You are here

Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

Page 1 of 2

Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 3:13 am
by hobbyist
It would be really useful if they built an SPL meter that had weighting for analogue loudness matching LUFS/LKFS in the digital world.

The real factor controlling analogue loudness is how far did somebody rotate the knob on the amplifier.

At home we can rotate to our taste.
But for a venue it would be nice to set the loudness to an objective standard once knowing that any digital content using a standard measured with LUFS would sound the same, loudness wise, in the venue when coming through our amplifier.

Anybody ever took an SPL meter apart to see how easy it could be modified?

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 10:27 am
by Hugh Robjohns
hobbyist wrote:It would be really useful if they built an SPL meter that had weighting for analogue loudness matching LUFS/LKFS in the digital world.

LUFS (LKFS) relates only to the full scale digital encoding and bears no relationship whatsoever to acoustic sound pressure level... so it would make no sense to have an LUFS scale on an SPL meter...

However, the idea of maintaining a consistent average acoustic loudness between venues (and production suites) is a sensible one... and it has been in place in film dubbing theatres and cinemas for decades, as well as in professional broadcast sound control rooms etc.

All that is needed is to establish a reference acoustic sound pressure level for a given digital signal level (the latter also relating to the desired target loudness as per R128 etc)...

H

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 4:56 pm
by hobbyist
Hugh Robjohns wrote:
hobbyist wrote:It would be really useful if they built an SPL meter that had weighting for analogue loudness matching LUFS/LKFS in the digital world.

LUFS (LKFS) relates only to the full scale digital encoding and bears no relationship whatsoever to acoustic sound pressure level... so it would make no sense to have an LUFS scale on an SPL meter...

However, the idea of maintaining a consistent average acoustic loudness between venues (and production suites) is a sensible one... and it has been in place in film dubbing theatres and cinemas for decades, as well as in professional broadcast sound control rooms etc.

All that is needed is to establish a reference acoustic sound pressure level for a given digital signal level (the latter also relating to the desired target loudness as per R128 etc)...

H


Thanks.

But now I am confused. But that is just me.

We do adjust to SPL levels. But I thought the theory behind LUFS was that SPL does not relate to perceived loudness.

Why could not the audio sound pressure be weighted with something other than A or C (or some other rarely used speciality ones) so as to get an idea of actual loudness.

I guess you are saying that loudness is what is set in digital and it is up to adjust the spl loudness. But that is the confusion here. Isnt taht why LUFS was invented so we did not have to keep adjusting the volume to keep the loudness the same for the listeners?

I guess we need a standard that says a program that is x LUFS should generate y pascals SPL under A (C? other?) weighting; or something equivalent.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 5:40 pm
by Hugh Robjohns
hobbyist wrote:I thought the theory behind LUFS was that SPL does not relate to perceived loudness.

It's more complicated than that. Clearly, a higher SPL sounds louder than a quieter one... but it's also about the ratio between peak and average SPLs -- or more specifically, sound energy and time...

Why could not the audio sound pressure be weighted with something other than A or C (or some other rarely used speciality ones) so as to get an idea of actual loudness.


A-weighting is all about indicating perceived loudness... but it's not as sophisticated as the technique used in BS.1770 .

There is nothing to stop the ITU, or EBU, or AES coming up with a method of measuring perceived programme loudness relative to a defined reference SPL... but they haven't chosen to do it.

Instead, they have agreed a way of measuring loudness within a digital file format relative to the maximum digital coding level. Other standards exist that relate digital audio coding levels to acoustic sound pressure levels for different applications.

Isnt taht why LUFS was invented so we did not have to keep adjusting the volume to keep the loudness the same for the listeners?

Partly. The whole BS.1770 programme is about maintaining a consistent replay volume for a given listener, to avoid the need to keep adjusting the volume, as you say. But there's no requirement for all listeners to listen at the same volume. Some may want it LOUD, others, quiet, and others somewhere in between -- all depending on their listening environment, activity, mood, and time of day. But whatever volume they choose to set their replay equipment, they shouldn't need to then keep on tweaking the volume as different tracks or programmes are selected.

I guess we need a standard that says a program that is x LUFS should generate y pascals SPL under A (C? other?) weighting; or something equivalent.

Would that be a defined SPL for listening in your car, or the kitchen, or a bedroom late at night, or a stadium concert?

H

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 7:31 pm
by hobbyist
Hugh Robjohns wrote:
hobbyist wrote:I thought the theory behind LUFS was that SPL does not relate to perceived loudness.

It's more complicated than that. Clearly, a higher SPL sounds louder than a quieter one... but it's also about the ratio between peak and average SPLs -- or more specifically, sound energy and time...

Why could not the audio sound pressure be weighted with something other than A or C (or some other rarely used speciality ones) so as to get an idea of actual loudness.


A-weighting is all about indicating perceived loudness... but it's not as sophisticated as the technique used in BS.1770 .

There is nothing to stop the ITU, or EBU, or AES coming up with a method of measuring perceived programme loudness relative to a defined reference SPL... but they haven't chosen to do it.

Instead, they have agreed a way of measuring loudness within a digital file format relative to the maximum digital coding level. Other standards exist that relate digital audio coding levels to acoustic sound pressure levels for different applications.

Isnt taht why LUFS was invented so we did not have to keep adjusting the volume to keep the loudness the same for the listeners?

Partly. The whole BS.1770 programme is about maintaining a consistent replay volume for a given listener, to avoid the need to keep adjusting the volume, as you say. But there's no requirement for all listeners to listen at the same volume. Some may want it LOUD, others, quiet, and others somewhere in between -- all depending on their listening environment, activity, mood, and time of day. But whatever volume they choose to set their replay equipment, they shouldn't need to then keep on tweaking the volume as different tracks or programmes are selected.

I guess we need a standard that says a program that is x LUFS should generate y pascals SPL under A (C? other?) weighting; or something equivalent.

Would that be a defined SPL for listening in your car, or the kitchen, or a bedroom late at night, or a stadium concert?

H

That would be a defined SPL at my ears. Different places might need to adjust because of the location.

Actually it would really be in churches and similar so that the audience is , on the average, happy. Adding upward compression to voices would fix the folks who claim things are too soft. normal compression on music would fix things for those who say it is too loud. Then setting an average 'loudness' level that most people were happy with would be much easier, but still a standard would be useful especially in EU where there are apparently some laws restricting loudness and at least one case where someone sued because their ears were damaged when it was too loud.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 8:27 pm
by Mike Stranks
hobbyist wrote:
Actually it would really be in churches and similar so that the audience is , on the average, happy. Adding upward compression to voices would fix the folks who claim things are too soft. normal compression on music would fix things for those who say it is too loud. Then setting an average 'loudness' level that most people were happy with would be much easier, but still a standard would be useful especially in EU where there are apparently some laws restricting loudness and at least one case where someone sued because their ears were damaged when it was too loud.

I thought churches had congregations... God save us from 'Performance Worship' :roll:

In my 40+ years of providing and working UK church audio I have yet to meet a congregation that's completely satisfied with audio levels. 30% think it's too loud; 30% think it's too quiet.

A colleague who does 'just right' levels in church audio was providing and operating the system for a big 'Convention' type event... Seats backed right up to the console. In an intermission a guy turned round and said, "Good sound, but a bit quiet." A couple of minutes later a guy in the same row but a few seats along turned and said, "Good sound, but a bit loud." With presence of mind my colleague then introduced them to each other and suggested they swap seats....

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 10:24 pm
by Sam Spoons
Mike Stranks wrote:A colleague who does 'just right' levels in church audio was providing and operating the system for a big 'Convention' type event... Seats backed right up to the console. In an intermission a guy turned round and said, "Good sound, but a bit quiet." A couple of minutes later a guy in the same row but a few seats along turned and said, "Good sound, but a bit loud." With presence of mind my colleague then introduced them to each other and suggested they swap seats....

Brilliant :clap: :clap: :clap: :clap: :clap:

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Thu Jul 11, 2019 11:17 pm
by hobbyist
Sam Spoons wrote:
Mike Stranks wrote:A colleague who does 'just right' levels in church audio was providing and operating the system for a big 'Convention' type event... Seats backed right up to the console. In an intermission a guy turned round and said, "Good sound, but a bit quiet." A couple of minutes later a guy in the same row but a few seats along turned and said, "Good sound, but a bit loud." With presence of mind my colleague then introduced them to each other and suggested they swap seats....

Brilliant :clap: :clap: :clap: :clap: :clap:


There is that problem until everyone wears a hearing assist device and sets their own level.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Fri Jul 12, 2019 9:51 am
by Mike Stranks
Sam Spoons wrote:
Mike Stranks wrote:A colleague who does 'just right' levels in church audio was providing and operating the system for a big 'Convention' type event... Seats backed right up to the console. In an intermission a guy turned round and said, "Good sound, but a bit quiet." A couple of minutes later a guy in the same row but a few seats along turned and said, "Good sound, but a bit loud." With presence of mind my colleague then introduced them to each other and suggested they swap seats....

Brilliant :clap: :clap: :clap: :clap: :clap:

I still chuckle at the memory of a large event where I was at the controls. The style was 'contemporary', the congregation older and 'traditional'. (Lack of thought and planning on the part of the organisers of course, but that's another story.)

I didn't notice it myself, but the keyboard player told me afterwards that she looked into the congregation after the first song to see a lady of mature years with a gents' handkerchief stuffed into each ear and flowing gently down over each shoulder... :lol:

I was sometimes told after a contemporary service that "It's not a disco you know", thus indicating that the speaker had never been within 500m of a disco in their life!

exactly my point

PostPosted: Fri Jul 12, 2019 9:23 pm
by hobbyist
Mike Stranks wrote:
hobbyist wrote:
Actually it would really be in churches and similar so that the audience is , on the average, happy. Adding upward compression to voices would fix the folks who claim things are too soft. normal compression on music would fix things for those who say it is too loud. Then setting an average 'loudness' level that most people were happy with would be much easier, but still a standard would be useful especially in EU where there are apparently some laws restricting loudness and at least one case where someone sued because their ears were damaged when it was too loud.

I thought churches had congregations... God save us from 'Performance Worship' :roll:

In my 40+ years of providing and working UK church audio I have yet to meet a congregation that's completely satisfied with audio levels. 30% think it's too loud; 30% think it's too quiet.

A colleague who does 'just right' levels in church audio was providing and operating the system for a big 'Convention' type event... Seats backed right up to the console. In an intermission a guy turned round and said, "Good sound, but a bit quiet." A couple of minutes later a guy in the same row but a few seats along turned and said, "Good sound, but a bit loud." With presence of mind my colleague then introduced them to each other and suggested they swap seats....



Thanks! That is exactly my point.

30% too loud - means on the hot end so compress it down
30% too low - means on the speakers so UPwards compress it higher

Then it would be easy to tweak the average if anybody still complained.

The problem is people assume that when half say loud and half say too quiet that they are doing good. Wrong! They are doing bad for both groups.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Fri Jul 12, 2019 9:54 pm
by Bob Bickerton
That’s really oversimplifying the situation.

Some people like to hear music at higher volumes, others prefer lower volumes and reducing dynamic range isn’t going to help!

Bob

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Fri Jul 12, 2019 10:23 pm
by AlecSp
Bob Bickerton wrote:Some people like to hear music at higher volumes, others prefer lower volumes and reducing dynamic range isn’t going to help!
This - completely.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Fri Jul 12, 2019 11:15 pm
by hobbyist
Bob Bickerton wrote:That’s really oversimplifying the situation.

Some people like to hear music at higher volumes, others prefer lower volumes and reducing dynamic range isn’t going to help!

Bob


Lowering DR allows us to have sufficient volume for intelligibility on voice while keeping the high end low enough to be safe.

What people do at home or in their car is on them. But I do want to provide a proper environment for our church when we do audio.

In my home I compress everything to a max 20dB DR and then normalise the volume level so I never have to keep adjusting the knobs on my stereo.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Fri Jul 12, 2019 11:43 pm
by Hugh Robjohns
hobbyist wrote:In my home I compress everything to a max 20dB DR and then normalise the volume level so I never have to keep adjusting the knobs on my stereo.

I'm not familiar with the broadcasting standards in the US, but it sounds a bit grim if you need to take that kind of extreme action.

In the UK, radio and TV broadcasters generally maintain a dynamic range of 24dB at most, and its usually less. That comes from a standard policy derived rom the traditional use of PPMs and the requirement to keep the needle between 1-6. For some stations the dynamic range will be very much less. Peak and average levels are generally pretty consistent, and I find I need to adjust the volume control pretty infrequently, either at home or in the car.

I understand the need to minimise dynamic range in noisy environments or to help when there are hearing impairments, of course, but personally I really enjoy good dynamic music.

H

.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 12:39 am
by hobbyist
Hugh Robjohns wrote:
hobbyist wrote:In my home I compress everything to a max 20dB DR and then normalise the volume level so I never have to keep adjusting the knobs on my stereo.

I'm not familiar with the broadcasting standards in the US, but it sounds a bit grim if you need to take that kind of extreme action.

In the UK, radio and TV broadcasters generally maintain a dynamic range of 24dB at most, and its usually less. That comes from a standard policy derived rom the traditional use of PPMs and the requirement to keep the needle between 1-6. For some stations the dynamic range will be very much less. Peak and average levels are generally pretty consistent, and I find I need to adjust the volume control pretty infrequently, either at home or in the car.

I understand the need to minimise dynamic range in noisy environments or to help when there are hearing impairments, of course, but personally I really enjoy good dynamic music.

H

.

That was for CDs and music I play back.

The audio for TV has reasonble DR. It is the wide variation in loudness that is a problem.

I just mute when they go to commercials and avoid most of the problems. But switching channels often causes major changes in volume. That could be the cable or streaming company doing that not the original source content.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 4:25 am
by Bob Bickerton
hobbyist wrote:
Bob Bickerton wrote:That’s really oversimplifying the situation.

Some people like to hear music at higher volumes, others prefer lower volumes and reducing dynamic range isn’t going to help!

Bob


Lowering DR allows us to have sufficient volume for intelligibility on voice while keeping the high end low enough to be safe.

Which does not address the issue that some like it loud, some do not..........

Bob

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 9:31 am
by Mike Stranks
As Bob says, in a live situation this is not a problem that can be sorted with gizmos.

In a church context in the UK it's often a generational thing and what is deemed 'seemly' in church.

But broadly speaking if you're having to apply bucket-loads of compression or limiting then either your gain-staging is wrong, the musos are playing too loud or they're not very good at playing as a group.

I say again... in the church context this is worship, not a concert.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 6:11 pm
by hobbyist
Bob Bickerton wrote:
hobbyist wrote:
Bob Bickerton wrote:That’s really oversimplifying the situation.

Some people like to hear music at higher volumes, others prefer lower volumes and reducing dynamic range isn’t going to help!

Bob


Lowering DR allows us to have sufficient volume for intelligibility on voice while keeping the high end low enough to be safe.

Which does not address the issue that some like it loud, some do not..........

Bob

It directly addresses it.

The too loud folks are complaining about the loud end.
The too soft folks are complaining about the speech on the low end.

COMPRESS it all UPwards at the low end and normally at the high end.
Then the loudness issue would be easily fixed with slight tweaks of the result if any were still needed.

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 7:48 pm
by AlecSp
hobbyist wrote:It directly addresses it.

The too loud folks are complaining about the loud end.
The too soft folks are complaining about the speech on the low end.

COMPRESS it all UPwards at the low end and normally at the high end.
Then the loudness issue would be easily fixed with slight tweaks of the result if any were still needed.
Of course it will.
Sounds like you're deducing this with no practical experience yet?
Give it a try wwith a real band and a real audience - good luck (you'll need it)

Re: Loudness - more than LUFS/LKFS but measuring SPL other than A or C weighting

PostPosted: Sat Jul 13, 2019 8:00 pm
by Mike Stranks
hobbyist wrote:
It directly addresses it.

The too loud folks are complaining about the loud end.
The too soft folks are complaining about the speech on the low end.

COMPRESS it all UPwards at the low end and normally at the high end.
Then the loudness issue would be easily fixed with slight tweaks of the result if any were still needed.

No! No! No!

Use. Your. Faders.