Recording levels? -18dbs

[The Weed]

Darryl, you know far more about this than I do, so if you could help me understand this better, I would appreciate it. With my 002 Rack, the manual said that 0 on my meters in Pro Tools was -14 dBFS. Based on that, I thought that if the meter read a peak of -6, that would mean the peak was at -20 dBFS and set my pre's gain accordingly. Is that incorrect? As far as I know, with my Pro Tools vanilla, I am unable to set any meter offset in Preferences.

Thanks,

[Darryl Ramm]

Originally Posted by The Weed

Darryl, you know far more about this than I do, so if you could help me understand this better, I would appreciate it. With my 002 Rack, the manual said that 0 on my meters in Pro Tools was -14 dBFS. Based on that, I thought that if the meter read a peak of -6, that would mean the peak was at -20 dBFS and set my pre's gain accordingly. Is that incorrect? As far as I know, with my Pro Tools vanilla, I am unable to set any meter offset in Preferences.

Thanks,

You mean this

"All Digi 002 audio inputs and outputs are set for
14 dB of headroom below 0 dB, or full code.
This means at the nominal reference input level
(+4 dBu or –10 dBV) you can have up to 14 dB of
headroom before input or output clipping occurs."

That is as statement about the *analog* design of the 002. dBu and dbV are all analog side signal level measurement. This does *not* say "0 on my meters in Pro Tools was -14 dBFS". dbFS anything is not mentioned here or relevant.

Pro Tools classic meters are digital, they display dbFS, or as the reference guide used to (may still?) say. They indicate the (digital) levels going to/coming from disk.

If you drive an input up to 0dbFS it damn well will start to digitally clip (go drive say a mic input on your 002 from a signal generator on an output and prove it to yourself). If you want say 16dB of headroom at the converter for transients then you just track at -16dbFS. You don't need to worry about the analog domain, at least not to the first order, you are worried about overhead at the converter, and the converter digital output tells you all you need to know. In more complex setups what you do as far as analog side behavior etc. gain or attenuator settings or analog console settings for analog signal reasons should be judged by you actually listening to the signals and what sounds best.

[Drew Mazurek]

Originally Posted by Darryl Ramm

Oh no headroom arguments again. Worrying about mixing headroom on modern DAWs is just not intersting, there is just so much headroom, and while the end-argument here of not wanting to track too hot may be good (why do so many people want to believe it's a good idea? Yikes), its not the reason.

Tracking levels are set for font-end/preamp/conversion reasons. Everything else is just math, and with 32/64-bit float processing in plugins, 64-bit float mixer there is more headroom available than anybody needs.

As for "fader sweet spot" hopefully you are meaning a tactile/UI sweet spot not an audible one.

I understand your underlying argument perfectly, but the fact is, recording hot impacts the headroom available for plugins without having to micromanage input/output controls everywhere.

[Park Seward]

Originally Posted by Chief Technician

Anyone who has worked in post is used to -20 as well.

I remember sending music files to Pro Tools from a Digital Betacam tape. The DB tape was recorded to the broadcast standard of -18 dBFS. The PT mixer raised everything to up to get close to "0" and then send the mixed file back to me and the DigiBeta. It was now too high, almost hitting clipping.

A sound recorded at -20 will sound the same at -50. It will just have a little more noise that you will not hear, especially if you dither.

[Bob Olhsson]

Originally Posted by Darryl Ramm

Tracking levels are set for font-end/preamp/conversion reasons. Everything else is just math, and with 32/64-bit float processing in plugins, 64-bit float mixer there is more headroom available than anybody needs.

This isn't true of anything but simple gain changes and the very simplest plug-ins. All analog modeled plug-ins have a sweet spot exactly the same as all analog gear has. Otherwise it isn't an accurate model! No plug-in developer assumes their plug-in needs to be accurate when sent signals above 0dBFS. It is operating outside of the design and may or may not be a problem.

[JFreak]

Originally Posted by Park Seward

A sound recorded at -20 will sound the same at -50. It will just have a little more noise that you will not hear, especially if you dither.

Again, this all comes down to numbers. If your AD converter has a dynamic range of 124dB (a good one!) then you have 20dB headroom and you gain NOTHING if you record any hotter than that (-20dBFS). So better be on the safe side.



And then again... there are not so many situations where you really need +120dB dynamic range. Your dialog or lead vocal or acoustic guitar or whatever has way lower dynamic range to capture. You need to think of gunshots or explosions or jet flying over your head under 300 feet if you would really be concerned about the NOISE FLOOR which is analog concept and usually totally nonsense in digital domain.



So to repeat myself, find that monitoring volume knob and turn it up!

[Darryl Ramm]

Originally Posted by Bob Olhsson

This isn't true of anything but simple gain changes and the very simplest plug-ins. All analog modeled plug-ins have a sweet spot exactly the same as all analog gear has. Otherwise it isn't an accurate model! No plug-in developer assumes their plug-in needs to be accurate when sent signals above 0dBFS. It is operating outside of the design and may or may not be a problem.

Yes, but my point was you as the mixer wetware have all the dynamic range to play with do whatever you like there. And I more wanted people not to assume there is yet another reason (plugin input levels) to have to track stuff at some specific level.

---

I'm not sure how many threads these never ending questions about tracking input levels takes for everybody here to try to help with from every possible angle but what I think new folks need to do here....

Have a read about some basic stuff on conversion, understand conversion bit depth and dB dynamic range, dB gain, dbFS, dbV, dBu, Vrms analog signal levels. (Maybe grab a voltmeter or scope and a signal gen plugin and confirm you understand this stuff in practice).

Get decent monitors/headphones/amps, understand how to get stuff turned up loud with the a volume knob not the input gain. (Seems a perennial issue in posts on DUC).

Start by not tracking too hot to avoid clipping. Understand dbFS and what the meters are telling you. Watch stuff in dbFS. Track at -16dBFS?, -20dBFS? I don't know, who really cares., but not normally -3 dBFS or -6 dBFS.

Listen carefully with your ears to analog front-ends, and what is happening with plugins. And yes OK realize that plugins as well as analog electronics have sweet spots.

If in doubt do some A/B tests for yourself. if you prefer something, do that.

[Chief Technician]

Originally Posted by Park Seward

I remember sending music files to Pro Tools from a Digital Betacam tape. The DB tape was recorded to the broadcast standard of -18 dBFS. The PT mixer raised everything to up to get close to "0" and then send the mixed file back to me and the DigiBeta. It was now too high, almost hitting clipping.

What he should have done was played the 1kHz calibration tone on the DB tape and used that to either align to an analog 0VU or to match the same -18 dBFS in Pro Tools.

Originally Posted by Park Seward

A sound recorded at -20 will sound the same at -50. It will just have a little more noise that you will not hear, especially if you dither.

That is only true if you apply +30dB of gain to the monitoring path to offset the -30dB of gain. Even then, I am skeptical that the audio will sound the same, since everything you could not hear before was just boosted by +30dB of gain.

[JuanPC]

Listening level is usually around 80dBSPL.
converters have -120dB dynamic & noise floor range.
And 24-bit files have -144dB.

The minimum recording level should be exactly the minimum audible...
The maximum should be exactly the maximum tolerable.
For example:
Adam a7x pair, has a Max SPL of 114dB @ 1mt.
108dB @ 2mts.

Minimum recording level should be -108dB.

Recording level and Playback level should be 1:1

http://www.adam-audio.com/en/pro-aud...technical-data
Recording level should be more than: -100dB.

People record hot because loudspeakers are not hot, To compensate overcompensate.

Recording a -6dB signal but listening at 50dB SPL, is not the same as listening at 110dBSPL.

People loose the sense/perception of loudness when listening at low volume.

When playback level is too low is not possible to detect errors as easy.

[Drew Mazurek]

Originally Posted by JuanPC

People record hot because loudspeakers are not hot, To compensate overcompensate.

No, people record too hot because they're looking at digital peak meters combined with the myth that you should record as hot as possible.

If everyone was forced to use real analog VU meters, we wouldn't be having this conversation because 0dBVU (which is -18dBFS) looks and feels "hot".