TBProAudio releases AB_LM - Loudness Match and Gain Staging Plugin for Windows and Mac OS X

[TB-ProAudio]

This plugin is basically used in mastering.. check out perception.. i definitely wouldn't be using it for gain staging the mix as such.. this plugin is about , um, level *perception* to make sure one doesn't make effects decisions on volume increase placebo.

Right! Not more not less.

I'm a bit concerned about you saying the RMS meter is wrong but that can be ignored i guess and one can use any other RMS meter in conjunction with the plugin. Is the ebu 128 on TB's plugin working according to specs? that's the most important one for mastering

AB_LM EBU measurement is working according EBU specs (ML/SL), that was never the question.
Even RMS measurement is correct, but is is summed RMS! Just compare EBU ML with RMS (400ms window). The small deviation comes from EBU pre filtering.

[Compyfox]

Right! Not more not less.

Wrong!

If the feature is there (and advertised as "gain staging", which is the most associated with per-channel gain staging - another Austrian company does even emphasize on that with their tools), people will use it. And people will(!) abuse it. If it's not according to specs, then mixes will be off. Those that don't know this, assume it's correct and continue without further thinking.

This(!) is what is annoying me the most. On top of the ignorance from BOTH users (no offence, Theo) and the developer himself (I still want to know who "we and others" are... again, I've read many posts on other forums regarding metering tools where my toenails basically folded themselves back up).

I'm a bit concerned about you saying the RMS meter is wrong but that can be ignored i guess and one can use any other RMS meter in conjunction with the plugin. Is the ebu 128 on TB's plugin working according to specs? that's the most important one for mastering

AB_LM EBU measurement is working according EBU specs (ML/SL), that was never the question.
Even RMS measurement is correct, but is summed RMS! Just compare EBU ML with RMS (400ms window). The small deviation comes from EBU pre filtering.

Once more - it's ITU-R BS.1770-x for once, especially if we talk variable reference level. EBU R-128 is a "preset" of this (fixed reference level, fixed gate values, fixed color codes). But R-128 is easier to remember and it's all the rage as of late. I get that.

And no, your RMS measurement is NOT correct, it's 3dB off due to the summing.

There is NOTHING to compare EBU R-128 ML (400ms, weighting filter) with RMS (raw loudness, 400ms, no weighting filter). It's like comparing apples with oranges.

If you insist on using RMS in the form you currently offer, which is NOT according to 40+ year old specs and mathematics, then remove the RMS+3 preset. Due to your summing matrix (again, not according to specs up until AES-17 r2015), a -18dBFS 1kHz reference signal doesn't resolve in -18dBFS but -15dBFS in RMS+3 mode. In RMS+0 mode it should be -18dBFS => -21dBFS RMS, but in your case it is -18dBFS due to the summing. This can be easily recreated in dpMeter.

So drop this one preset and write it ACCORDINGLY in both your manual (if present) and on your page... then things should be fine. The AES-17 "compensation" is the most used setting for RMS meters these days anyway.


It is still not according to specs however! (again: the summing of the L and R signal)
And I do insist on that matter

[TB-ProAudio]

It is still not according to specs however! (again: the summing of the L and R signal)
And I do insist on that matter

No one said that TBProAudio RMS calculation (summed or not) is according to any specification! Not me! Please be precise, thank you.

[Lesha]

It is still not according to specs however! (again: the summing of the L and R signal)
And I do insist on that matter

No one said that TBProAudio RMS calculation (summed or not) is according to any specification! Not me! Please be precise, thank you.

An option for summed / not summed signal would be a welcomed addition, including dpmeter.

[TB-ProAudio]

An option for summed / not summed signal would be a welcomed addition, including dpmeter.

Hmm, what is the use of 2 different RMS levels? 2 different gain stages?
As is said before, i´m not a fan of auto gain-staging during production phase, just manual gain-staging. In this case both channels should be treated together, right?

[Lesha]

As I have stated earlier regarding dpmeter, I would like to have a proper RMS readout. If you don't want to implement it to AB_LM that is your choice, but I would like dpmeter to have the option.

[TB-ProAudio]

Sorry, but i still do not get it.
You mean with "proper": RMS per channel, right?

[Lesha]

Sorry, but i still do not get it.
You mean with "proper": RMS per channel, right?

3 dB quieter, ie. not summed.

[Compyfox]

It is still not according to specs however! (again: the summing of the L and R signal)
And I do insist on that matter

No one said that TBProAudio RMS calculation (summed or not) is according to any specification! Not me! Please be precise, thank you.

If you use the tag "RMS" for a meter, it implies certain specs, just like if you tag a meter readout MLk (400ms integration, k-weighted) or SLk (3s integration, k-weighted).

A basic RMS Meter is 300ms time window (about 95% of the time, except for Dorrough's, which is 600ms according to patents or certain desired applications), unweighted (sometimes called z-weighted), individual/per channel readout. RMS meters are also called "Raw Loudness Meters" (Steinberg does that for example). The AES-17 papers up until revision 2015 state that a 1kHz sine test signal with -18dBFS should read -18dBFS, which also takes care of the r2004 re-addressed (originally addressed in 1995 IIRC) +3dB offset.


Now here is the problem:
If you tag your meter RMS+0 and RMS+3, people assume that 1kHz Sine at -xyz value read out specific values. RMS+0 with -18dBFS = -21dBFS, RMS+3 with -18dBFS = -18dBFS. And here it doesn't matter if it shows on the L or the R channel.

You however "sum" the channels into one, therefore adding +3dB during summing. So RMS+0 is actually +3dB off, resulting in RMS+3, and RMS+3 is also 3dB off, technically resulting in RMS+6. You created your "own" standard - like so many others before you.

An option for summed / not summed signal would be a welcomed addition, including dpmeter.

Hmm, what is the use of 2 different RMS levels? 2 different gain stages?
As is said before, i´m not a fan of auto gain-staging during production phase, just manual gain-staging. In this case both channels should be treated together, right?

Again, you need to clear the RMS+0 and RMS+3 thing first. THEN we talk about "per channel" Gain Staging according to RMS values (which will happen if the feature is there, people will use the plugin for something else as well).

Gain Staging per channel or with a summed mix usually always(!) goes by the max RMS peak of either channel.

Example:
- a Kick Drum in Mono or Stereo peaks up to -16dB RMS, you want to gain stage it to -18dBFS, you lower it by 2 dB

- a choir spread through the stereo field has more impact on the right side, so L reads -20dB RMS and R ready -17dB RMS. Therefore the loudest section is -17dB RMS, you lower the signal by -1dB to get to -18dB

- a stereo mix down has a healthy RMS level of -14dB RMS, but the left side is 1 dB stronger due to drums and cymbals than the right side. So L = -14dB RMS, R = -15dB RMS. You want to gain stage all tracks of the album to about -16dBFS in order to use proper EQ, compression and the loudness raising up to -12dB RMS avg... so you adjust the whole signal down by -2dB


Works the same way for PPMs, digital meters, VU's, the K-System meter, etc.
And actually, there are a lot of tools on the market by now with "auto gain" settings that work the same way. One VU that comes to mind, is Hornet Plugin's VU meter.


Granted, it is simpler with ITU-R BS.1770-x meters due to the summed signal, and the developer or rather the software therefore only needs one value to check (also considering integration time and weighting filter). But the principle is the same... unless you declare an average between lowest SLk and highest SLk value, you always go by the highest peak of the meter. Case in point => -14LUFS SLk max, but you want to level it down to -16LUFS before editing further, reduce the signal by 2dB.

I don't see where this is so hard to understand?



Solution (simplest form):
Drop the RMS+3 preset if you INSIST on summing the signal for easier readouts, as RMS+0 with the summed signal is basically the AES-17/RMS+3 setting. Then write in your manual (if there is a manual!) AND on your page:

"The RMS meter sums the signal and analyzes it according to a time frame of 300ms (regular RMS meter, variable setup depending on the plugin). This is equal to the AES-17 +3dB compensation. This feature is to simplify the code of the plugin"


If you do that, and make people DEAD AWARE of this... then I don't see a problem. But if your users don't know/understand the concept... then things go awry really fast. At some point, some user will point out that "all other tools are off" or some crazy article writer is like "this is how it works, you should do so as well!". Which in turn will spread like a wildfire, "hearsay" will turn into "standards". Even if they're wrong.


And then people are surprised why certain Audio Engineers are so darn pedantic if it comes to proper labeling of signal strengths, or being very touchy on accuracy of metering tools in general.

[ObsoleteAcc99]

One thing I am not sure of, is why is TB not implementing the same RMS standard as the other pro meters? Why argue about it if it's wrong?

Ian shepherd has said the waves WLM one is spot on, so will check that and compare it to TB's and see what i get.

He said logic's and a couple others are way out.

[Lesha]

One thing I am not sure of, is why is TB not implementing the same RMS standard as the other pro meters? Why argue about it if it's wrong?

Ian shepherd has said the waves WLM one is spot on, so will check that and compare it to TB's and see what i get.

He said logic's and a couple others are way out.

I already did it here: http://www.kvraudio.com/forum/viewtopic ... 7#p6274777

[ObsoleteAcc99]

One thing I am not sure of, is why is TB not implementing the same RMS standard as the other pro meters? Why argue about it if it's wrong?

Ian shepherd has said the waves WLM one is spot on, so will check that and compare it to TB's and see what i get.

He said logic's and a couple others are way out.

I already did it here: http://www.kvraudio.com/forum/viewtopic ... 7#p6274777

no not waves you didn't.. I said i was going to compare it with *waves* as Ian Shepherd endorses that one, and you said you already did it, with a link, and i click, and you haven't.

[Lesha]

I don't use Waves, but I compared it with several others.

[Compyfox]

Theo, Waves WLM is a ITU-R BS.1770-x type meter.

I do talk about the plain, old, boring, digital RMS meter... Apples and Oranges

[ObsoleteAcc99]

Theo, Waves WLM is a ITU-R BS.1770-x type meter.

I do talk about the plain, old, boring, digital RMS meter... Apples and Oranges

who cares? it matters only for his free meter.. people will use a proper meter in conjunction with this.. this is for the master bus or a track here or there to compare volume matched effect processing. You are going way OTT.

yet on his free plugin, lesha showed above that one of the settings matches the other, correct plugins. So what' the problem?