Loudness or, How to measure what doesn't exist?

[AudioPostExpert]

I'm starting this thread in response to another thread where an IT member was using audio (dB) levels as a reference to loudness. I pointed out that there is no correlation between dB levels and loudness, which eventually led to several members expressing interest in knowing more about loudness and the way it is measured.

To many here this subject may appear utterly uninteresting and arguably it is! However, the consequences of this dull technical subject have affected a US Presidential campaign and resulted in new laws and/or regulations of TV production and broadcast in virtually all "Western" countries in the last couple of years. Furthermore, these regulations are starting to spread to online distribution and there has been discussion of them in commercial theatrical circles too. So, boring or not, it's something every filmmaker should know a bit about and aspiring professional filmmakers should have a pretty reasonable understanding of. So, IMHO, it's worth having some reference material on the subject available on IT.

I'm going to explain a little of the history about where we are, how we got here and why it was a big enough problem to involve a US President, various international bodies and national governments.

So far, this story sounds mildly interesting but in all honesty, it really isn't! It's quite technical but I'll do my best to keep it simple when I actually start explaining in the next post.

G

 

[stef]

I think you must have misunderstood? There can be no correlation between dB and loudness if measuring loudness also depends on another unrelated measurement, the frequency in this case.

Sure it can. If loudness depends on db, and it does, then it is correlated. It also depends on the frequency of the track.

In other words, there is no way of ascertaining loudness from a dB value alone and therefore there is no correlation.

That's not what correlation means, at least not where I am or in any of the scholarly papers I've read. Correlation is a statistical term to determine if something depends on something else, and it's provable to a specific level of confidence. Some things are so closely related, there's little need to run a regression on it. Loudness, by definition, is a function of dB, frequency, and ear sensitivity to that frequency, and thus, correlated to all of those. You cannot have loudness without dB. It's a necessary component, and I'd even claim, dB is highly correlated to loudness. You cannot determine loudness of a track without knowing the dB (among other things).

I believe your point is "You can't determine loudness from dB alone, thus a VU meter is useless in determining loudness without more information", and I get it. But the proof of correlation is simply, "If you listen to a track, then replay it and increase the dB, does the perceived loudness increase or decrease or stay the same?" In all cases (except maybe extremely low frequencies where our ear response is very poor), the perceived loudness always increases with dB. Bingo: correlation. The graph you posted demonstrates this correlation... loudness increases as db increases across every single frequency range. If it didn't, the lines would cross each other a lot, and only then would there be some question to correlation.

This is just a nomenclature nit-pick that I noticed, and in no way am I arguing any of your points. I was merely being pedantic about the term. It struck me as odd because people are generally far too quick to claim a correlation between two unrelated things. I see false correlations drawn on fb all the time, and this may be the first true non-correlation claim I've seen.

I am going to go back and comb through this thread to answer my other questions, and I sincerely appreciate you posting it! If only I had been able to use this as a primer 3 months ago... would've saved me a ton of research.

 

[stef]

This might all sound like a bit of a nightmare to the uninitiated and I suppose it is, but it's still way easier than it used to be. If all deliverables do become standardised, that might not be such a good thing for us audio post professionals because some producers might decide it's easy enough to be handled by the video editor for a few extra bucks rather than having to pay knowledgeable audio post pros to produce compliant mixes. Fortunately though deliverables won't be standardised in the foreseeable future!

G

It doesn't just sound like a nightmare, it is a nightmare for the uninitiated. A video editor could do it with the right tools and perfect studio sound, but you guys have job security. Anything that's difficult, and there's going to be a lot, will require a pro. Edit: and that doesn't even count full sound design. I've cringed at what an editor did to one of my films, and I see myself making the same mistakes when I try to do it.

You can see the frustration that bit me in this thread here, where APE, AA, and others helped me out. This thread will go down as one of the most important threads for indies thinking of broadcast TV that nobody knows they need... yet.

In my case, while I understood the concept of "loudness" existed, it was the first time I ever saw it used or limits defined, and you can imagine my complete ignorance when reading a broadcaster's spec. They aren't joking around, either, and they basically will not air your show if you ignore their reqs. While I kind of understood the basics of a VU meter, I didn't know a LUFS from a dBTP from a K-weight. Apparently, Dolby had a chunk of hardware (loudness meter) that studios were using for a long while, and for some time I was thinking it was impossible to measure loudness without one.

After many searches without really knowing what to search for, I found a few defined broadcaster's specs, which gave me a starting point to look up how to determine the units and such. That's when I came across Orban, the free loudness meter, too. But the background and more importantly, the definitions and specs that are given in this thread that could've saved me a couple days of wasted time looking for the wrong things.

 

[AudioPostExpert]

Correlation is a statistical term to determine if something depends on something else, and it's provable to a specific level of confidence.

Ah, you are using the word "correlation" in it's general sense. In the audio world "correlation" has a much more specific meaning which would probably align more with a general meaning of "direct correlation". In audio we often talk about "correlated signals" and "de-correlating signals" and it's also used commonly in digital audio theory in regards to dithering and quantisation error. For example, a signal at 80dBSPL could sound uncomfortably loud or it could sound almost inaudible, which is a pretty extreme range of loudnesses for the same dB level! By the same token, you can easily have a sound at 60dBSPL which sounds far louder than another sound which has double that amount of SPL (66dBSPL) and of course vice versa. So, there is no direct correlation between dB and loudness. Only by changing the dB scale to a different scale can we approximate loudness. And, although I've avoided mentioning it, there is another set of variables which affects our perception of loudness which is NOT accounted for in the new loudness measurements! In other words, even the new loudness measurements are only approximately accurate and even then, only within a certain range of conditions!

The graph you posted demonstrates this correlation... loudness increases as db increases across every single frequency range. If it didn't, the lines would cross each other a lot, and only then would there be some question to correlation.

Ah, but draw a straight horizontal line (to represent equal loudness) at say 60dB, and that line bisects 7 different contours, even though each successive contour (from the bottom up) represents roughly double the level of the previous contour!

Apparently, Dolby had a chunk of hardware (loudness meter) that studios were using for a long while, and for some time I was thinking it was impossible to measure loudness without one.

It used to be impossible to meet many broadcasters' specs without a Dolby meter but now the ITU-R BS 1770 has superseded the Dolby measurements, although DialNorm still has to be set as all HDTV is broadcast with Dolby Digital! VU Meters are not related to any of this loudness discussion though, either the ITU or the Dolby predecessor.

...you can imagine my complete ignorance when reading a broadcaster's spec. They aren't joking around, either, and they basically will not air your show if you ignore their reqs.

It won't even get as far as the broadcaster if the audio doesn't precisely match the specs! Most broadcasters have some sort of ingest department where QC compliance is measured and if it fails the program is simply rejected. In many cases broadcasters use specialist 3rd party QC companies so your film/program can be rejected before the broadcaster even gets to see it. I know the indie philosophy is usually "ähh, there's always a way around it" but in this case there really isn't!!

This thread will go down as one of the most important threads for indies thinking of broadcast TV that nobody knows they need... yet.

It's even more important than that because it's not only broadcast TV anymore! Loudness normalisation has already been implemented in iTunes radio and Spotify. It is currently being worked on by the EBU for all European radio stations and it's likely to start being introduced at some stage by VOD distributors and even possibly for theatrical presentation. So, there may come a time in the not too distant future where pretty much every commercial distribution channel will require a good understanding of loudness normalisation.

G

 

[rayw]

After many searches without really knowing what to search for, I found a few defined broadcaster's specs, which gave me a starting point to look up how to determine the units and such. That's when I came across Orban, the free loudness meter, too.

A rashy pox upon you fellas for "making" me go hunt down this link*!

http://www.orban.com/meter/



* FWIW, said link hunt took all of ten seconds.




Alright, G, just through sheer repetive reviewing I'm picking up on the subtle difference between dBFS and LKFS/LUFS et al.

But... I'm still having a hard time, working with my NLE's or freeware DAW's crude dB & RMS adjustments, to not just call anything red lining, resulting in a flat-topped clipped output, as "too loud" or "the volume's too high."

However, as a result of this four parter class I can now clearly see the GIGANTIC leap in audio standards from youtube to broadcaster demanded specs.

Yeah, there's no real way to use nickel and dime audio software and/or hardware to provide what broadcasters and top tier festivals require.
There's a pretty big gap.

Thank you.

 

[rayw]

Yeah, there's no real way to use nickel and dime audio software and/or hardware to provide what broadcasters and top tier festivals require.

Eh... not so fast...

http://www.softpedia.com/get/Multimedia/Audio/Other-AUDIO-Tools/Audio-Loudness-Meter.shtml

And:
http://www.softpedia.com/get/Multimedia/Audio/Other-AUDIO-Tools/

 

[AudioPostExpert]

A rashy pox upon you fellas for "making" me go hunt down this link*!http://www.orban.com/meter/
* FWIW, said link hunt took all of ten seconds.

Why did you go and "hunt it down"? I put a link to the Orban meter in Part 4!

But... I'm still having a hard time, working with my NLE's or freeware DAW's crude dB & RMS adjustments, to not just call anything red lining, resulting in a flat-topped clipped output, as "too loud" or "the volume's too high."

It's an important distinction because you could have a "sound" where the level is so high that the signal has been clipped AND AT THE SAME TIME is so quiet it is inaudible.

However, as a result of this four parter class I can now clearly see the GIGANTIC leap in audio standards from youtube to broadcaster demanded specs. Yeah, there's no real way to use nickel and dime audio software and/or hardware to provide what broadcasters and top tier festivals require. There's a pretty big gap. Thank you.

You're welcome. Bare in mind that I've only discussed in this thread one aspect of the technical specs of broadcasters, loudness. In all honesty, this loudness spec is pretty easy to meet (with the right meters). There are other tech specs for many broadcasters (and distributors) which are much more difficult (and expensive) to meet, creating an acceptable M&E mix for example. And, even with all this we are still only talking about MINIMUM, common denominator, technical specifications. In theory we could meet these loudness specs with an otherwise crappy mix, which would fail QC on other technical or artistic aspects. I haven't mentioned these other aspects of audio QC because they can vary from broadcaster to broadcaster. Also bare in mind that the top tier film festivals (and theatrical films in general) do not use this new loudness paradigm and it's not absolutely certain that they ever will. The loudness of films is controlled, but it's much harder to comply with because it's based on an experienced professional set of ears and a highly calibrated monitoring environment, which of course is way, way more expensive than even a $500 software plugin!

Eh... not so fast...

http://www.softpedia.com/get/Multimedia/Audio/Other-AUDIO-Tools/Audio-Loudness-Meter.shtml

And:
http://www.softpedia.com/get/Multimedia/Audio/Other-AUDIO-Tools/

I'm not sure what you mean/what your question is? The first link is to Nugen's loudness meter which is definitely one of the better loudness meters on the market and the second link is to a page of general audio software tools.

G

 

[rayw]

Why did you go and "hunt it down"? I put a link to the Orban meter in Part 4!
Pfft.
No brains, no headaches.

One of my minor petty grievances with this forum's site set-up is that the links are underlined just like plain old underlined words, and I'd hazard a guess only a third of the underlined words and text are linked to something.

I just failed to click on those two.


It's an important distinction because you could have a "sound" where the level is so high that the signal has been clipped AND AT THE SAME TIME is so quiet it is inaudible.
Principle understood.
I've yet to run across it, but then my experience is minuscule relative to a professional's, such as yourself.


You're welcome. Bare in mind that I've only discussed in this thread one aspect of the technical specs of broadcasters, loudness... There are other tech specs for many broadcasters (and distributors)... Also bare in mind that the top tier film festivals (and theatrical films in general) do not use this new loudness paradigm and it's not absolutely certain that they ever will.
Understood.
Good to know.


I'm not sure what you mean/what your question is?

Oh, I was just hunting around for some free loudness history programs, if such were even available, and ran across those.
Thought someone else might benefit from them and the other tools.
No question, just sharing.​

 

[AudioPostExpert]

Oh, I was just hunting around for some free loudness history programs, if such were even available, and ran across those.
Thought someone else might benefit from them and the other tools.

Bare in mind that the Nugen VisLM-H plugin is NOT free, it costs $449. You linked to a free, time limited trail version!

I've yet to run across it, but then my experience is minuscule relative to a professional's, such as yourself.

While you probably haven't run across this extreme example, you certainly will have run across less extreme examples where your sound had a fair amount of low frequency energy which made the dBFS levels higher than you might have expected relative to it's loudness. You just might not have noticed or put it together though. Just for anyone else's benefit and to clarify, these loudness measurements affect pretty much all sounds, not only those with a lot of bass content, although those with a lot of low bass content will be affected the most.

G