When they say its good to record at -18dbfs they mean peaks right?

[meymia]

When they say its good to record at -18dbfs they mean peaks right?

[Recording David]

It's an ideal way to calibrate your AD converters against a 1 kHz 0VU signal coming from a +4 analogue mixer. So 0VU (analogue) = -18 dBFS.

[Jack Ruston]

No. The calibration level refers to the average operating level in analogue. You would typically see peaks maybe 10dbs higher.

[mikeon]

I thought it was -18dbfs as average signal strength.

[Waltz Mastering]

Originally Posted by meymia

When they say its good to record at -18dbfs they mean peaks right?

I think "they" could have been referring to average rms recording level.
-18 dbfs is also used as reference/calibration for converters. Who's "they"?

[dbbubba]

Originally Posted by Recording David

It's an ideal way to calibrate your AD converters against a 1 kHz 0VU signal coming from a +4 analogue mixer. So 0VU (analogue) = -18 dBFS.

ding, ding, ding, ding THE CORRECT ANSWER!

I had lengthy discussions with the top field tech for Sony Pro Audio Group out of the California office regarding this matter back in 1995.
He was at our facility in Dallas repairing our PC 3324 24 track DASH thing-o-matic
We hung for two days and we had eight FULL digital rooms running then.
I had to know because we had ten engineers doing hundreds of radio/TV spots weekly and we needed a standard that was set in stone and correct at the same time.

I still have a SONY DAT tape (that I have duped to a CD) which he made me specifically for calibration of digital equipment.
It was a DAT made in the Sony main shop out there on their calibration / test rig.
For their units to poot out a +4 signal the digital signal is -18dBFS.
So, a sine wave 1Khz tone should have the meters sitting at -18 dBFS.

The reason I am stressing the SONY practice of -18dBFS is because it was SONY and Phillips in a joint agreement that set most all of the standards we are SUPPOSED to use in digital audio.
Because of the design of the actual chip sets used in other gear, this practice IS NOT adhered to 100%.

This is how I have my stuff calibrated to this day, BUT.... BUT....
I know for a FACT that MOTU doesn't really adhere to this standard.
I had a tech at MOTU admit and explain why... I forget the details now.
I got them to admit that the headroom on their stuff is a bit lower.
I also know for a FACT that DigiDesign's converters are NOT set up to adhere to this either.
A robust signal (drums) WILL often show clipping on the meters and waveform when the units are set to -18 dBFS.
You can argue this point, but I know too many people who have confirmed this to be true.

[meymia]

Please simplefy it for me.

If i have a blur robbie preamp (wich doesnt have a vu meter) and i record vox into cubase 5.


I just tell the singer to sing and i start turning the gain knob on the robbie till i see in the channel on cubase what? -18? and thats the robbie's 0?

[JoeyM]

Originally Posted by dbbubba

...
A robust signal (drums) WILL often show clipping on the meters and waveform when the units are set to -18 dBFS....

Is that phenominon worse if the drums are gated?

I wonder because, you'd think it there's more signal mojo going from zero to -18dBFS.

Every wrong attempt discarded is another step forward - Edison

[Recording David]

Originally Posted by meymia

Please simplefy it for me.

If i have a blur robbie preamp (wich doesnt have a vu meter) and i record vox into cubase 5.


I just tell the singer to sing and i start turning the gain knob on the robbie till i see in the channel on cubase what? -18? and thats the robbie's 0?

OK, if you're simply recording straight from the Robbie into your ADC/audio interface the easiest way to set the gain would be 'by hand' keeping an eye on your digital meters in Cubase (this is the dBFS reading). If you're recording at 24 bit (and you should be) just make sure your peaks don't go over about -8 dBFS on the input (i.e. the input meter in Cubase).

[u b k]

Originally Posted by JoeyM

I wonder because, you'd think it there's more signal mojo going from zero to -18dBFS.

This depends completely on the design of the analog input stage. Some boxes, like the 2192 and Burl, were designed to sound good when hit hard by saturating and distorting in a sweet, analog way. Others, like Mytek or Pacific Microsonic, are just awesome circuits and stay totally clean thru to digital clipping, and even the clipping doesn't sound bad in small doses.

But on a lot of converters the analog stage has a sweet spot where it has the most transient fidelity and least distortion, and this is often found near the conservative input levels: RMS around -18, peaks around -10.

You gotta go by ear for this, see where your converters like to breathe.


Gregory Scott - ubk
.

[JoeyM]

Thanks Gregory for the answer and everyone else for a thought-provoker in an area I feel fairly illiterate.

One night at a time