|
|
Post by MorEQsThanAnswers on Mar 15, 2021 18:09:01 GMT -6
Started getting into the idea of gain staging my multi-tracks to 0vu before starting.
Looking at a guitar loop and checking a few different style meters to figure out where this thing needs to be and some clarification could go a long way if anyone can help.
Using the Klanghelm VUMT meter calibrated to -18, I'm peaking at 0 VU. My SilverBullet is displaying peaks around +6VU. This is a little funky because both of these work consistently and accurately when using test tones, but the real kicker for me is the fact that my DAW is peaking at -8 dBfs to get these meters to read 0VU.
I understand that there is a difference between VU (loudness) and instantaneous digital peaking, but if my interface, SilverBullet, and VUMT plugin are calibrated to -18dBfs/+4dbBu, I'm surprise that I'd have to send in signal that's digitally peaking at -8 dBfs to get my VU to read 0. It seems as if I'm blasting signal into these meters that would be 10dB above the point at which they correlate with 0 VU to get them to read as such??
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 15, 2021 20:41:56 GMT -6
1) 0 VU is +4 dBu with a 300ms integration time. 0 dBfs depends on your converter. You have to calibrate your digital VU meters.
2) Klanghelm VUMT is not a true peak meter. It is inaccurate and based on the samples, not the recreated signal the samples represent, a common flaw of digital metering and processing. VUMT will usually be a few VU off at almost all times during playback.
3) PSP Triple Meter has a true peak option that is much more accurate than VUMT at 44.1 and 48 kHz. Waves VU also has great needle movement but I don’t know if it tries to estimate the true peaks. A brief comparison with psp in digital mode vs true peak will show it.
4) Physical VU meters vary a great deal too.
|
|
|
|
Post by jmoose on Mar 15, 2021 20:50:12 GMT -6
|
|
|
|
Post by mrholmes on Mar 16, 2021 15:09:53 GMT -6
I calibrated everything a year or two ago I can't remember I just listened before this, and user svart seconded this. The newest plug ins crap out like hardware if I feed them over the top. It was much surprised that Ray also simulated this behavior - the sound gets small and ugly. |
|
|
|
Post by EmRR on Mar 16, 2021 15:22:29 GMT -6
|
|
|
|
Post by srb on Mar 16, 2021 16:23:05 GMT -6
Used the Free G for years. Did not know about the Orban! 👍 |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 16, 2021 18:55:31 GMT -6
I calibrated everything a year or two ago I can't remember I just listened before this, and user svart seconded this. The newest plug ins crap out like hardware if I feed them over the top. It was much surprised that Ray also simulated this behavior - the sound gets small and ugly. Yep. They get nasty quickly driven hot and not aliased or clipped nasty like a Waves plug. Some of the Vpres really need trimming. The 2c breaks up in a cool way; the 376 and 31a sound like shit hit too hot. |
|
|
|
Post by mrholmes on Mar 16, 2021 20:37:30 GMT -6
I calibrated everything a year or two ago I can't remember I just listened before this, and user svart seconded this. The newest plug ins crap out like hardware if I feed them over the top. It was much surprised that Ray also simulated this behavior - the sound gets small and ugly. Yep. They get nasty quickly driven hot and not aliased or clipped nasty like a Waves plug. Some of the Vpres really need trimming. The 2c breaks up in a cool way; the 376 and 31a sound like shit hit too hot. It’s enough that they sound nice in their sweet spot. I am extremely thankful for the new plug in generation. Can’t help myself to acknowledge that even the tape sims are extremely close that’s at least what I learn mixing my current song. Tracked Neveish - Tape Sim - PA SSL - optional compression is the most often used combination. As soon I bypass BRA Magnetide a natural bigness, fatness is gone. Gain- staging finally is fun in the digital domain.... 😉 |
|
|
|
Post by jmoose on Mar 17, 2021 14:29:48 GMT -6
Started getting into the idea of gain staging my multi-tracks to 0vu before starting. Looking at a guitar loop and checking a few different style meters to figure out where this thing needs to be and some clarification could go a long way if anyone can help. Using the Klanghelm VUMT meter calibrated to -18, I'm peaking at 0 VU. My SilverBullet is displaying peaks around +6VU. This is a little funky because both of these work consistently and accurately when using test tones, but the real kicker for me is the fact that my DAW is peaking at -8 dBfs to get these meters to read 0VU. I understand that there is a difference between VU (loudness) and instantaneous digital peaking, but if my interface, SilverBullet, and VUMT plugin are calibrated to -18dBfs/+4dbBu, I'm surprise that I'd have to send in signal that's digitally peaking at -8 dBfs to get my VU to read 0. It seems as if I'm blasting signal into these meters that would be 10dB above the point at which they correlate with 0 VU to get them to read as such?? Ok with more time a more serious answer... and hopefully skip the heavy math. Basic idea with gain staging is that it should happen while recording. If you've got good sounds coming in and some idea of what the goal was on the outset, in theory there should be a fairly decent mix with most channel faders parked at or near unity gain. More often then not when I get "home toolz" projects in for mix, people tend to record things at max gain or wherever they happen to fall and then go to DAW faders for balancing... which is how you end up with some faders at +5 and others at -25 to try and balance a mix. So in cases like those I'll plop trims on the tracks that need them so I can all of the channels & faders into a smaller, workable range. There are four distinct kinds of metering all related but different. Peak - RMS - analog - digital Most digital is peak. Most analog is RMS. Not a hard & fast rule there are exceptions! Meter ballistics of any sort don't tell the whole story, you need to calculate the averages. Fer instance drums & stuff with high transients will show a much higher peak then RMS. Might peak at -6dBFS but RMS is -18. Most "steady state" things like distort-o guitars don't have a huge peak and are crushingly way too loud if tracked at say, -8dBFS. If you get into calibrating converters and checking meters, never a bad thing, results are better when using test tones. Often a 1kHz sine wave though I still follow the old analog tape routine of checking 1kHz, 10kHz and 100Hz. So I would start there. Fire up the tone generator to 1kHz, set at -18dBfs in the DAW and see where other things start to land. |
|
