stuff is what can make a forum so worthwhile. Thanks for taking the time out of your day to help us out. I've been at this a long time and this subject is somthing I'd never heard before. I would see it's effects, but didn't know what was going on. For instance, when I'd engage the Oxford eq with filters and engage a HP filter, the overload light would sometimes appear. I had no idea why then, but now I understand what's going on. So from what you said, is it reasonable to generalize that an eq which makes the meters dance would be an oversampling EQ? Thereby concluding the GML EQ is an oversampling EQ and the others, including the Oxford are not? |
RKrizman wrote on Tue, 31 May 2005 22:35 |
Incredible reading. Thanks, Paul, for your generosity. As a workaday PT guy I have some questions. In light of all this, what's the best specific methodology to follow.? It seems that if a (typically prefader) eq plug is creating illegal overs then pulling down a fader after the eq is too late, right? Does the material really need to be recorded in at a lower level to begin with? Or do you need to insert a trim before the eq's? Thanks, Rick |
Paul Frindle wrote on Wed, 01 June 2005 18:56 |
The overload thing - if the illegal overs have not caused numerical saturation (i.e. no red lights on normal meters) then moving the fader down WILL recover it correctly - providing you wind the gain down before hitting another plug like an EQ. |
Paul Frindle wrote on Wed, 01 June 2005 18:56 |
The overload thing - if the illegal overs have not caused numerical saturation (i.e. no red lights on normal meters) then moving the fader down WILL recover it correctly - providing you wind the gain down before hitting another plug like an EQ. |
Quote: |
So then in Protools, for instance, where the plugins are prefader normally, it would be wise to put a trim as your first plug and wind it down before hitting the eq's if the recorded signal is particularly hot. Right? |
Quote: |
I guess I'm trying to translate the theory into specific methodology. "In light of Paul Frindle's analysis, here's what you should do to make sure your mixes ITB have depth and beauty...." |
Paul Frindle wrote on Wed, 01 June 2005 18:56 |
...Thanks for the encouraging words - I am just really glad that I may have managed to help people a bit |
TheArchitect wrote on Tue, 09 August 2005 18:01 |
Allow me to restate what I think I have read here. Input levels while recording should be in the -15 to -10 range ballpark on peaks. Ideally, similar levels per track (plus or minus as needed of course) on playback with any needed make up gain on the output bus peaking in the -7 to -5 range to leave room for the mastering process. I'm thinking even those levels may be a little hot if I am comprehending things correctly I guess I am still a little confused on where the problem actually is occurring, at the A/D conversion coming in or in the summing process in the DAW? |
Mark du Plessis wrote on Wed, 10 August 2005 11:14 |
Sorry all if I say something that's already been said, but I joined this one late. I used to think Digital was a swear word (unless used to describe the area BK inhabits, which is cool, of course!). I have had a fair number of decent analogue consoles in my short studio-owning career. A Calrec broadcast console, an Amek Rembrandt and an MTA 980 with Uptown. All into PT or Nuendo thru Apogee. All nice and warm, etc... Then thru a bad accident at the studio I was kinda "forced" by the insurance to get a digital (aaaggghhh) console. A nasty Neve Capricorn. By George (the king, not the engineer/producer) was I surprised. What a difference good digital makes. My mixes have a depth and clarity I had not achieved before. They had a back-to-front-ness which ,even with the Amek (which I loved), I struggled to achieve. I can (and I swear this to be true)even hear dirt on PT plugs. Add to that the ability to recall a complete mix, and you're smiling. I agree with BK in this - digital done well rocks!! I have done all sorts of mixes on it, Rock, metal, Hip-Hop and RnB, dirty, filthy, 60's sounding squished stuff and it has just made my choices easier. The EQ may not be as sweet as the Amek or crunch as much as the MTA, but, hey I can just load another plug into PT, knowing that the Cap won't be doing any major colouring. Cool! Mark |
Ronny wrote on Wed, 10 August 2005 10:51 |
Plus the eq's, dyn's, gates and fx aren't plug-ins they are built in processors. No latency... |
Ronny wrote on Wed, 10 August 2005 00:39 |
Not too hot for me. As long as the peak gain isn't above -0dBFs it doesn't matter how high the peak level is, because I'm going to attenuate the input of any processor that will boost gain at output, OTOH when you record peaks at -15dB, mix at -7dB and your material gets mastered, the noise floor of your mics, mic pres and mixing console, is going to be raised by at least -15dB when the ME sets final perceived gain. This may be inaudible on some tracks and not so inaudible on others, for example if you raise the inherent noise of a U87 which is -82dB (cardioid pattern) by 15dB on a vocal track, your final floor will be -67dB, while that's not bad and around the floor of good annie tape, if you record a single coil pickup guitar with a high noise floor amp, typical guitar amp buzz or a guitar DI box or fx processor that is outputting -40 to -30dB of noise due to the cheapo DAC's and the gain gets raised in mastering by +15dB to make up gain, it's going to be quite audible and introduce more noise in your final songs than if you would track close to peak, but without going over. I only track live concert orchestra at -12dB, but for studio work where it's not a get it in one take or else scenario, there isn't any reason to record peak at -15dB, that's too much headroom for some instruments and will raise the noise in your tracks. You don't have to squeeze every bit on the 24 bit A/D conversion that many people used to advise with 16 bit, but the lower your peak gain is at the ADC, the higher the noise floor of the mics and instruments will be raised on the final master. |
blairl wrote on Wed, 10 August 2005 16:33 | ||
Some plug-ins don't sound very good. Some plug-ins sound amazing. A digital console running a DSP process is nothing more than software interfacing with hardware to create the desired process, (sounds a lot like a plug-in). There is latency involved in all DSP processes, however, dedicated DSP processors such as those included in stand alone digital consoles and Pro Tools TDM systems can have lower latency than native powered DAWs. There seems to be a theory out there that DSP processes in a dedicated console are inherently better sounding than all plug-ins. This isn't entirely true. If you have excellent code interfacing with the right DSP you can get excellent results. As an example, the folks at Sony Oxford have brought the same processes included in the stand alone Oxford digital console to a plug-in format. If you ask them they will tell you that the plug-ins are the same processes as the original processes found on the Oxford console. They are not watered down versions with lesser fidelity. They are the same. George Massenburg makes an EQ plug-in that he has said he likes better than his analog original. There are many more great plug-ins out there. |
Ronny wrote on Wed, 10 August 2005 17:21 | ||||
I'm not knocking plug-ins, but they are 3rd party items, they don't always work well with all DAW programs, whereas the digi console is designed from the ground up, processors included to incorporate with the design. Also, there is always some latency, even with an analog processor, but when I speak of latency with digi console processors, I'm speaking so low as to be inaudible, quite the contrary with many plug-ins on a DAW platform. You also have Waves processors on slot cards, so you have some of the DAW plug-in options as well. You have cheapo digital consoles that don't have as good sounding processors as some high end plug-ins, so it's not one is better sounding than the other, both platforms have varying degrees of sonic integrity, typically but not always, associated with price. The major benefit is no overpowering the CPU and having hardware faders like an annie console instead of mixing with a mouse. World of difference. Couple a dedicated to audio only digi console, with a dedicated to audio only HD-R and you have the most reliable digital recording system available, much more reliable than recording on a PC or Mac system that isn't audio dedicated and has instances of other programs always running in the background. |
Ronny wrote on Tue, 09 August 2005 23:39 | ||
Not too hot for me. As long as the peak gain isn't above -0dBFs it doesn't matter how high the peak level is, because I'm going to attenuate the input of any processor that will boost gain at output, OTOH when you record peaks at -15dB, mix at -7dB and your material gets mastered, the noise floor of your mics, mic pres and mixing console, is going to be raised by at least -15dB when the ME sets final perceived gain. This may be inaudible on some tracks and not so inaudible on others, for example if you raise the inherent noise of a U87 which is -82dB (cardioid pattern) by 15dB on a vocal track, your final floor will be -67dB, while that's not bad and around the floor of good annie tape, if you record a single coil pickup guitar with a high noise floor amp, typical guitar amp buzz or a guitar DI box or fx processor that is outputting -40 to -30dB of noise due to the cheapo DAC's and the gain gets raised in mastering by +15dB to make up gain, it's going to be quite audible and introduce more noise in your final songs than if you would track close to peak, but without going over. I only track live concert orchestra at -12dB, but for studio work where it's not a get it in one take or else scenario, there isn't any reason to record peak at -15dB, that's too much headroom for some instruments and will raise the noise in your tracks. You don't have to squeeze every bit on the 24 bit A/D conversion that many people used to advise with 16 bit, but the lower your peak gain is at the ADC, the higher the noise floor of the mics and instruments will be raised on the final master. |
Extreme Mixing wrote on Wed, 10 August 2005 17:33 | ||||
Ronny, I think you have a lot of bad information concerning how the noise floor works in your post. What possible difference does it make if you increase the gain of an 87 or a single coil guitar before the converters of after? The noise floor would stay at the same level in relation to the balance in the mix either way. As long as you have enough bit depth to capture the full signal down to the noise floor, nothing will improve from printing hot. Steve |
Ronny wrote on Wed, 10 August 2005 20:09 |
The analog mic pre or guitar amp in this case, as all analog devices do, operate at an optimal range, so now we we set the recommended input and output on it, noise floor is lowest on that device as we don't have to crank it because the ADC is peaking low and we send that optimal signal to the converter with peaks at -1dB, we don't have to turn up gain, because it's already there, the noise floor doesn't go up. |
blairl wrote on Thu, 11 August 2005 11:43 | ||
(See message #82389 in this thread.) OK but how do you get the guitar to peak at -1dbfs? Self noise measurements of analog equipment have been taken at standard nominal levels. If in setting your record levels you deviate from these standard measurement settings then the noise floor of analog components like guitar amps and microphones will change as well. If you have set the amp at it's optimal level and you have set the mic pre at it's optimal level (+4dbu=1.23 Vrms) and you have the AD converter calibrated to match the optimal level of the mic pre, (between -20 and -18 dbfs = 0VU), and the guitar player is playing at an average dynamic level, you will naturally be peaking at around -16 to -14dbfs. To peak higher than this you have to turn up the volume somewhere. You would have to raise the gain either at the amp or at the preamp which would change the level at which the self noise measurements were taken and would raise the noise floor of the amp. Raising the gain above optimal levels pre ADC or post ADC would raise the Vrms level and noise level equally. There is no difference. If you were talking about low recording levels being too close to absolute noise floor levels of things such as 16 bit dither or quantization noise then I would understand. Raising the gain of a 16 bit source with low recording levels would be a concern. Not because of the noise level of mics or instruments, but because of the absolute noise floor of a 16 bit source. With 24 bit recording, this is no longer a concern. |
Ronny wrote on Thu, 11 August 2005 14:58 |
I think you guys are missing my point. |
Quote: |
...you may be surprised at how many people don't optimize the analog side... |
Quote: |
If you record a guitar using a DM2000 for front end, optimize the amp and it's -16dBFs, you can turn the mic pre's up to peak at -1dB with no degradation, because they are designed to output the signal at -0dBFs. |
blairl wrote on Thu, 11 August 2005 19:04 | ||||||
I think that perhaps we are not understanding each other.
Absolutely, this is a problem.
I think this statement is where we are not understanding each other. They way I see it is that when the signal to noise ratio of analog equipment is measured, it is done according to a standard where all analog equipment used to measure the noise is set to +4dbu. If you turn the mic pre up beyond +4dbu at this point to make the ADC peak at -1db you have thrown out the standard and when you bring the mic pre up you are bringing up the noise floor of the guitar amp with it. Whether you bring the guitar amp level up before the ADC with the mic pre or you bring it up after the ADC digitally, it is the same. The noise floor of the guitar amp that was properly optimized for noise is not the problem when it comes to bringing the level up in mastering. It is the noise floor of the 16 bit or 24 bit recording that could be an issue. Again to clarify. Boosting the level beyond +4db with a mic preamp to make it peak at -1dbfs at the ADC will bring up the noise of the guitar amp exactly the same amount as if you were to leave the mic pre at +4db, having the ADC peak at around -14dbfs and then boosting the level digitally after the ADC to peak at -1dbfs, either in mixing or mastering. Try an experiment to verify this. Set up a guitar amp and optimize the levels. Set your mic pre so that the the ADC is peaking around -14 dbfs with someone playing guitar. Record the ambient noise with nobody playing. Now boost the level digitally +13 db and look at the noise floor on your meters. Make a note. Now go back and set your mic pre to peak at -1dbfs while somebody is playing guitar. Record the ambient noise with nobody playing. Look at the noise floor on your digital meters. If the guitar player was playing the at the same dynamic level when you set the different levels then the noise floor of the guitar amp will be the same in both cases. |
Extreme Mixing wrote on Fri, 12 August 2005 10:57 |
Ronny, I agree that sounds should always be recorded with as little noise as possible. What I don't agree with is your statement that recording the guitar hotter in your example above, would have made any difference in the signal to noise ratio. Both the guitar and the noise would be louder. The point is that you have to do something to push the noise floor lower in relation to the guitar. Simply cranking the output of the mic pre won't do that. You have to change the pedal, examine the ground issues, eliminate RF--do something, but just printing hotter will not advance the cause at all. Steve |
Ronny wrote on Fri, 12 August 2005 16:37 | ||
You and Blair are both correct, however the fellow that started the thread said that he was peaking at -15dB and sending mixes to the ME at -7dB and was asking if these levels were correct, as I mentioned nothing is written in stone, but we all agree that opimizing every gain stage or at least as close as you can get will give the best results and my advice still stands that he doesn't need to give himself that much headroom. He needs to hear how loud the guitar noise is in that example and be aware that it's going to increase per ratio of guitar note gain on the final master. Absolutely the best solution is to take care of noise at the source. Speaking of single coil guitar hum, you can often eliminate that by either getting out of the EMI field from the amp, which extends about 4 feet out the front and back of the amp in an elliptical pattern that looks much like the magnetic flux lines when you put filings on a bar magnet in science class. If the room is too small to do this, the guitarist can turn the neck of the guitar 90 degrees perpendicular to the amp face and reduce the hum by mucho decibels, because the guitar strings are basically antennae that pick up the field, run the neck parallel along the flux lines and you'll hear an immediate improvement in signal to noise ratio. It all starts in the tracking doesn't it. |
Paul Frindle wrote on Fri, 12 August 2005 19:51 | ||||
I have been though this discussion before (not least of all when doing the input stages of the R3) it's a headache to work through because of different perceptions of what constitutes levels within the analogue and digital domains - of course they are the same and IMLE the best way to illustrate this is in terms of dynamic ranges. In this case we need to think of dynamic range as the difference between the noise and the max signal modulation. For the Guitar amp this is the noise and hum versus the sound pressure produced when the guy plays. If we assume this to be 80dB for argument: We then have the Mic which is almost certainly miles better than the amp - producing a signal level that easily stretches comfortably between the hum and noise of the amp and the level of the guitar playing - so probably very little loss of dynamic range here - we still have 79dB or so. We then hit the converter - the converter has a dynamic range of say 110dBr in the digital domain - so in order to avoid introducing excessive converter noise we must modulate it to at least -30dBr or so. So in this case there's plenty to play with and you can get it badly wrong without any damage (even modulating at -20dB or less) cos the source has such a restricted dynamic range. But if we actually modulate the converter optimally so that the guy playing hard gets nearly full level out of the converter, the dynamic range of the track is still around 79dBr - in the digital domain within the mixer. Now almost whatever you do within reason with gains in the mixer this will remain unchanged, since increasing it will make it clip immediately - and you can reduce it by as much as 60dB (and bump it back up again) without hitting the 24bit noise floor of the system. So (apart from deliberately using EQ to boost the line hum freqs) the only thing you can practically do to DECREASE the dynamic range of that signal (i.e. make the amp hum and noise proportionally louder wrt to the guy playing) is to compress or limit it - i.e. dynamically make the soft bits louder and/or the loud bits softer. A compressor/limiter reduces dynamic range by definition - and that's why the hum may get louder in the mastering stages, as the guy struggles to achieve 'commercially acceptable' levels of 'fashionably continuous' modulation. So if the mastering guy put on (for an unreasonable illustration) say 20dBs of total compression - if the guitar was a solo instrument peaking flat out and there was no other music or significant noise in the track, when the guy stops playing the amp noise will wander up from -79dB to -59dBr, i.e. change from being just acceptable to annoyingly loud. BTW, obviously this would also occur if you put this much compression on the track within the mixer. |
Extreme Mixing wrote on Sat, 13 August 2005 01:47 |
So if I follow you correctly in your example, Paul, it is the compressor pulling the noise floor up by 20 db, Printing the guitar part hotter, or closer to full scale would not change the outcome on the noise at all, because the guitar is going to sit where it needs to in the mix, and it will be hit by the compressor with the same result either way. That was really my only point of disagreement with Ronny. Thanks for putting the science in a way that makes it easy to follow the facts. Steve |
Quote: |
Ronny, I think you have a lot of bad information concerning how the noise floor works in your post. What possible difference does it make if you increase the gain of an 87 or a single coil guitar before the converters of after? The noise floor would stay at the same level in relation to the balance in the mix either way. As long as you have enough bit depth to capture the full signal down to the noise floor, nothing will improve from printing hot. Steve |
Extreme Mixing wrote on Sat, 13 August 2005 06:47 |
So if I follow you correctly in your example, Paul, it is the compressor pulling the noise floor up by 20 db, Printing the guitar part hotter, or closer to full scale would not change the outcome on the noise at all, because the guitar is going to sit where it needs to in the mix, and it will be hit by the compressor with the same result either way. That was really my only point of disagreement with Ronny. Thanks for putting the science in a way that makes it easy to follow the facts. Steve |
Paul Frindle wrote on Sat, 14 May 2005 16:34 |
Yes it is quite easy to make a meter plug-in that reconstructs the signal and shows you the value (I have made such processing as integral parts of other applications) - and you could put it where you like in the mix - i.e. the main outputs for a start. It is fairly costly in terms of processing load though, so you certainly wouldn't want to pepper such a thing all over the tracks in your PT mix!! |
geosound wrote on Tue, 13 September 2005 03:46 |
hi nathan i did a test with a mackie 1604 and mixing itb it contain 6 tracks and no eq or plugs al faders at odB conclusion the sound that came out the mackie was more punchier and every instrument has more detail and better heard individually so summing on a cheap desk will get better results also what i discovered on a bigger desk soundtracs jade (easier to mix less time consuming goes also for a cheap crappy desk like the mackie. i wouldn't know for soundcraft but in my opinion the low cost range of soundcraft is useful for everything except for audio georges sterkenburg |
Ronny wrote on Sat, 13 August 2005 03:11 |
]OTOH when you record peaks at -15dB, mix at -7dB and your material gets mastered, the noise floor of your mics, mic pres and mixing console, is going to be raised by at least -15dB when the ME sets final perceived gain. |
RKrizman wrote on Tue, 13 September 2005 13:11 | ||
Yes, but it will also be the case even if you record at -7 db to begin with. Whatever level you record at, when you mix you'll put that guitar at a certain level. Perhaps you'll record at minus 15 and boost to, say, minus 10. Or perhaps you'll record at -7 and then attenuate to -10 when mixing. In either case you get the same guitar track, with the same signal to noise relationship between the buzz of the amp and the guitar signal. In either case, when you go to mastering and the levels are raised the result will be the same. So there's nt signal to noise benefit to be had by recording hotter, (unless you're talking about the noise from the input converters themselves, which you're not) Either way you're capturing the same picture. The exception, of course, is if you record too hot and distort your analog components somewhere along the line, or don't give yourself enough headroom to do further manipulations. -R |
Nathan Eldred wrote on Tue, 13 September 2005 16:20 |
I'm not surprised really one bit. I'm sure ultimately really great mixers might be able to get better than a Mackie results ITB, but they are stemming to high quality outboard,etc so it's not the same as plugins by any stretch of the imagination. I've heard A/B comparisons between ITB and on a high level console from the same guy (he has talent). While his ITB mixes were better than 80% of everybody out there, his console mixes were 90% better than everybody out there. Just my opinion. |
Bob Olhsson wrote on Sun, 18 September 2005 17:27 |
Why limit a mix at all? Unless you have the song before and the song after up at the same time, there's no way to know what final level it needs to be at or how much peak limiting to use. Even when you do have them available, album sequences can often change. |
Ronny wrote on Tue, 13 September 2005 14:30 |
"when the ME sets final perceived gain." Limiting raises noise floor relative to peak gain. Anytime that you decrease crest factor you raise noise floor. I would assume that everyone knows that if you record at -10dB and turn the "volume" up, that the noise floor goes up 10dB, no arguement there. The ME is typically going to "alter dynamics and decrease crest factor" and this is where several of you guys are missing it. |
RKrizman wrote on Sun, 18 September 2005 20:20 | ||
That's right, but the results will be the same whether you recorded soft or loud. -R |
Bob Olhsson wrote on Sun, 18 September 2005 16:27 |
Why limit a mix at all? Unless you have the song before and the song after up at the same time, there's no way to know what final level it needs to be at or how much peak limiting to use. Even when you do have them available, album sequences can often change. |
Ronny wrote on Sun, 18 September 2005 22:43 |
[The point that I'm making is if the tracking and mix engineer is recording and mixing too low, he's not getting an accurate picture of where the final noise floor is going. If he peaks higher and mixes higher, he will hear more of the noise floor and will realize that he should have taken care of the noise "before" the tracking, not at the mastering stage. If I had a dime for every client that said, "wow, I didn't know that the guitar effects or 60 cycle hum sounded that noisy" until it went to mastering and had perceived gain processing, I'd be a dimondaire. |
RKrizman wrote on Mon, 19 September 2005 13:45 |
Hey, I decide my playback level with my monitor volume control. I don't need to track something louder to hear it louder. -R |
RKrizman wrote on Mon, 19 September 2005 15:45 | ||
Hey, I decide my playback level with my monitor volume control. I don't need to track something louder to hear it louder. -R |
Denny W. wrote on Tue, 20 September 2005 11:17 |
How loud the monitors are is not what I believe he was driving at. What I believe he is saying is that the relative level of the noise floor to source material will be greater if recording the signal at -5db vs -15db |
RKrizman wrote on Tue, 20 September 2005 11:47 | ||
If he's referring to the noise from the source, then he is incorrect. -R |
Bob Olhsson wrote on Sun, 18 September 2005 18:22 |
In a documentary you still don't know how loud or dynamic the music needs to be until the final mix happens against picture. |
Quote: |
My point is why guess? Unless you luck out and guess exactly right to a tenth of a dB., the end results will almost always be more distorted. |
lukejs wrote on Sun, 25 September 2005 20:53 |
I've seen that many of you have decided to mix outside the box.... I have an Apogee AD16-x and am considering purchasing the DA16-x in order to mix on my Allen and Heath GS3000 recording console.... Should I go for it ?? Any of you heard anything about the quality of the Allen and Heath GS3000 console and would the expenditure on the DA converter be worth it ? I heard that only the best Analogue will beat digital... will this console suffice ? I appreciate any advice !! thanks, Luke |
bobkatz wrote on Sun, 25 September 2005 21:19 | ||
I have a client who uses the Apogee AD16 and DA16 with his Soundcraft Ghost and gets excellent results. I can't imagine the Allen and Heath being worse than the Ghost... Neither one has much of a pedigree sonically. The key to my client's good mixes is his use of good outboard gear with the console along with good ears, of course, and relatively little use of the onboard EQ in the Ghost. BK |
button wrote on Sat, 01 October 2005 18:22 |
Recording digital and using analogue desks and processing to mix is a happy compromise ... a musical moment preserved for the consumer, >thanks< to digital mastering and delivery mediums. |
compasspnt wrote on Sat, 17 December 2005 12:59 |
In my opinion, there is much less chance in that case of digital harshness, than if you had it constantly peaking at '0,' and then had to push the fader down. |
Phi Lion wrote on Sat, 17 December 2005 11:44 |
Lets say i have a track that was not recorded hot at all say -40 db and during the mix in order to be able to hear it in the mix i have to push the fader all the way up. Will this introduce digital harshness. |
Bob Olhsson wrote on Sat, 17 December 2005 11:22 |
There's no question that mixing itb requires a very different work-flow. Because it's so much more "left brain," it's best done for brief periods of time with lots of breaks. Being able to recall everything becomes an absolute necessity because getting at your gut reactions is so much harder. I've been absolutey amazed by how good stuff that peaks to -20 sounds when it's 24 bit files. Don't just take our word for it, try it yourself. Check your digital signal path out with tones including all the plug-ins. Some of your favorites might sound worlds better with some extra headroom. A DAW mixer is still a console with all of the issues of a console. |
Bob |
There's no question that mixing itb requires a very different work-flow. Because it's so much more "left brain," it's best done for brief periods of time with lots of breaks. Being able to recall everything becomes an absolute necessity because getting at your gut reactions is so much harder. |
Quote: |
A DAW mixer is still a console with all of the issues of a console. |
Paul Frindle wrote on Tue, 20 December 2005 10:07 | ||
Exactly so - well said |
cerberus wrote on Tue, 20 December 2005 23:07 | ||||
You appear to reject floating point mixers out of hand. There are such DAWs whether you sanction them or not. So I think the assertion is incorrect in this context. jeff dinces |
Paul Frindle wrote on Tue, 20 December 2005 20:45 |
For instance - quite apart from all the discussion about overloads and reconstruction errors - IMO, much of the reason people perceive that sound quality increases so dramatically when dropping levels by as much as -20dBr before mixing, is precisely because they are relieved from the distraction and disruption of permanently watching for 'red lights' and continually destroying their balances (and artistic concentrations) by trying to avoid them. The effect of this factor alone on the flow of a mixing session is immense IMLE - and it is only one issue of a great many facing people who have to mix entirely 'in the box'. |
Paul Frindle wrote on Tue, 20 December 2005 10:07 |
(Digital Workstation)...internal mixers have evolved from simple apps (limited by processing restriction) needed to manage HD storage systems, not necessarily with the advantage of all the knowledge aquired by the mixing console fraternity over the decades. Interaction between the kit and the user is a very subtle thing and it's not always obvious to people why they get better results under some circumstances - this leads to speculation of all sorts. When people report better results from external analogue mixers a large part of it IMO is the artistic facility of having interactive tactile involvement along with the immediacy of analogue control. |
Ron Steele wrote on Wed, 21 December 2005 13:07 |
It's been so long since I grabbed any kind of faders, that I often wonder, if it would make a difference how I would balance a mix just because of the actual physical "touch". Any more thoughts on that? |
compasspnt wrote on Wed, 21 December 2005 02:23 | ||
Excellent point, Paul. There are few, if any, tactile surfaces which give the harmonious feedback to the operator that a "good old" analogue desk does (any of several makes). |
howlback wrote on Thu, 22 December 2005 18:12 |
Terry, Bob, and Paul, you would know better than me, but it seems that ITB mixing is much less a "group" activity than analogue mixing used to be (particularly before automation). I might be wrong, but my impression is that "back in the day" musicians would participate in the process on occasion, and that assistants would be involved in recording outboard settings, etc. The first record I made (analogue) certainly was more of a "group" thing than more recent projects ITB. This group dynamic seems to be a bit missing now-days, the process is bound to be different. |
Quote: |
I do understand that. It is only partially true that I don't need to "worry" at all about red lights until just before the converter, I still need to be able to read all my meters; the ones on most of my plug-ins don't tell me much after they go red. Also the threshold on none of my dynamics plug-ins can be set to a value over zero, so that would be another barrier which keeps my particular floating point mixer acting like any other mixer in a real world situation. But I think that is more about the way the plug-ins I use are designed, I am not sure that mixer itself sounds any different or is even mathematically more or less internally precise when levels peak at -20dBfs than when they do at -.02dBfs; so i think gain variances within reason for float that would not be considered "ideal" in fixed shouldn't affect a float process in theory, but in practice they do. |
Quote: |
And I don't dispute that Bob is hearing what he says he is. It just sinks in that he chose to change his computer to a PC which is -said to have a poorer workflow for audio production than a Mac-. Rather than change his DAW, which I understand he feels he is required to purchase and update regularly in order to do his job. But we are talking about the interaction about sound quality and workflow, it is impossible not to look at every part of these systems and ask if there is a weakness we can solve.. |
Quote: |
I think that eventually the improvements for digital will be very incremental and subtle... just like newer analog, which has evolved and is mature. These kinds of differences in the math between fixed and float are very subtle, but I'd like to see you designers work out the science, the ergonomics too if you will, otherwise I think digital cannot overcome the prejudice it now deserves. Since I work in digital, I have to defend it to my clients "i hear tape is better, i am going to master with a tape guy.. because i get a real pro job then"..etc... Because analog is sanity. We can't even decide...you can't even say.. one kind of mathematical paradigm is better.. and they are different.. a lay person can grasp that i think. |
Quote: |
But why can't we agree on the kind of math we want to use? I think it's a big reason why ITB mixing and fully digital mastering are not taken seriously in some quarters. |
Quote: |
This could be one aspect... so maybe we can call this problem: 1. collaboration vs. isolation. here is another terry has mentioned: 2. the necessity to commit vs. the ability to tweak endlessly. paul has suggested that many factors could affect : 3. the pace and rhythm of the workflow. such as dfferences in headroom on different systems i notice two important factors which support his contention that there are "many" such issues : a. with tape, there is always a dramatic pause before playback (rewind). b. with a daw, one can loop a section; the lack of need for rewind can lead much faster to ear fatigue in my experience. then there is another factor we touched on: 4. making music could be considered a physical activity. for me, mixing in analog is more like football, on a daw, it's more like chess. the first three aspects are clear to me, but i think we could perhaps overcome them through better daw design and by changing our work habits. however, the fourth...if it is really a factor, means to me that analog mixing will have this advantage over daw mixing until daw latency becomes less than one sample. jeff dinces |
maxim wrote on Thu, 22 December 2005 00:24 |
i've never touched a console, yet i've mixed numerous cuts itb i'd hazard a guess that, for my brain, mixing on a console would seem an unintuitive process which would take a while to get used to i have no trouble grabbing a fader with my mouse, the brain kicks in automatically (after 10 years, it's not surprising) i'm dubious that the experience would be much enhanced if my finger was touching a real fader this is aside from the fact that it would render my studio immobile |
maxim wrote on Fri, 23 December 2005 05:26 |
paul, these are very good points and i can see why my analogue desk trained friends and colleagues lament having to work on a daw however, i think, your analogy is a little stretched riding a fader requires a simple 'digital' movement, like rolling a trackball (which is what i use) it's not the same as getting tone out of a fretboard mixing itb is no longer a non-realtime experience (as it was when i started with deck on 7200 powermac) as it is now, in dp on a g3, setting balances and pans is a completely right hemisphere activity (for me) likewise, patching in eq, compression and delays is a fairly straightforward affair (more so than a desk, i would guess) the only thing i'm missing out on is being able to ride more than one fader at the same time (although submixing and automation can compensate for this) but i'm used to concentrating on one fader at a time how much left hemisphere was involved in syncing multiple desks, patching in reverbs etc, that has now been left behind? i think, it is a question of what you're used to |
maxim wrote on Fri, 23 December 2005 17:48 |
as i said, after 10 years of daily daw use, i no longer think about it, and the right hemisphere kicks in straight away the 'ghosts' between speakers materialise easily, and the control is fairly invisible the only time when the stop/start feeling arises is when i'm comping, but there again, i'm sure it would be even worse with tape |
NoWo wrote on Fri, 03 February 2006 18:21 |
what about all your pencils, papers, remote controls, favourite magazines, keyboards and mouse-pads? |
maxim wrote on Sat, 24 December 2005 09:48 |
i'm not certain about encouraging, but then, i'm not sure, how encouraging the console is either i know for certain that it would take me a while before it became intuitive as i said, after 10 years of daily daw use, i no longer think about it, and the right hemisphere kicks in straight away the 'ghosts' between speakers materialise easily, and the control is fairly invisible i do think it's all about the cerebellar feedback, like riding a bike the only time when the stop/start feeling arises is when i'm comping, but there again, i'm sure it would be even worse with tape also, i chose dp as my platform of choice, primarily, on the ease of use while mixing i think it's VERY important to have your system set up well, so the issues like cpu overload don't arise also, make sure all your submixes, auxilliaries, delays etc are set up before you start, so the right hemisphere is not distracted |
Keyplayer wrote on Sat, 23 April 2005 16:04 |
Terry Manning [Whatever Works] requested this thread be a sticky Have fun. F ------------------------------------------------------------ ---- ------------------------------------------------------------ ---- With all the debate over the supeior ease of automation in the DAW vrs that of most mixing consoles, I was wondering if anybody was actually using their DAW like a tape deck/editor and mixing from their consoles to a mixdown deck or even back to a stereo or 6 stem tracks on their DAW? I'm pretty sure those of you with access to Neve's, API's, SSL's etc are doing just that. But for those running in the "Mid-Line" (I.E. DM2K, R-100, Soundcraft Ghost etc.) are you doing this or letting the DAW do all the work and having your desk just act as a router? |
Daniel Asti wrote on Thu, 25 May 2006 15:42 |
He truly believes that his particular board is magical and has a special sound. I think that you need some mojo like that to make good mixes. Making something amazing requires confidence in your equipment and abilities. You also need a lot of hard work. In my opinion the most important tools are your ears and the will to do something amazing. |
yngve hoeyland wrote on Sat, 23 September 2006 17:13 |
Morale of the day? Stop looking at what you're hearing. |
compasspnt wrote on Fri, 25 May 2007 15:24 |
Hi Jim, Not knowing Cubase, I would say that it appears you are doing things better than many who track and mix ITB. The only real change I would make to your annotation would be to input the initial tracking levels about 6 dB lower than you now show. This may mean running your final analogue stage a bit lower, or calibrating your A>D a bit lower, or both. Once this is accomplished, the rest is virtually automatic, and little will "go wrong," level-wise. As for adding a limiter at the final stage, if not sending out to Mastering, yes, that is where you would do so. Bringing up level properly is a bit of an art, however, and in some cases you might be better actually doing so within two separate plug-ins, rather than one alone, depending, of course, upon the quality of the software (or upon how pretty the GUI looks). Ignore the GUI comment. |
jdvmi00 wrote on Fri, 25 May 2007 16:06 |
Thanks! So basically I should record at -12 to -15 peak instead of -6? I'm guessing that's because it will give me enough headroom to add in the plugs and stay within the -6 range for the entire track? |
Quote: |
a) The hands-on approach. I like doing two or three things at once, say autoing levels on lead vocals and bv's simultaneously. |
audio2u wrote on Sat, 13 October 2007 08:59 |
Wow! * Most informative thread I've read on an audio forum EVER. This has shed a lot of light on a whole bunch of things for me. |
noeqplease wrote on Fri, 21 March 2008 19:25 |
Yes I do still mix sometimes on my mixer. More umph. More blending mystery. More headroom. More pleasing distortions. Faster. Fuzzier...not as clear as the ITB thing. No weird low end and high end...weirdness. Cheers |
mustgroove wrote on Tue, 08 April 2008 06:52 |
The theories featured in this thread aren't just limited to Pro Tools, and apply equally to 32bit floating point DAWs, don't they? |
PaulyD wrote on Wed, 09 April 2008 00:19 |
Yes, they do. A thread related to this one can be found at the top of the Whatever Works forum. The entire thread is a worthwhile read, but if you want an explanation as to why this is, jump right to page 8 of that thread and read afroshack's post. Paul |
Elbowgeek wrote on Mon, 26 May 2008 08:17 |
If I can just jump in with the experiences of a recording engineer still very much with training wheels on... I asked an experienced engineer (he'd worked with some pretty big names) about recording levels, and his answer was simple: hot as possible, and don't be afraid if you get a few "overs" and redlines, but keep those to a minimum. But make sure you're using crap inputs; on my MOTU 896 workhorse I could push the levels regularly into red territory and get almost analog smoothness to the distortion, never any digital nasties. I just hope the Firestudio I replace it with will have the same qualities. Otherwise, I've learned a megatonne about the deeper aspects from the great, experienced members here - I'll be reading through this thread for a long while yet. Cheers |
noeqplease wrote on Mon, 26 May 2008 16:25 | ||
Just to put a wrench in ye olde "recording less hot at 24 bit" thing that's floating around. I always try to optimize the recording before it hits the converter, and try to hit a good enough leve that will allow the most dynamic range going through the converters. To me this means reaching around -3 on peaks, and is an occasional peak goes higher, then no worries. This usually ends up meaning I have an RMS level good enough to capture the sound as best as I can. What happens, IMO, is that I end up with a great recording, and then I can use the faders to set up the basic levels, instead of trying to the the very stupid (IMO) thing of trying to aim for all faders near unity level, and recording each instrument at different levels, ie. "mixing" by changing the recording levels. This is a great way to increase bad recordings, and decrease fidelity. Cheers |
PaulyD wrote on Sat, 14 November 2009 01:25 |
... Terry is a genuine good-guy and I'm sure his suggestion was a friendly one. He's just trying to help us all make the best records we can. |
Ron Steele wrote on Wed, 21 December 2005 10:07 |
It's been so long since I grabbed any kind of faders, that I often wonder, if it would make a difference how I would balance a mix just because of the actual physical "touch". Any more thoughts on that? |
thedoc wrote on Wed, 25 November 2009 21:12 |
Just a quick note on recording levels during recording to a digital medium: If you record too hot (such as -3dB below clipping) you may be actually clipping intersample peaks that you do not know about. Pulling down faders later will not cure clipped recordings. I know I hear an echo in here somewhere...sigh... |
Nick Sevilla wrote on Wed, 02 December 2009 13:11 |
Sometimes I get recordings done by other engineers, and I find myself actually gaining DOWN some of the individual instruments, because as as soon as I put a plug-in on there, it immediately just clips at any level I set the plugin to. I had this happen to an electric bass recording for a whole album. I guess no one was looking at the clipping red LED lights... |
compasspnt wrote on Wed, 02 December 2009 10:53 |
Bass especially can act very differently in a software recording programme than in an analogue console, level-wise. For those with a desk, check the level of a bass guitar on the desk meters, compared to the PT ones...it's almost always recorded too hot, even if the levels look fairly OK on the DAW. |
merrymerry wrote on Sat, 31 October 2009 21:39 |
In samplitude and sequoia in the "general" settings there is an option to lower all mixer strips by 6, 12 or 18 db!! |
Quote: |
Thats right... you can record so you are peaking at -3 db to get the fullest possible recording quality.... |
Quote: |
Bobkatz , maybe you could suggest this to your customers that use samplitude or sequoia? Warm regards, Merry. |
Geoff Emerick de Fake wrote on Fri, 26 February 2010 12:56 | ||
|
Tomas Danko wrote on Fri, 26 February 2010 07:24 |
And once more it's down to actual implementation. Most DAW engines are floating point 32 bit and most people record and store audio data at 24 bit. Whenever you shift it one bit to increase or decrease the gain by 6.02 dB you are moving the 24 bit set of data within those 32 bits, maintaining full data integrity even at extreme gain reductions. In theory you might lose LSB's, but in practice you usually do not. |
Mark D. wrote on Thu, 13 May 2010 07:49 |
Ideal A/D & preamp input levels can vary. But I've heard some going close to 0db at the A/D, or even preamp. At 0db input the preamp, with no further boost, peaks at -18db at the A/D. Lower preamp gain will reduce S/N of that & the A/D. Many pres have a 90db S/N, converters 110db. So unless it distorts, I can't imagine negatives in being near 0db at the pre. The question is if harm is done above -18db at the A/D. Driving a preamp input to distort...bad idea. But an output level control can bring what it sends to the A/D to, say, under -6db. Closer to the 110db S/N of the A/D. Well below clipping or intersample peaks, but with better S/N, which should be beneficial. I just want to hear thoughts that and experiences with that. (Edited & revised to clarify what I'd said.) |
jetbase wrote on Sun, 16 May 2010 19:16 |
But if I recorded everything peaking at -6dBFS, even if I had no distortion in the recording, I would most likely get distortion when mixing. |
Quote: |
When I get sessions in to mix & everything is recorded hot I gain individual tracks down so that the line ins on my console are not being overloaded. |
Quote: |
I'm not sure if there are negative consequences if mixing ITB with tracks peaking at -6dBFS. |
Quote: |
But anyway, if the mic preamp sounds better at a higher gain setting then isn't the solution to calibrate the input of the AD converter to suit? |
Geoff Emerick de Fake wrote on Tue, 18 May 2010 03:05 | ||||||||
In fact it is important to make a distinction between gain and output level, although they are related. Again, in a fixed-topology preamp, increasing gain generally increases distortion (by reduction of gain-margin), and the related increase of output level (for a given input level) increases the distortion created in the output stage (except for crossover distortion). But it is also very possible that a certain balance of input/output distortion be particularly euphonic. This should be your particular operating level for the preamp, to which you should match the A/D converter's operating level. The presence of transformers in the signal path may also steer to a different conclusion. |
jetbase wrote on Mon, 17 May 2010 20:25 |
Perhaps I could adjust the tape inputs on the desk so that they could accept higher levels from my DA converters (assuming that the DA converters themselves are fine with these levels), but then my analogue MTR will not be delivering the right level to the console. I also have to consider what levels the MTR & DAW will deliver directly to each other. |