Topic: Record Levels. -12 db?

[Fredman]

Greetings, all!  First Third post here.  I come from the old school of taping, ran hundreds of hours using state of the art (for the 1980s!) analog gear (D5>Nak 300+CP4).  I pretty much fell out of the taping scene in the late 90s and am just now getting back into it.

I've purchased an M10 and have the Church Ca14 package on the way.

Back in the day the rule was to run your levels at 0 db, with some peaks up to +3 (maybe a little higher with metal tape).  Everything I'm reading says to run the M10 to an average peak around -12db.

My question is Why?  Seems like there's some missing headroom with running levels that low.  Is there a concise FAQ on recording levels and why?  Any help would be much appreciated!

[adrianf74]

Hi Fredman,

When it comes to recording, digital is a much different animal compared to analog.  With analog, you have headroom that was sufficient while in the digital realm, the headroom is technically to "zero" -- anything beyond that could clip/distort/etc. and it's nowhere as forgiving as the analog world was.

The notion of -12dB for peaks is a mere guideline.  Assuming you'll be recording 24-Bit, 48kHz, you have the ability of raising your levels in post instead of risking clipping or distortion.  Unlike analog tape, you'd be increasing very little hiss at -12dB if it needed to be raised +4 or +5dB.  At some shows, -12dB might be a touch low because the show is either not very loud or is very well compressed but you never know if you're gonna have some yahoo next to you that feels the need to whistle or yell right into your mics.  A lot of it'll be trial and error but a lot of people here run as "hot" as -8db or -6dB for peaks; I normally wouldn't but that's just me.

As long as you're rolling in 24/48, you'll be good at -12 to -10dB.

[Cheesecadet]

I generally run my levels on my M10 around -12dB - -10dB with the peaks up around -6dB FWIW.  I like to run a little hot tho

[Fredman]

That makes perfect sense, with being able to normalize without introducing noise.  I guess it's all about the S/N limitations of analog tape vs the lack of those limitations with digitial.

I'm beginning to transfer my old analogs to the M10, at 24/48, and looks like I'll keep them around -12 then bump up as needed in Audacity.

Hi Fredman,

When it comes to recording, digital is a much different animal compared to analog.  With analog, you have headroom that was sufficient while in the digital realm, the headroom is technically to "zero" -- anything beyond that could clip/distort/etc. and it's nowhere as forgiving as the analog world was.

The notion of -12dB for peaks is a mere guideline.  Assuming you'll be recording 24-Bit, 48kHz, you have the ability of raising your levels in post instead of risking clipping or distortion.  Unlike analog tape, you'd be increasing very little hiss at -12dB if it needed to be raised +4 or +5dB.  At some shows, -12dB might be a touch low because the show is either not very loud or is very well compressed but you never know if you're gonna have some yahoo next to you that feels the need to whistle or yell right into your mics.  A lot of it'll be trial and error but a lot of people here run as "hot" as -8db or -6dB for peaks; I normally wouldn't but that's just me.

As long as you're rolling in 24/48, you'll be good at -12 to -10dB.

[yates7592]

That makes perfect sense, with being able to normalize without introducing noise.  I guess it's all about the S/N limitations of analog tape vs the lack of those limitations with digitial.

I'm beginning to transfer my old analogs to the M10, at 24/48, and looks like I'll keep them around -12 then bump up as needed in Audacity.

[

The -12dB rule of thumb works well when recording rock concerts with a digital recorder. But when transferring old analog tapes to digital, I would personally try to get the levels much closer to -1dB, given that you're in a much more controlled environment (you could run trials to determine the optimum record level), and you don't want to add any more noise to an analog tape.

[John Willett]

Greetings, all!  First Third post here.  I come from the old school of taping, ran hundreds of hours using state of the art (for the 1980s!) analog gear (D5>Nak 300+CP4).  I pretty much fell out of the taping scene in the late 90s and am just now getting back into it.

I've purchased an M10 and have the Church Ca14 package on the way.

Back in the day the rule was to run your levels at 0 db, with some peaks up to +3 (maybe a little higher with metal tape).  Everything I'm reading says to run the M10 to an average peak around -12db.

My question is Why?  Seems like there's some missing headroom with running levels that low.  Is there a concise FAQ on recording levels and why?  Any help would be much appreciated!

This is because anything over 0dB is distortion and clipping.

For a 16-bit system the standard is to have -12dBFS as your 0VU setting and then peak a little over this.

For a 24-bit system the standard is to have -18dBFS (Europe, -20dBFS in the US) as your 0VU setting and then peak a little over this.

This allows for headroom without distortion - it also allows for inter-sample peaks to be recorded without distortion as these may not register and be higher than indicated.

I hope this explains it clearly for you.

[fguidry]

That makes perfect sense, with being able to normalize without introducing noise.  I guess it's all about the S/N limitations of analog tape vs the lack of those limitations with digitial. ...

Actually the reason for the difference is that they are two completely different scales. Your +3 was referenced to 0 dBVU which was an RMS oriented metering system. Digital systems are calibrated to a peak level of "all bits on" as 0 dBFS (dB Full Scale). Normal calibration for digital systems varies, but a common choice has -18 dBFS at a level similar to your 0 dBVU. So a -10 or -12 dBFS level is actually hotter than your +3 dBVU.

It's quite likely that a lot of "bad digital sound" is related to a misunderstanding of the relationship between these two metering concepts.

Fran

[stevetoney]

IMHO, nobody's given the basic answer for running -12db with todays recorders.  Back in the day you were recording in 16bit technology.  Today even the lowest end flash recorders use 24bit technology.  Back in the day we ran hot to make sure we got the full dynamic range on our recordings.  However, the resolution provided with 24bit means you don't any lose dynamic range if you record with alot of headroom and then bump the levels in post. 

The second concern to running with lots of headroom was decreasing the SNR and amplifying the noise floor in post.  However, noise specs have also improved on gear and most people find that, unless you record nature sounds, for the type of recording most around here do, the noise floor never becomes an issue.  I don't always run at -12db, but I have no concern bumping my levels that much in post (or even more) as long as I've recorded in a 24bit format.

[stevetoney]

When it comes to recording, digital is a much different animal compared to analog.  With analog, you have headroom that was sufficient while in the digital realm, the headroom is technically to "zero" -- anything beyond that could clip/distort/etc. and it's nowhere as forgiving as the analog world was.

Running with extra headroom is not a digital vs. analog issue.  For example, the guideline was to still to run hot levels when we were using a 16bit Sony d8 DAT which is of course digital technology.  See my previous response. 

[runonce]

That -12db mark has been in the Sony literature since the first portable DATs.

Never quite got what that was about...and remember, back then - nobody had PC editing abilities..so it wasn't like "Oh I'll just normalize later..."

I always assumed it was an RMS type mark - and read it as "at least -12db"

In practice - I think most of us ran it just short of the "over" light (otherwise...what's the point of that light?!)

At the time - I wondered if it was a level that, on playback, would guarantee it wouldn't overload any downstream analog gear, while still producing adequate signal...probably wrong about that.

[stevetoney]

That -12db mark has been in the Sony literature since the first portable DATs.

Not the same.  Sony's -12db was just a demarcation on the scale.  Tapers never used -12db as a basis or a guideline to run at when we were using 16bit DATs.  You gave up too much.

It was only when 24bit came on-board that tapers accepted that you could safely run with that much headroom without giving much up.  The tradeoff became worth it.

[runonce]

That -12db mark has been in the Sony literature since the first portable DATs.

Not the same.  Sony's -12db was just a demarcation on the scale.  Tapers never used -12db as a basis or a guideline to run at when we were using 16bit DATs.  You gave up too much.

It was only when 24bit came on-board that tapers accepted that you could safely run with that much headroom without giving much up.  The tradeoff became worth it.

Gotcha - just saying that's what the manuals said/suggested...I think the recorders even had a dot on the 12.

I always ran short of the over light...(and like a lot of others - eventually put the D5 back in front and used it's pres.)

[stevetoney]

Maybe this will help...

http://www.homestudiocorner.com/24-bit-vs-16-bit/

[stevetoney]

Jon you live in a different dimension than many.  Sometimes simplified examples and just keeping it simple can be helpful. 

[cavekelly]

I have just recently began using an M10 for recording.  Most of my projects are of the stage-lip, jazz variety.  The music often has a very wide dynamic range.  I have been pushing my levels well past -12 ( but no overs) using the logic that the softer portions are then being recorded at as high a level as possible.  Perhaps I am reverting back to my old D5 analog days, but this makes sense to me.  Thoughts?