Old 09-17-2018, 05:40 AM   #321
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

Quote:
Originally Posted by heda View Post
The problem I see is Max TP is the maximum found. Let's say you have a 0.0 Max at the beginning a click or something... then all the measurements of PSR will be affected after that. For dynamic range the MaxTP should be more like SL. something like a Max TP that autoresets each 3 seconds maybe... I don't know. Or maybe an average of MaxTPs for 3 seconds minus the SL.

I underline this. At the end it is a matter of definition and currently it is not 100% clear to me.
I have not yet found any definition of Peak in the PSR idea:


1) maximum peak from start to the point where the SL is captured or

2) maximum peak within a sliding window of e.g. 3sec or

3) maximum peak reset every 3sec or

4) maximum peak with a decay of 3sec or
5) average of maximum peak within a certain window or

6) whatever :-)



As the PSR term is fairly new and not yet defined in any official specification I tend to define it so it makes sense :-)


So will spend some brain on it. And I like nr. 5 from my list. Because it combines the old and still good VU meter concept with the modern loudness concept...
__________________
www.tb-software.com/TBProAudio
TBProAudio is offline   Reply With Quote
Old 09-17-2018, 05:44 AM   #322
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

Quote:
Originally Posted by doppelganger View Post
my 2 cents:
i think it's: max (true) peak in range of 3s - Short-Term in the same range, calculation is on every sample.

Update rate for meter the same as for Short-term = not less than 10 Hz.

Thank you.So, nr. 2 from my list.
Any references?
__________________
www.tb-software.com/TBProAudio
TBProAudio is offline   Reply With Quote
Old 09-17-2018, 05:54 AM   #323
doppelganger
Human being with feelings
 
Join Date: Feb 2017
Posts: 371
Default

Quote:
Originally Posted by TBProAudio View Post
Thank you.So, nr. 2 from my list.
Any references?
nope=), just summed all the info i read, plus specifications: https://tech.ebu.ch/docs/tech/tech3341.pdf.

I just think, that if max peak for PLR is taken from the whole range, then for PSR max peak should be taken from the short-term range, but that is just my thought. Also ones say max peak, other say true max peak, which one to choose, nobody knows=)
doppelganger is offline   Reply With Quote
Old 09-17-2018, 05:58 AM   #324
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

@doppelganger
I think you are right !


I asked Youlean some infos about this. His plugin is certified 100% "ITU-R BS.1770" compliant We'll see if he can share his expertise with us
X-Raym is offline   Reply With Quote
Old 09-17-2018, 06:01 AM   #325
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

Quote:
Originally Posted by X-Raym View Post
I asked Youlean some infos about this. His plugin is certified 100% "ITU-R BS.1770" compliant We'll see if he can share his expertise with us

Hmm, ITU-R BS.1770 has not yet defined PLR or even PSR value. At least to my knowledge
__________________
www.tb-software.com/TBProAudio
TBProAudio is offline   Reply With Quote
Old 09-17-2018, 06:04 AM   #326
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

@TBProAudio
It is right, but it is an indicator of his expertise :P


I'll tell you as soon as I have more infos.
X-Raym is offline   Reply With Quote
Old 09-17-2018, 06:19 AM   #327
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

Quote:
Originally Posted by X-Raym View Post
@TBProAudio
It is right, but it is an indicator of his expertise :P


I'll tell you as soon as I have more infos.

Right. I just want to make sure that the definition is clear and makes sense.

Otherwise i implement something different...
__________________
www.tb-software.com/TBProAudio
TBProAudio is offline   Reply With Quote
Old 09-17-2018, 06:34 AM   #328
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

OK, just to concluded on this:

Ian Shepherd:

"Micro-dynamics are measured using the Peak to Short-term loudness Ratio (PSR). PSR is similar to PLR, but covers shorter time periods. The measurement uses a 3 second loudness window and uses a realtime peak measurement."

ref: https://www.meterplugs.com/blog/2016...nt-matter.html

There is also a AES paper available.

In short: for PSR the Peak value seems to be defined as Max (True) Peak within a sliding 3s window.
__________________
www.tb-software.com/TBProAudio
TBProAudio is offline   Reply With Quote
Old 09-17-2018, 11:38 AM   #329
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

Quote:
In short: for PSR the Peak value seems to be defined as Max (True) Peak within a sliding 3s window.

But is this window calculated at fixed timecode or is from play start position?
Cayse in this case, the curve may be different if you start playing song from timecode 0 and then from timecode 2. (maybe not very different, but it can be different).


I guess it is from play position per default.
X-Raym is offline   Reply With Quote
Old 09-17-2018, 11:45 AM   #330
TBProAudio
Human being with feelings
 
TBProAudio's Avatar
 
Join Date: May 2014
Location: Germany
Posts: 497
Default

Quote:
Originally Posted by X-Raym View Post
But is this window calculated at fixed timecode or is from play start position?
Cayse in this case, the curve may be different if you start playing song from timecode 0 and then from timecode 2. (maybe not very different, but it can be different).


I guess it is from play position per default.

I do not really understand what you mean with time code, but sliding window means: from current position 3secs back


At smp = 3000: max peak[pos 0; pos 2999]
At smp = 3001: max peak[pos 1; pos 3000]
...
Of course as long sample 3000 is not reached you start always from 0.
__________________
www.tb-software.com/TBProAudio

Last edited by TBProAudio; 09-17-2018 at 10:25 PM.
TBProAudio is offline   Reply With Quote
Old 09-18-2018, 07:57 PM   #331
lachinhan
Human being with feelings
 
lachinhan's Avatar
 
Join Date: Nov 2014
Location: Can Tho - Viet Nam
Posts: 127
Default Loudness Penalty: ANALYZER

Crest Factor, PSR and PLR


Peak to Loudness Ratio


I need it for LOUDNESS PENALTY: ANALYZER

If you can add features Loudness Penalty: ANALYZER, it's great
__________________
NK Recording Studio
Email: lachinhan@gmail.com or admin@thuamninhkieu.com
Website:nkpro.top
lachinhan is offline   Reply With Quote
Old 09-23-2018, 05:57 PM   #332
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

Youlean confirmed me was PSR was Max True Peak during the short term loudbess window (3s) - Short term loudness val, just like Ian Shepered quote brought by TBProAudio.
X-Raym is offline   Reply With Quote
Old 12-14-2018, 08:58 AM   #333
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

Hi !


When I use Loudness Grpah and many tracks (like 10), I got huge CPU usage like 60% and audio troubles.
This also happens on more acient projects (for which Loudness Graph was already set and so which wasn't supposed to have this huge CPU usage and audio problems).


Has anyone experienced such performance issue with Loudness Graph lately ? I don't remember it has such limitations.
X-Raym is offline   Reply With Quote
Old 12-15-2018, 03:08 AM   #334
heda
Human being with feelings
 
heda's Avatar
 
Join Date: Jun 2012
Location: Spain
Posts: 5,351
Default

Hi X-raym. How are you doing man?! I have not updated Loudness Graph(I should)in long time so I dont know why would it be different.
Check that the hidden tracks are not duplicated or something. Each track has one dpmeter(which is the element that uses more cpu) but 10 dpmeters shouldn't use 50% cpu either.
heda is offline   Reply With Quote
Old 12-15-2018, 03:56 AM   #335
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

@heda
I'm of, thx ! I hope you are well too :P I'm finally taking a bit of time to work on creative projects ! And that's where this script come handy !



You are right the problem isn't with the script, it's dpmeter which is CPU intensive. It only take 0.77% in loudness mod of my CPU but if I monitor 15 tracks,, it means 11%! just for the loudness monitoring. Add to that various other FX like convolution reverb and you get the problem I had. (i starts crackling at 25-30%, even with buffer set to maximum)

I tried with dpMeter 3 and it is the same than the 2 version.
Not sure why I have problem now and not before but clearly


I tested other Loudness measurement and Youlean goes to 0.2 CPU and MLoudness to 0.3. But they d'ont provide the automation/parameter output needed there
X-Raym is offline   Reply With Quote
Old 05-07-2019, 01:59 PM   #336
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

Hi Heda,


I found what was causing the audio dropouts with 10+ analysed tracks:



it seems that it is the audio_statitics JS. Indeed, if they are activated I get dropout, but if they are not, I don't have dropouts. They don't create huge CPU increase on the performance meter but somehow having too many of them isn't good for audio.


A quick fix: having a button to not analyse RMS (and so, deactivating audio_statistics plugins which is used only for that purpose) should fix my performance issue. I don't use RMS meters, only Loundess.



I know you haven't updated Loudness Graph for a while but really, it is a great script.


Thanks !
X-Raym is offline   Reply With Quote
Old 05-08-2019, 06:07 AM   #337
heda
Human being with feelings
 
heda's Avatar
 
Join Date: Jun 2012
Location: Spain
Posts: 5,351
Default

very strange, since audio_statistics uses less CPU.
I know... very old script. But I have good news. I have been working on rewriting it (still half way) as a module for HeDaMixer instead. So Loudness Graph V3 will be HeDaMixer V3
heda is offline   Reply With Quote
Old 05-08-2019, 06:12 AM   #338
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

Quote:
very strange, since audio_statistics uses less CPU.
Yep I don't know why either but it is the behavior I got, I can make a video if you need ^^


But as you have other plan, yes maybe odn't bother pushing a fix (especially hard to get back to old code).



I really wonder what it would look like in HeDaMixer 3... I'll keep an eye on it !
X-Raym is offline   Reply With Quote
Old 05-19-2019, 05:37 PM   #339
X-Raym
Human being with feelings
 
X-Raym's Avatar
 
Join Date: Apr 2013
Location: France
Posts: 5,960
Default

For those who want to use it with dpMeter4,
create a Preset with EBU, Sync On, Output On and Ref level on, name it



Code:
LUFS EBU R128 Sync Autom

And set the following code in



Code:
"C:\Users\Anymord\AppData\Roaming\REAPER\Scripts\HeDaScripts\Loudness Graph VIP\Global Options.ini"

Code:
-- settings for dpMeter    4
    dpMeter_to_use = "dpMeter4 (TBProAudio) (6ch)"
    dpmeter_preset = "LUFS EBU R128 Sync Autom"
    dpmeter_reset_on_cursor_change = 0
    dpmeter_min = -144 
    dpmeter_max = 24
    dpmeter_minIL = -70
    dpmeter_lramax = 144
        dpmeter_paramIL = 24 
        dpmeter_paramML = 25
        dpmeter_paramMLMax = 26
        dpmeter_paramSL = 27
        dpmeter_paramSLMax = 28
        dpmeter_paramTPMax = 29
        dpmeter_paramLRA = 30
X-Raym is offline   Reply With Quote
Old 08-18-2019, 03:23 AM   #340
alkamen
Human being with feelings
 
Join Date: Nov 2017
Posts: 6
Default

I just got VIP. But, unfortunately, LUFS measurement does not work in Reaper 5.981 (( I see RMS, but I do not see LUFS changes...
alkamen is offline   Reply With Quote
Old 08-19-2019, 12:24 AM   #341
heda
Human being with feelings
 
heda's Avatar
 
Join Date: Jun 2012
Location: Spain
Posts: 5,351
Default

Quote:
Originally Posted by alkamen View Post
I just got VIP. But, unfortunately, LUFS measurement does not work in Reaper 5.981 (( I see RMS, but I do not see LUFS changes...
Which dpMeter version do you have installed?
I am updating this soon as a module for HeDaMixer VIP which supports the latest version of dpMeter. Stay tuned.
heda is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 03:10 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.