![]() | All Advertisers |
| | #1 |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | About recording levels in DAW I've recently started recording all my projects peaking at 0dbv (-18Dbfs), and that has really paid off in terms of sound quality, ease of mixes, depth, width etc.. I find a massive difference in the quality of my mixes now, big revelation to me.. However there's one point that I'm not sure about; Mic preamps and most gear have their sweet spot, at which they perform best. Say if the best setting on my pre gave an output level superior to the 0dbv desired, should I reduce the virtual input channel from within the DAW(in my case Cubase 4)? My signal path is always the shortest possible; Mic->Preamp->Aurora AD->soundcard. I only recently found out that one could lower the recording level from within Cubase, but that's after the AD stage, so I guess it won't have the same effect as hitting the AD softer, am I right? What are the alternatives for that?
__________________ http://www.myspace.com/jeangillet |
| | |
| | #2 |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Maybe I should have titled this thread; How do you calibrate your converters? |
| | |
| | #3 |
| Lives for gear Join Date: Mar 2005 Location: Germany
Posts: 1,086
| I think as long as you don't hit your AD hotter than -6dBFs you should be fine with most converters. Then in the DAW I'd use a gain plugin as a first instance to bring down the level to your DAWs sweetspot, so in your case -18dbFs peak. |
| | |
| | #4 |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | I find even -6dbfs is too much for a full production with many tracks, you end up having to turn down all tracks too much, and it doesn't sound as good. |
| | |
| | #5 |
| Gear Guru Join Date: Mar 2005 Location: Long Beach, CA
Posts: 14,279
| Your DAW software itself (assuming it's a modern 32 bit float or better summing engine and no plug ins) should be very flexible with regard to level -- internally. Obviously, once you output to a DA interface or export/print data to a fixed bit depth file, you will have a potential for overs, including "stealth" intersample peaks. (You can watch out for those by using an oversampling intersample peak meter. [edit: looks like the info and download page for SSL's free intersample-aware oversampling peak meter is missing in action. I don't know it that's temporary or it's gone for good. Too bad if it is. They had a free one for a long time.]) Of course, that's just your DAW's floating point internal math. Where things get unpredictable -- and where a more cautious and 'conventional' approach to internal DAW gainstaging becomes important, is in the sends to plug ins. Some plugs may be "level agnostic" but others (perhaps most) will have sweet spots or perhaps use fixed point internal math that will not be as forgiving of overs. So, for that reason, maintaining a nominally optimized gain structure -- even in a DAW's floating point environment -- is usually best practice.
__________________ day job | A Year of Songs | music and social stuff | mutant pop on facebook | roots acoustic on facebook |
| | |
| | #6 | |
| Lives for gear Join Date: Mar 2005 Location: Germany
Posts: 1,086
| Quote:
So if the slightly hotter sweetspot of you preamp can't outperform the attenuation by a couple of dB of a trim plugin I'd stay with recording at 0dBV. But honestly, I don't believe so. But then I'm on Logic, where I haven't notice a quality drop when doing so. Andreas | |
| | |
| | #7 |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Your DAW doesn't care about attenuation...you will not hear any loss in a modern-day DAW with even 800db of gain or attenuation. No, seriously. So your gain staging issues come down to your preamp vs. your converter, and anything else in-between. Generally your AD converter is totally fine up to at least -6dbFS peak, and often past -1dbFS peak. Also, most high-quality converters have about 30db less noise than most preamps. So you can feel at ease getting your preamp to sound the way you want it without having to worry much about what the converter thinks. And don't worry at all about what the DAW thinks. You may have plugins that the programmer has, for some often inexplicable reason, decided to make hard-clip at 0dbR, or to saturate as it nears that point, which you might not like the collective effects of. But the programmer thinks he is trying to help you out (maybe...or maybe he's just a dope). Nonetheless you are free to gain stage however you want ITB with no concern whatsoever about loss from gain itself. |
| | |
| | #8 | |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Quote:
Pardon my French.. | |
| | |
| | #9 | |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Quote:
I wouldn't use gain reduction plugins anyway, I'm happy with my gain staging, I posted the question after a friend asked me how to record at -18dbfs if he wanted to run his pres hot | |
| | |
| | #10 | |
| Lives for gear | Quote:
Is it high SPLs you're having a problem with? I bought some Shure inline attenuators for drums and guitars for my pres that don't have pads. Maybe you should look into that, not an expensive solution. | |
| | |
| | #11 |
| Lives for gear Join Date: Oct 2003
Posts: 554
| To answer the OP question and restatement of the thread I think what you need is some sort of attenuation before A>D/post micpre output. You like the results and workflow of -18dBfs nominal levels. Great. You want those levels to print so mix time all your DAW faders start at 0 without any funny gain plug ins happening. Why make work and complication for yourself with trimming levels after they are in? Why not just print the levels you intend to mix from? Your pres have a sweet spot so you want their gain to hit that spot and not worry where their output hits the A>D. A>D's also have an optimal input range but are much more forgiving. Usually the manufacturer will specify that or some spec Yoda will have sussed it out and shared their findings. The best scenario is a pre/channel with an output level like most Neve 1084 rack modules' or many of the channel strip style units have into an A>D that has an input calibration trimmer such as a Digidesign 192I/O. That way you can calibrate the A>D for -18dBfs and trim the output of the pre after you've found it's ideal gain setting on the fly for any source you put into it. If on the other hand you have something like API pres in a lunchbox into an A>D with no analog trim your best off buying/making an external passive attenuator to trim the pres' level to what the A>D likes best. Either way you're not fooling around inside the DAW more than needed and it's much nicer, simpler, and faster to just reach over and twist a couple knobs on the pre to get going.
__________________ "Wow, that's really exciting and new and underground and authentic. Let us take it and bring it into our dark hearts." -John Stewart on marketing. |
| | |
| | #12 | ||
| Lives for gear | Quote:
The 'trim' or similar function in daws is simply a gain change that occurs before the file-writing stage. (in some daws, it is simply a gain change that occurs pre-fader). It is no different than just moving the fader... except you don't move the fader ![]() Quote:
__________________ The best microphone for an accordion is none. (This rule-of-thumb also applies to bagpipes and rappers.) | ||
| | |
| | #13 |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| I wouldn't introduce an attenuator into the chain unless you are ACTUALLY getting unpleasant non-linearity from overdriving the AD converter. Because attenuation post-conversion, at a DAW's internal resolution, is far more perfect than its analog counterpart. The "keep levels low" priesthood is totally screwing people up.... tutt |
| | |
| | #14 | |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Quote:
This has been the greatest quality improvement for me , now everything has the depth, accuracy, dynamic that I've long been longing for, I highly recommend reading this Proper Audio Recording Levels | Rants, Articles | My MASSIVE Blog | |
| | |
| | #15 |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Most likely, you were using your converters and your DAW fine before all this. You just didn't understand how to use your plugins. That priesthood doesn't understand this, by and large, and they are transferring their ignorance onto the unsuspecting. (Indeed, that article you linked to doesn't mention "plugin" once... tutt) |
| | |
| | #16 | |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Quote:
Yes I used it fine except that I had to turn ALL my tracks at -25 or something at mixdown, and that it sounded half as good as now. And regarding plug-ins, they sound better now and easier to setup.. | |
| | |
| | #17 | |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Quote:
You cannot clip the bus of a DAW, nor does the bus have a "sound" despite hue and cry from thousands of misled engineers. The bus in a floating point DAW has 900db of headroom, and in PTHD, many dozens of db of headroom. If you have plugins on the bus, or are rendering to a fixed format at the filesystem or a DAC, then you just pull the master fader down on the bus, and you will not clip or saturate anything you don't want to. No need at all to pull all the little faders down, trim all the channels, normalize, or track at lower levels. Don't take advice from anyone who doesn't know these very basic and instantly self-verifiable FACTS. Mastering engineers lecturing people how to mix is bad enough. Analog heads who don't think plugins can saturate like analog and don't know that they are doing so is just as bad. People who were overdriving a bus plugin, and solved the problem by tracking at lower levels are hilarious. You could get rid of a hangnail by cutting off your arm too. ![]() | |
| | |
| | #18 | |
| Lives for gear Join Date: May 2005 Location: Paris
Posts: 929
Thread Starter | Quote:
All I can say is that it sounds much better this way, mixes are effortless, I barely have to touch the faders! As far as the plug-ins, if the cause is overdriving them, then it's obviously a good solution to track less hot; more headroom for the plugins to breathe. Anyway, I try not to use plug ins, when I use them it's the UAD eqs and limiters, usually cutting rather than boosting.. | |
| | |
| | #19 |
| Gear addict | You're missing the point. This isn't about headroom within a digital system. It's about headroom in the analog stage of a converter.
__________________ Aspiring member of tomorrow. |
| | |
| | #20 | |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Quote:
At least, you don't have to go all way way down -18dbFS peak if you don't want to. -6dbFS peak will be fine with nearly any competent ADC. -1dbFS peak will almost always be fine with a good ADC. My 2192 gets some color up there at -1dbFS, but that's not necessarily a bad thing. I can fix the occasional unpleasant over with izotope RX quite well too. Sounds better than a limiter gumming up the chain. People can relax and make music. If they know where the problems are coming from, they are easy to avoid. | |
| | |
| | #21 |
| Lives for gear Join Date: May 2005
Posts: 3,647
| I don't understand the fuss. People have been driving preamps for decades. Just turn the gain up on the pre where your happy with the sound and pull the analog fader or trim pot on the preamp down until you hit the converter at the desired level. It doesn't matter if your hitting a convertor or tape machine - that is how you do it. And PS, don't forget that converters and tape machines have sweet spots as well - but that's another can of worms. |
| | |
| | #22 |
| Gear Head Join Date: Mar 2007 Location: the western edge
Posts: 43
| and if your pre has no trim control after gain, you add one with this: A-DESIGNS ATTY knowledge is good, gear is great |
| | |
| | #23 | |
| Gear maniac | Quote:
peace beau | |
| | |
| | #24 | |
| Lives for gear Join Date: Aug 2006
Posts: 534
| Quote: Soundawg | |
| | |
| | #25 |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Having no HD system here to test with, I'd appreciate it if you or someone else would determine precisely how much headroom an HD bus has. Sonalksis freeG is the free plug I use for this testing as it offers up to 54db of gain. Put the signal generator with a 600Hz, 0db (or -6db) sine wave on an aux input, follow it with freeG. Then go out on the bus you are testing, and back into another aux input or audio track. Put another freeG on that track and lower the gain an equal amount to what you had raised it. Let us know at what level the bus clips. You can see the clipping on an RTA, or you can hear it distorting to square wave. I have heard 48db of headroom, 96 db of headroom...there may be more headroom on the master bus because they do go 56bit fixed somewhere in the architecture. Use multiple FreeG inserts to get more than 54db of gain. If you don't believe in FreeG, because of RTAS/TDM differences or whatever, use whatever plugin offers clean gain in TDM, e.g. Trim or maybe EQIII. You may need more of those strung together to get appreciable gain of course. In PTLE, and in any other 32bit floating point DAW, you can string over 16 FreeG's together at max gain before you run out of headroom... |
| | |
| | #26 |
| Gear maniac | so... after just doing a little experiment, here is what i find.... within pro tools.......master busses are different than aux inputs. an aux input seems to clip very easily. for instance... a recorded drum stem... peaking at 0dbfs outputting to an aux input will light up the clip meter on the aux input.. as well as the master fader.... now.. you push the drum stem(audio track) fader up to +12 and the aux input is distorting... here is where aux inputs and master faders differ.. if you pull the aux input fader down... the volume comes down on the aux input.. but it is still clipping.... and no dynamics are restored, that were lost during the clipping... but if you feed that same drum stem(audio track) with the fader at +12 to a master fader, it will be clipping... but... if you pull the master fader back, the dynamics return, and clipping goes away.. make sense? peace beau |
| | |
| | #27 | |
| Lives for gear Join Date: Aug 2006 Location: No longer participating here.
Posts: 6,705
| Quote:
Those red lights are MEANINGLESS in PTLE. They are simply a gauge of level: there is no clipping going on internally when they light. That may or may not be true in HD. Your description of master faders is correct: a master fader is pre-insert, and is 100% mathematically identical to lowering the all faders feeding that bus proportionally. Note that you must pull down the bus' aux input channel (via a trim plugin or FreeG) an equal amount as you increased earlier, or else you are really just clipping your DAC, which demonstrates nothing. The test I described is very easy to do, I would be done faster than I can type this comment. Please do exactly as I specified to get an authentic result. Are you claiming there is really NO headroom on a PTHD bus? I think that contradicts what the manual and white paper both say. OK I am going to get to the bottom of this by uploading a PTLE session which a few people can load into PTHD, record enable a few tracks, record 5 seconds, zip up the files and reupload them and we will then know for absolute certain what the headroom of PTHD is in practice. ![]() | |
| | |
| | #28 | |
| Gear addict Join Date: Jun 2007 Location: Bristol, UK
Posts: 427
| Quote:
__________________ I am no slut. It's just hard to say no Never believe what an artist says, only what they do - David Hockney Ars longa, vita brevis | |
| | |
| | #29 |
| Gear maniac | ya, i cant really hear distortion until i am pushing the audio track fader up til around +6-10. up until then, it just sounds like limiting. but i am assuming that the amount of distortion that is audible, will vary with the content that was recorded to begin with. let me know how the experiment goes.. peace beau |
| | |
New Reply
Facebook
Twitter
LinkedIn
| Thread Tools | Search this Thread |
| Similar Threads | ||||
| Thread | Thread starter | Forum | Replies | Last Post |
| DAW Mixing: Lower Levels=Better Sounding Mixes? | uptoolate | So much gear, so little time! | 189 | 11th December 2008 04:26 PM |
| Recording at lower levels in a DAW? | TechnoCowboy | Mastering forum | 34 | 14th May 2008 10:28 PM |
| Ref. Levels - camera to DAW | ergalthema | Post Production forum! | 1 | 12th February 2008 08:18 PM |
| Debunking myths - DAW Mix Bus Levels | Ethan Winer | Music computers | 57 | 6th November 2006 07:33 AM |
| Calibration levels for DAW | audioboy6 | So much gear, so little time! | 22 | 2nd July 2005 12:12 PM |
| |