I've - perhaps wrongly - considered VU meters roughly RMS meters, that is, giving me an average level affected in a dynamics processor. So -3dB on a RMS meter or a VU is the amount on the average of signal attenuated. Whereas a peak meter is "we shaved -3dB off the transients".

Live and learn, obviously I've got a bit to learn here.

 

No, no, that's correct, that's why I wrote On one hand this is correct, but there are some minor differences in the exact way of the measurings. So pretty much the same, but not identical.

 

same thing, only different.

 

Thank fuck I'm not a ME. But thanks, I had enough problems figuring out the various LUFS weightings but at least I know my basic approach wasn't wrong.

 

Roughly is the key word. RMS is not wrong, but inaccurate. All digital loudness standards are based on dBFS RMS. VU reaches 99% of the FS RMS value in 300 ms and has an overshoot of min 1%- max 1.5%. It is therefore RMS with modified ballistic.

 

VU meters are irrelevant if you are using a DAW

 

I always thought a VU meter displays the average volume level of an audio signal and thought VU stood for volume unit.
I also always found it useful visually no matter the idiom.
I have read other comments and will leave it as "I have read other comments". There are different quality and accuracy levels of third party vst VU meters, and meters in a DAW, but a good one is always useful for checking your work (for me), especially the ones that also have switchable PPM.

 

+4 dBu

 

VU (Volume Unit) meter is a Standard Volume Indicator (SVI). It shows loudness expressed in dB. The zero level of VU meter should not be confused with 0 dB - the threshold of hearing. Zero indicates the maximum distortion-free level that can be handled by the device.

https://producelikeapro.com/blog/vu-meter/

 

= -14dBFS ?

 

Last edited: Sep 17, 2021

https://audiosex.pro/threads/can-so...-a-vu-meter-in-a-daw.60633/page-3#post-582852
'they cannot find intersample peaks due to the vey slow release time'
Who has written this shit??? He's talking about the digital domain and DAWs and the real reason why nothing can find intersample peaks is because they don't exist! It's a prediction, not a measurement.
And BTW, peak is the most important value in the digital domain because if you really screw this up no skill can save your mix (I hate them with a passion - what a bullshit).

loudness over 400ms ... extremely useful as we can gauge what is happening ... over the last 400ms
loudness over 1-3 seconds ... extremely useful as we can gauge what is happening ... over the last 1-3 seconds

Logical, no doubt. How about
loudness over 20ms ... extremely useful as we can gauge what is happening ... over the last 20ms
or
loudness over 800ms ... extremely useful as we can gauge what is happening ... over the last 800ms
or maybe more important
loudness over 1-3 minutes ... extremely useful as we can gauge what is happening ... over the last 1-3 minutes

This allows us to make mix critical edits
Now I understand why every single mix from the pre-LUFS area sounds like shit. They couldn't make these edits.

Really, mate, it's been a long time that I've read such a nonesense. I know, not your words, just a pic, but utterly nonsense (as an explanation/reason).

 

Be nice. If you compare the mixes of various periods in musical history and I will use George Martin as one example only, some were extraordinary considering what they had to work with, also keeping in mind they only had stereo 2-tracks and then 4-tracks.
Sounds like shit is an opinion too

I know of one engineer who has awards in mixing who was employed to mix the last Geri Allen CD before she passed away. He made it perfect spectrally. It made acoustic jazz sound like a dance mix. The kick drum in acoustic jazz plays accents, the pulse is on the ride cymbal and closed hats. He boosted the kick to sound like EDM. Blue Note Records unceremoniously fired him because it sounded like shit because he played God and forgot the essence of the style and stuck to theoretical guidelines instead of bothering to learn the style. But that too, was only Blue Note's opinion, not everyone's.
I know a lot of other engineers who do the same thing and they're all out of work. It comes down to one thing, what the client wants.
So consider the style and the period PLEASE.

As for who writes this shit as you put it - engineers, unfortunately. I am sure if you google it you''ll find it and include audio engineer in your search because that's what came up - mate.
All a musician cares about is that it sounds like they heard it in their head and they can see what is happening when it sounds right or wrong. Not how technically perfect it is. That is your job, to give them what they want while retaining what you know to be right without compromising their needs.

 

Last edited by a moderator: Sep 17, 2021

A vital point is where to set 0dB. In Hi-Fi you were supposed to meet the (now obsolete) Hi-Fi standard of less than 0.1% of total distortion (Klirrfaktor). So setting the zero in a particular Hi-Fi amp was meant to indicate the max. loudness level you can get with that amp without trespassing the mentioned Hi-Fi standard.
With tape recorders it meant mostly the loudness at which the tape saturation begins while recording. (Of course every type of saturation leads to analogous clipping, but it is not abrupt, but progresses slowly).
So significant clipping was not due to occur in the red zone, but it indicated a higher and higher probability that it might happen, and also that unwanted increase in distortion will happen for sure.
Therefore meaningful setting of the VU zero is relative, i.e. suited for a particular piece of gear, in order to help user to stay within the standards (if he wanted to) when using that gear.
In digital realm with no quality standards it is just a RMS meter with no further meaning, and more of an ornament, to persuade users that they deal with a truly 'analogous' modelled software.
Only FS shows (each and every) /digital/ clipping.

 

Well, actually, it's both. It is a prediction of the D/A signal, but a valid measurement of the D signal. Nevertheless, we are always in the speculative range with dBTP.

If someone wants to read more about dBTP and Loudness Standards, you can do so here:
Design Specifications for a DSP Broadcast Mastering System

Also funny, because the peak detection has nothing to do with release. important is the attack and PPM has an attack <1ms. It might actually be difficult with QPPM, with its 10ms attack. But since all DAWs I know use SPPM, I don't see his problem.

VU meters being the closest to the logarithmic hearing.
Well, then I wonder why the EBU R 128 specification was developed.

 

Last edited: Sep 17, 2021

You guys should probably take it all up with Eddie Bazil. He's the author apparently from where that screen capture came.

 

Last edited by a moderator: Sep 17, 2021

Exactly.
Broadcasters do not use VU meters in their studio for measuring their audio signals

The reason for using a VU meter nowadays is pretty much either:

Or they simply do not know how to measure transients more efficiently for their so called "sweet spots".

 

Last edited: Sep 17, 2021

You do realize that my comment was meant ironically, don't you? There're a bunch of great mixes from the pre-LUFS area, probably even more than from the LUFS area.

These measurement don't change according to the style. LUFS = LUFS in Jazz and HM.
The period? Well, they didn't have it in the 70s, that's right.

Means? They can't be wrong???

Absolutely true, but I can't see the context to your post or my reply.

 

Oh! Inter-sample means between the samples and between the samples there's nothing. How can this be measured? Calculated over several samples, I'll give you that, but measured between?

 

It's been a while since I've been in a broadcasting studio but I assume VU meters are still used for analogue broadcasting and LUFS for digital?