Mo Facta
Farts of Nature
A HAH! EUREKA!
Better sound BELOW -12dBFS! Amazing!
Mathematical theory please?
Cheers
A HAH! EUREKA!
Better sound BELOW -12dBFS! Amazing!
Mathematical theory please?
Cheers
S
sjjohnston
New member
I've been on the Internet long enough (a) to recognize someone who's only interest is engaging in a pissing context and (b) to have found by experience that there are no cash prizes available for winning one. I don't think I have a responsibility to respond to a combination of some obvious and accurate, though irrelevant, facts, plus some odd gobbledygook.
In the interest of conveying my only point to people who might actually read this thread to learn something, a shortened and to-the-point summary of what I said earlier:
- There's no significant harm, and some benefit (in ease of use, if nothing else) in leaving some extra room below 0 dBFS when tracking
- But the number of tracks you intend to include in your final project is not relevant to how much headroom you should allow.
In the interest of conveying my only point to people who might actually read this thread to learn something, a shortened and to-the-point summary of what I said earlier:
- There's no significant harm, and some benefit (in ease of use, if nothing else) in leaving some extra room below 0 dBFS when tracking
- But the number of tracks you intend to include in your final project is not relevant to how much headroom you should allow.
S
sjjohnston
New member
Oh yeah, and though this wasn't even on the table earlier:
bouldersoundguy's advice,* as usual, makes perfect sense: it doesn't make any sense to master your backing tracks, then mix the vocal over them, and master them again.
Everyone knows (I hope) that, when mixing, you want to watch all the levels so that you don't clip the output. That doesn't really have a lot to do with where you want the levels when tracking. The whole reason you've got channel and master faders is so that you can change the levels when mixing.
* and Massive Master's as well, as he also said that.
bouldersoundguy's advice,* as usual, makes perfect sense: it doesn't make any sense to master your backing tracks, then mix the vocal over them, and master them again.
Everyone knows (I hope) that, when mixing, you want to watch all the levels so that you don't clip the output. That doesn't really have a lot to do with where you want the levels when tracking. The whole reason you've got channel and master faders is so that you can change the levels when mixing.
* and Massive Master's as well, as he also said that.
Last edited:
Massive Master
www.massivemastering.com
Running an unfinished mix through a plugin preset which is arguably responsible for more damage to otherwise perfectly decent audio than most other plugs combined is most definitely not "mastering" -- You're dynamically compromising (very, very likely in a very, very non-flattering manner) and skewing the stereo image of an unfinished mix and then trying to stack dynamic, non-skewed content on top of it.
Mix when you're done tracking. And be extremely careful when using Blowzone presets... I'm not saying that Ozone on its own is evil (as it's only as good or horrible as the person using it), but the presets I've heard on that plug are some of the most damaging, horrific sounding, "I can't believe someone would actually do that to a mix" presets I've ever heard (period). Some of them are very unique (as I can't even think of another plug that can completely trash the phase coherency in the low end faster or more efficiently). And again - I'm not trying to trash it outright. I know a few engineers that use a particular module or two (the limiter, mostly - Occasionally dither as well). But the presets are just whacked and most everything else relies on a maul-the-band (multi-band) approach (which is rarely used in typical mastering - or mixing for that matter - at least not on the 2-buss).
FTR: "Mastering" is taking a collection of completed mixes and creating a production master (hence the name) for replication and distribution.
Mo Facta
Farts of Nature
No one is here for a pissing contest. We are here for debate about audio.
I did not agree with some things you said, and I challenged you on it.
If you want to rack that up as a pissing contest, be my guest. And as far as "gobbledygook" goes, the only one spouting any was you. What exactly is "dumb" summing?
You are still yet to answer any of my questions in a straight forward manner as well.
Anyway, I'm off the internet today in light of making records. See ya later.
Cheers
Bobbsy
Boring Old Git
A few thoughts, probably more to muddy the water than anything else.
First off, the OP specifically mentioned a VU meter. Perhaps he really meant this or perhaps he just used it as a generic term for "meter". However, this can be important.
A true VU meter is an old fashioned analogue device that is fine for measuring average/constant levels but pretty much useless as a way to judge peaks. The ballistics of the bouncing needle simply aren't up to it--they're a bit slow to react to anything and start to fall long before you've seen the level of a peak.
However, very few DAWs have VU meters. Most have some form of PPM (Peak Programme Meter) which, being electronic, react almost instantly to peaks, then hold this peak briefly before starting to fall again. This makes them much more accurate at measuring actual peaks--not perfect, but much better on most sources.
What does this mean in practice? Well, with a VU meter you are pretty much forced to record very low on the meter just to leave a lot of headroom for the peaks you're not seeing. With a PPM, you can trust a bit more that the peaks you're seen are accurate and leave a bit less "Tennessee windage" for the peaks.
Now, this still doesn't mean you want to record everything super hot. There's no good reason to do so and lots of reasons not to. The advice of having the average level on your tracks at around -18dBFS and peaks about 6dB higher (-12dBFS) is a very good starting point and one I use myself. This is equivalent to most things being around 0dBu with peaks at +6dBu which has worked well in the analogue world for decades. It gives a good starting level and still leaves 18dB margin for unplanned peaks which caters for virtually everything.
What I take a bit of issue with is the advice of NO peak EVER going above -12dbFS. If you get one or two peaks going above the -12 level in a track where everything else is sitting nicely, there's no reason to drop everything just for a couple of rimshots or whatever--so long as nothing ever goes above the magic 0dBFS clipping number. Frankly, one or two peaks like that will almost certainly be tamed with some light compression when you're mixing anyway--no sense having to boost a whole five minute track for two or three random peaks!
The other curve ball to throw in is that, with modern A to Ds and DAWS using 24 bit or 32 bit Floating Point systems, it's probably far more important to watch for clipping at the mic, pre amp or input to the A to D. They dynamic range at 24 or 32 bits is silly-huge anyhow...certainly more forgiving than your analogue stages.
Finally, I disagree with the advice to set your record levels based on the number of tracks you will be using. Yeah, adding together lots of tracks at -12 can result in clipping on the final mix--but that's why you have faders and master faders. Just as there's no good reason to record everything super hot, there's also no good reason to keep every fader at 0 during the mix. Modern mics and interfaces are lovely and quiet, often with a signal to noise down near -90dB. This gives you S/N to waste--but it's foolish to throw too much away for the sake of keeping faders at zero!
Bob
First off, the OP specifically mentioned a VU meter. Perhaps he really meant this or perhaps he just used it as a generic term for "meter". However, this can be important.
A true VU meter is an old fashioned analogue device that is fine for measuring average/constant levels but pretty much useless as a way to judge peaks. The ballistics of the bouncing needle simply aren't up to it--they're a bit slow to react to anything and start to fall long before you've seen the level of a peak.
However, very few DAWs have VU meters. Most have some form of PPM (Peak Programme Meter) which, being electronic, react almost instantly to peaks, then hold this peak briefly before starting to fall again. This makes them much more accurate at measuring actual peaks--not perfect, but much better on most sources.
What does this mean in practice? Well, with a VU meter you are pretty much forced to record very low on the meter just to leave a lot of headroom for the peaks you're not seeing. With a PPM, you can trust a bit more that the peaks you're seen are accurate and leave a bit less "Tennessee windage" for the peaks.
Now, this still doesn't mean you want to record everything super hot. There's no good reason to do so and lots of reasons not to. The advice of having the average level on your tracks at around -18dBFS and peaks about 6dB higher (-12dBFS) is a very good starting point and one I use myself. This is equivalent to most things being around 0dBu with peaks at +6dBu which has worked well in the analogue world for decades. It gives a good starting level and still leaves 18dB margin for unplanned peaks which caters for virtually everything.
What I take a bit of issue with is the advice of NO peak EVER going above -12dbFS. If you get one or two peaks going above the -12 level in a track where everything else is sitting nicely, there's no reason to drop everything just for a couple of rimshots or whatever--so long as nothing ever goes above the magic 0dBFS clipping number. Frankly, one or two peaks like that will almost certainly be tamed with some light compression when you're mixing anyway--no sense having to boost a whole five minute track for two or three random peaks!
The other curve ball to throw in is that, with modern A to Ds and DAWS using 24 bit or 32 bit Floating Point systems, it's probably far more important to watch for clipping at the mic, pre amp or input to the A to D. They dynamic range at 24 or 32 bits is silly-huge anyhow...certainly more forgiving than your analogue stages.
Finally, I disagree with the advice to set your record levels based on the number of tracks you will be using. Yeah, adding together lots of tracks at -12 can result in clipping on the final mix--but that's why you have faders and master faders. Just as there's no good reason to record everything super hot, there's also no good reason to keep every fader at 0 during the mix. Modern mics and interfaces are lovely and quiet, often with a signal to noise down near -90dB. This gives you S/N to waste--but it's foolish to throw too much away for the sake of keeping faders at zero!
Bob
bouldersoundguy
Well-known member
Just to be clear...
Although I often say to keep peaks below -12dBFS, that is to allow for some peaks that unexpectedly go past that level. If you use analog emulation plugins then it's a good idea to keep your average levels around -18dB (or whatever the plugin considers the equivalent of 0dBVU) so the emulation works as intended.
It's like driving on a two-lane highway. On the one side you have the guardrail and on the other you have the oncoming traffic. Running up against the guardrail gets a bit noisy, and going across the center line is catastrophic. The approximate middle is the safest place to be.
Bobbsy
Boring Old Git
Thanks bouldersoundguy...you and I are definitely on the same page on this. It just worries me that sometimes newbies read this stuff and take it completely literally.
Re-reading this thread, one other point I'd question is the long discussion about "resolution" based on the bit depth. Perhaps this is just a question of different terminology used in different countries, but I'd always describe the bit depth as determining the dynamic range of the recording rather than "resolution". Frankly, "resolution" simply isn't a word I use to discuss digital recording--but, if it did, I'd be more inclined to use it in a discussion of the sample rate. However, let's not get into Nyquist frequencies here--the thread has already kicked off enough!
Re-reading this thread, one other point I'd question is the long discussion about "resolution" based on the bit depth. Perhaps this is just a question of different terminology used in different countries, but I'd always describe the bit depth as determining the dynamic range of the recording rather than "resolution". Frankly, "resolution" simply isn't a word I use to discuss digital recording--but, if it did, I'd be more inclined to use it in a discussion of the sample rate. However, let's not get into Nyquist frequencies here--the thread has already kicked off enough!
N
Nathan1984
New member
No you misunderstood me. I don't use ozone on the vocals after I use ozone, I have my master mix consisting of guitars, and drums, and then I just record vocals over top. But I am taking you guys advice, I will get my tracks before I "butcher them with ozone" lol. To the guy who hates ozone, I only use it because I am not even close to experienced enough to do it manually. I haven't gotten terrible results with it, I use it more or less for a slight polish, not really as an effect or anything.