|
Post by M57 on Jul 13, 2017 7:07:33 GMT -6
The "leave 6dB for the mastering guy" thing is -- well, I'm not even sure where it came from. Avoiding the need to add an additional analog attenuator to knock levels down where they interface well with comp thresholds, EQ overload points, etc. I prefer to leave headroom, but I'll play devil's advocate. Why Analog? Why can't the mastering guy use clip gain? You're just giving him a file with the max amount of 1s and 0s. Let him decide how much of it to knock out.
|
|
|
Post by EmRR on Jul 13, 2017 7:31:39 GMT -6
We're speaking to the origin of the request, and a purist mindset of changing the least possible before hitting analog domain. I haven't personally run through all the path options in that era to be able to fully explain or defend. From memory it was based on assessment of the various sonic penalties that might be paid for various approaches.
|
|
|
Post by jazznoise on Jul 13, 2017 8:17:03 GMT -6
I use a limiter for reference on things like FX balances and ambiance levels. Usually end up drying out the mix a touch and bringing the vocal up a touch - no harm in it. Using one all the time has the danger of you getting too fond of driving instruments into the limiter, or at least that's been my experience. Turn off the limiter for the first time all day and the snare and vocal are blasting!
|
|
|
Post by dandeurloo on Jul 13, 2017 9:19:07 GMT -6
For some reason I NEVER use MAster faders in PT. I only use AUX's. I have a feeling I am missing out. I also don't use Group faders. I use AUX's. I keep thinking I need to learn or reshape my work flow to use those? But mostly I am just trying to get out of the DAW into the console! You talking about VCA faders? Yeah, I so rarely use them I don't even know what they are properly called. haha
|
|
|
Post by jimwilliams on Jul 13, 2017 9:32:57 GMT -6
My two mix buss levels are at about +14 dbu. The ADC clips at +20 dbu. That leaves around 6 db headroom out of the current feedback designed sum amps, fader amps and output drivers. The signal level is an additional 10+ db above the console residual noise level that is already extremely low. I never use a compressor/limiter on the two mix, never have. That's for mastering.
|
|
|
Post by jin167 on Jul 13, 2017 11:04:50 GMT -6
I'd rather choose dither over noise floor of ANY analog line amplifier in existence.
|
|
|
Post by EmRR on Jul 13, 2017 11:33:48 GMT -6
I'd rather choose dither over noise floor of ANY analog line amplifier in existence. You might not have 15 years ago.
|
|
|
Post by jin167 on Jul 13, 2017 11:52:46 GMT -6
I'd rather choose dither over noise floor of ANY analog line amplifier in existence. You might not have 15 years ago. I can agree with you and that's what I mentioned in my previous comment. Also proves that 6dB headroom thingy is no longer valid.
|
|
|
Post by noah shain on Jul 14, 2017 1:00:11 GMT -6
I'd rather choose dither over noise floor of ANY analog line amplifier in existence. Point missed
|
|
|
Post by jin167 on Jul 14, 2017 1:32:36 GMT -6
I'd rather choose dither over noise floor of ANY analog line amplifier in existence. Point missed Don't think so. I'm not saying it's the only reason but it is one of many reasons. Mr. Williams also made a good point about the maximum output of a converter and leaving 6dB headroom but even then it's something a mastering engineer should be capable of dealing with in digital domain. Mixing with 6dB headroom could be a good practice to follow but not necessary as long as you know what you're doing (if you don't know what you are doing then yeah sure, leave the 6dB headroom).
|
|
|
Post by noah shain on Jul 14, 2017 8:43:23 GMT -6
Don't think so. I'm not saying it's the only reason but it is one of many reasons. Mr. Williams also made a good point about the maximum output of a converter and leaving 6dB headroom but even then it's something a mastering engineer should be capable of dealing with in digital domain. Mixing with 6dB headroom could be a good practice to follow but not necessary as long as you know what you're doing (if you don't know what you are doing then yeah sure, leave the 6dB headroom). I meant that I missed your point! Noise certainly can't be the deciding factor when building a 2 bus chain in this day and age...can it?
|
|
|
Post by adamjbrass on Jul 14, 2017 8:51:56 GMT -6
I gotta disagree with the "choose dither" over analog noise comment. That's like choosing between a hamburger and a veggie burger.
|
|
|
Post by EmRR on Jul 14, 2017 9:09:47 GMT -6
Not necessarily noise. High res environments can reveal the differences between a longer and a shorter signal processing chain, even the addition of a passive device. Additional cables, additional metal contacts.
Noise: for all we gain in increased dynamic range, the more we seem to over-process and further reduce dynamic range. That increased dynamic range reveals that decreased S/N from processing. A long processing chain with a lot of compression can easily reveal a lot of previously unnoticed noise floor. Doesn't matter with steady state music, but really obvious with dynamic music having breaks, stops, quiet sections, etc.
I can't think of a good reason to ignore the 6 dB 'rule' from a mixing perspective. There's nothing I see to be gained from mixing hotter, not in any genre I work in, nor really from any I don't either. A 32 bit float bounce has no noise floor penalty, and has room to move in either direction. A 24 bit analog mix has room to move too, and a lot of analog equipment starts to sound fairly brassy and harsh when pushed hot enough to print close to 0dBFS with a lot of modern converters being +24dBU = 0dBFS. Common +/-18V power rails = +24.31 dBu max level unless there's output transformer step-up, so multiple parts of the chain are up against the wall. +/-24V rails = +26.81 dBu, only buys 2.5dB before flatlining.
|
|
|
Post by notneeson on Jul 14, 2017 9:31:26 GMT -6
I think the OP was talking about ITB mixing.
But, if I understand what I've read on the subject correctly, when you're printing anything hot (including a mix) to your DAW it's theoretically possible to have intersample peaks that clip the audio, but do not meter as clipping. (I'm getting this from stuff Paul Frindle has posted over the years.)
|
|
|
Post by jimwilliams on Jul 14, 2017 9:43:52 GMT -6
My console mix at +14 dbu is sent to the Burrbrown PCM1794 ADC set to 24 bits, 44.1k for CD prep, 48k for DVD. That then feeds the digital input sound card of the PC, no dither nor buss compression is ever used until disc burning.
THD of the console outputs is .0006% THD+noise, .00015% CCIF IMD, the residuals of my Audio Precision analyzer. That's 1.5 ppm, a very small error rate. The 30 mhz bandwidth and 2000V/us slew rate assure the fidelity is not affected by bandwith limitations.
There are 3 classes of sum amp designs: one is voltage feedback of an analog signal, used since the 1920's with tubes, transistors or opamps.
The second is digital summing with a lot of math involved. It is a lossy algo as resolution decreases with decreasing level, the opposite of analog suming. The hottest signal is the cleanest, the lowest level signals are dirty in comparison.
The third class is analog current feedback suming, very rare and not seen in commercial console designs. That topolgy uses a small amount of current rather than voltage for feedback correction. It is immune to bandwidth and slew limitations. Yes, it has a completely different sound than the previous topologies. It's akin to your ears popping after a long jet flight.
|
|
|
Post by massivemastering on Jul 14, 2017 9:48:44 GMT -6
I get that -- I'm just pointing to the whole digital element here. You don't need analog attenuation.
And don't get me wrong here -- I'm an absolute whore for headroom at every possible stage -- but once those analog signals are in the digital world, the "analog guidelines" of headroom give way to the "absolute rules" of digital headroom. If I'm introducing another analog stage (which I almost always do), establishing the output level into that stage can be done in the digital realm without any harm.
And don't get me wrong again -- I'm not arguing "against" peaking at -6dBFS... I just don't find it to be really any different than peaking at -3 or -1 or -12dBFS. *Some* headroom is just good form (if absolutely nothing else). How much headroom on the other hand...
|
|
|
Post by M57 on Jul 14, 2017 9:52:05 GMT -6
I think the OP was talking about ITB mixing. Yep. That's me. Once ITB - it stays there. One thing that got my attention is that when I put Slate's FG-X (I assume it's a limiter) on the master buss and get it to start peaking (at 0.1), if from there I send it to Logic's Adaptive Limiter with the input at 0 and "True Peak Detection" enabled, the levels go above 0db and additional reduction is applied by Logic. I dunno, it could be a bug in the FG-X (or Logic's limiter), but I'm not willing to risk the mix. I love the little transient and dynamic perception touches the FG-X can add, but I make sure Logic's limiter is the last thing the music sees.
|
|
|
Post by ChaseUTB on Jul 14, 2017 12:34:08 GMT -6
I think the OP was talking about ITB mixing. Yep. That's me. Once ITB - it stays there. One thing that got my attention is that when I put Slate's FG-X (I assume it's a limiter) on the master buss and get it to start peaking (at 0.1), if from there I send it to Logic's Adaptive Limiter with the input at 0 and "True Peak Detection" enabled, the levels go above 0db and additional reduction is applied by Logic. I dunno, it could be a bug in the FG-X (or Logic's limiter), but I'm not willing to risk the mix. I love the little transient and dynamic perception touches the FG-X can add, but I make sure Logic's limiter is the last thing the music sees. You need to set the ceiling or outputs for at least -.3 to -.5 on your final Limiter. You may need to reduce the output even more depending if MP3 streaming a la SoundCloud is the Final destination for the wav file. I had an issue with a master last night where I had to pull down almost a whole DB ( -1.0 ) output to get the 320kbp/s MP3 not to clip after conversion... i also find that this happens more with 32 bit float wav to MP3 files than 24 bit wav To MP3 conversion files
|
|
|
Post by jeremygillespie on Jul 14, 2017 14:18:47 GMT -6
In the loudest parts of a mix, Im usually pinned pretty hard on the VU meters on the console. I know when things get crunchy, and I stay away from that - but you have to push the desk pretty hard to get there. At that point, (generally with the bus comp getting tickled, the a,b,c,d compressors handle most of that for me), it's out to a bus EQ, then onto the atr102 where the level gets swallowed up pretty good. Out of tape, into a pair of VP-28's to boost the signal, and finally into a Lavry Gold ADC. You can hit the Gold pretty good and it never seems to fold at all.
I've never looked at what the level was as far as meaaurements of my final mix. If it sounds good it is good. Never had a complaint from a Mastering Engineer.
|
|
|
Post by jin167 on Jul 14, 2017 14:35:46 GMT -6
I get that -- I'm just pointing to the whole digital element here. You don't need analog attenuation. And don't get me wrong here -- I'm an absolute whore for headroom at every possible stage -- but once those analog signals are in the digital world, the "analog guidelines" of headroom give way to the "absolute rules" of digital headroom. If I'm introducing another analog stage (which I almost always do), establishing the output level into that stage can be done in the digital realm without any harm. And don't get me wrong again -- I'm not arguing "against" peaking at -6dBFS... I just don't find it to be really any different than peaking at -3 or -1 or -12dBFS. *Some* headroom is just good form (if absolutely nothing else). How much headroom on the other hand... My point also. My opinion is that there's more to lose in using an analogue line amp stage to attenuate or boost level when compared to changing levels in digital domain. 6dB headroom is like I said a 'good practice' to keep around but not a rule as long as one understands how digital audio works. Note that none of my comment has anything to do with the dynamic or loudness. It's a different subject.
|
|
|
Post by M57 on Jul 14, 2017 16:22:59 GMT -6
You need to set the ceiling or outputs for at least -.3 to -.5 on your final Limiter. You may need to reduce the output even more depending if MP3 streaming a la SoundCloud is the Final destination for the wav file. I had an issue with a master last night where I had to pull down almost a whole DB ( -1.0 ) output to get the 320kbp/s MP3 not to clip after conversion... i also find that this happens more with 32 bit float wav to MP3 files than 24 bit wav To MP3 conversion files I tried -0.5 and it worked - Of course, then I found that dialing up the "Dynamic Perception" knob could again push things over peak. I love the sound and functionality of many of the Slate plugs, but it can be frustrating when you don't really know how it's doing what it does. Steven has definitely figured out that marketing supersedes descriptive accuracy. Just turn up the 'happy.' Like I said, I'm keeping Logic's Limiter at the end of the 2, whether or not it ever does anything. Mind you, this is only for psuedo mastering purposes for SC and the like. If I ever decide to get something professionally mastered, I'm taking massivemastering and jin167's advice to heart, Leave some headroom. -6.0db doesn't sound like bad target to aim for when mixing.
|
|
|
Post by Guitar on Jul 24, 2017 19:01:35 GMT -6
In 24 bit digital, you get foot room. In analog audio you get a lot of headroom/voltage, but less dynamic range? Different approaches for sure. A little tip I picked up, from making mistakes: I finally upgraded to the latest version of Cubase a few months back. If my 2-bus level is WAY TOO HOT, and I don't want to upset my balance, I can just apply a VCA fader 'group' to all of my tracks, pull them down 'symmetrically' with a single fader, and then delete the fader... ahh. Don't know if this applies to other DAWs or not. That way I can apply my normal compression, tape sims, etc, at the right levels ITB. One of my best weapons is to have a nice 2-bus chain. I do love slamming it like massivemastering said. But I also fantasize about deleting the last half of it, and sending it to a proper mastering engineer. I can only imagine what that would sound like..
|
|
|
Post by drbill on Jul 24, 2017 19:31:28 GMT -6
In 24 bit digital, you get foot room. In analog audio you get a lot of headroom/voltage, but less dynamic range? Different approaches for sure. A little tip I picked up, from making mistakes: I finally upgraded to the latest version of Cubase a few months back. If my 2-bus level is WAY TOO HOT, and I don't want to upset my balance, I can just apply a VCA fader 'group' to all of my tracks, pull them down 'symmetrically' with a single fader, and then delete the fader... ahh. Don't know if this applies to other DAWs or not. That way I can apply my normal compression, tape sims, etc, at the right levels ITB. One of my best weapons is to have a nice 2-bus chain. I do love slamming it like massivemastering said. But I also fantasize about deleting the last half of it, and sending it to a proper mastering engineer. I can only imagine what that would sound like.. PT has had VCA faders for what?? Probably a decade or so. Yes, your workflow is cool, and works great, eh? If you've already got automation written, leave the VCA in the session without deleting it.
|
|
|
Post by drbill on Jul 24, 2017 19:36:26 GMT -6
As for the OP question - I try to mix at least -6-8 dB below full scale on the mix bus. IMO, it's not so much how the software handles elevated levels above that as it is how the analog electronics sound in your DA and downstream outboard when you're pushing that close out and back into the DAW. Between the two, if you're pushing the crap out of things, mixes become crunchier, more congested, less open, blurred, edgy. You know....like modern pop hits. No reason to push that hard. That's for mastering, not mixing.
|
|
|
Post by BradM on Jul 24, 2017 19:48:29 GMT -6
In 24 bit digital, you get foot room. In analog audio you get a lot of headroom/voltage, but less dynamic range? Different approaches for sure. A little tip I picked up, from making mistakes: I finally upgraded to the latest version of Cubase a few months back. If my 2-bus level is WAY TOO HOT, and I don't want to upset my balance, I can just apply a VCA fader 'group' to all of my tracks, pull them down 'symmetrically' with a single fader, and then delete the fader... ahh. Don't know if this applies to other DAWs or not. That way I can apply my normal compression, tape sims, etc, at the right levels ITB. One of my best weapons is to have a nice 2-bus chain. I do love slamming it like massivemastering said. But I also fantasize about deleting the last half of it, and sending it to a proper mastering engineer. I can only imagine what that would sound like.. Why not just pull the master fader down? It's floating point architecture. Brad
|
|