Are you REALLY listening at Reference Volume??? Maybe Not...

[ddude003]

Be nice if there was software to do this! Dolby LM100 is available for short money.

Are you looking for something like this software https://nugenaudio.com/vislm/ ???

 

[JStewart]

Are you looking for something like this software https://nugenaudio.com/vislm/ ???

Well first I'll need a DAW and a clue about what to do with them. lol

Dolby has some products/api services I also wouldn't know how to use either, but they appear to specifically deal with dialog levels.

Dolby’s Speech-Gating Reference Code

Include Dolby technologies in your loudness measurement products at no cost.

professional.dolby.com

Dolby OptiView Documentation

Developer documentation, guides and APIs for THEOplayer, OptiView Ads, OptiView Ad Engine, Open Video UI, Millicast and THEOlive.

docs.dolby.io

And this older device to measure dialog that can be picked up for not much at all on eBay

Dolby LM100 Broadcast Loudness Meter | Product Catalog | TVC

Dolby LM100 Broadcast Loudness Meter

www.tvcinc.com

 

[Paul Hurt]

Well first I'll need a DAW and a clue about what to do with them. lol

Dolby has some products/api services I also wouldn't know how to use either, but they appear to specifically deal with dialog levels.

Dolby’s Speech-Gating Reference Code

Include Dolby technologies in your loudness measurement products at no cost.

professional.dolby.com

Dolby OptiView Documentation

Developer documentation, guides and APIs for THEOplayer, OptiView Ads, OptiView Ad Engine, Open Video UI, Millicast and THEOlive.

docs.dolby.io

And this older device to measure dialog that can be picked up for not much at all on eBay

Dolby LM100 Broadcast Loudness Meter | Product Catalog | TVC

Dolby LM100 Broadcast Loudness Meter

www.tvcinc.com

I have all the DAW stuff obviously. The issue is not the DAW or plug-ins, the issue is capturing the audio accurately from the disc (or a stream) into a bunch of WAV files. It needs to be done either by ripping the disc and extracting WAVs at least for each base layer channel, or taking it from HDMI. No easy way to do either. I could capture it via analog outputs, but the digital levels on the disc won’t really correlate with the levels that we capture.

I could get close maybe by calibrating input levels with the Spears & Munsil discs…

But it’s a big operation, much harder than measuring SPLs. Even then it’s affected by dialNorm and any errors introduced there.

 

[VJM]

I have all the DAW stuff obviously. The issue is not the DAW or plug-ins, the issue is capturing the audio accurately from the disc (or a stream) into a bunch of WAV files. It needs to be done either by ripping the disc and extracting WAVs at least for each base layer channel, or taking it from HDMI. No easy way to do either. I could capture it via analog outputs, but the digital levels on the disc won’t really correlate with the levels that we capture.

I could get close maybe by calibrating input levels with the Spears & Munsil discs…

But it’s a big operation, much harder than measuring SPLs. Even then it’s affected by dialNorm and any errors introduced there.

This is what they do for BEQ, someone will extract the audio and they will analyze it, that's how they come up with MV offset

 

[VJM]

I came across this, haven't read through everything yet but it looks like Netflix's sound mixing requirements

Netflix Sound Mix Specifications & Best Practices v1.6

Skip to Translations Table of Contents 1 Nearfield Audio Prerequisites for Mix Facilities 2.1 Near Field Atmos® Mix - Mastering and Archive - Original Language Version 2.2 Near Field 5.1 Surround M...

partnerhelp.netflixstudios.com

• This is a nearfield mix. The most common monitoring levels for nearfield are 79db or 82db

 

[ddude003]

You can pick up a external SCSI Blue Ray/DVD reader/writer fairly cheap and there is lots of extraction software that can copy to your hard disk or SSD...

 

[Paul Hurt]

yes… same technique people use to rip discs for Zappiti etc. Can it rip Dolby Digital/Atmos tracks to WAV files tho? I guess BEQ folk are doing it somehow, I just haven’t had time to investigate.

 

[Paul Hurt]

Netflix and everyone tend to use a variation on the broadcast ITU / ATSC / EBU R128 LKFS-target based spec. Thing is, that’s not how movies are mixed… tho it may be how some HE mixes of movies are mixed.

But since i believe the aim is try and duplicate the theatrical experience, it makes sense to me to approach this the way movie theaters approach it. ie having a calibrated reference level and fairly consistent dialog levels measured in SPL. Not that it’s not interesting to see the levels present on the disc.

 

[JStewart]

yes… same technique people use to rip discs for Zappiti etc. Can it rip Dolby Digital/Atmos tracks to WAV files tho? I guess BEQ folk are doing it somehow, I just haven’t had time to investigate.

Their software is freely available on GitHub. Links in Post 3 of this avs thread.

Bass EQ for Filtered Movies

BEQCatalogue Why are so many movies coming out with reduced bass levels lately? We can’t be 100% certain, but the Bass EQ project is here to help. Some of the early instances that made the “nerfing” of the Bass on a home release obvious was when DVD’s were re-released in Bluray format. Some...

www.avsforum.com

Not sure it helps much though. It gets you a file the analysis of which will be far more complicated.
EDIT: Correction, BEQ Designer does not do the ripping.

Would a catalog of presets be nice? Of course. Necessary? Nope!

Give me a SPL phone app and a minute of dialog, I’m good.

 

[Paul Hurt]

FWIW I think I’ve figured out ripping UHD and extracting WAVs. Have never had any need to do it before. Amazing what you can figure out with 10 min on Google.

Let me decide if I want to go quite this deep down the rabbit-hole :-)

I agree JS - when the object is just to find good volume settings for our movie collections, it’s arguable whether there’s a need to go beyond REW and the SPL logger. I mean, it’s relatively easy, I’m getting good results, and AFAIK it’s how large theatre setups are checked.

But a rip of the audio sure would allow me to employ a few different tools…

 

[ddude003]

Oh man, I totally forgot, SOX can read just about any audio format file, do multichannel and provide a very nice status... And its free...

sox -V4 /Users/dd/Desktop/Audio\ Tests/ryanadams441.flac -n stats

Input File : '/Users/dd/Desktop/Audio Tests/ryanadams441.flac'
Channels : 2
Sample Rate : 44100
Precision : 16-bit
Duration : 00:10:49.32 = 28635012 samples = 48699 CDDA sectors
File Size : 60.4M
Bit Rate : 744k
Sample Encoding: 16-bit FLAC
Endian Type : little
Reverse Nibbles: no
Reverse Bits : no
Comments :
ALBUM=2006-10-17 Das Haus - Ludwigshafen, Germany
ALBUM DYNAMIC RANGE (DR)=12
ALBUM DYNAMIC RANGE (R128)=17.492994
ARTIST=Ryan Adams & The Cardinals
COMMENT=Soundboard > Edirol R-09
DATE=2006
DYNAMIC RANGE (DR)=12
DYNAMIC RANGE (R128)=17.037037
GENRE=Rock
REPLAYGAIN_ALBUM_GAIN=-2.04dB
REPLAYGAIN_ALBUM_PEAK=0.977508
REPLAYGAIN_TRACK_GAIN=-2.94dB
REPLAYGAIN_TRACK_PEAK=0.977508
TITLE=Magnolia Mountain
TRACKNUMBER=2

Output File : '' (null)
Channels : 2
Sample Rate : 44100
Precision : 16-bit
Duration : 00:10:49.32 = 28635012 samples = 48699 CDDA sectors

/usr/local/bin/sox INFO sox: effects chain: input 44100Hz 2 channels (multi) 16 bits 00:10:49.32
/usr/local/bin/sox INFO sox: effects chain: stats 44100Hz 2 channels 16 bits 00:10:49.32
/usr/local/bin/sox INFO sox: effects chain: output 44100Hz 2 channels (multi) 16 bits 00:10:49.32
/usr/local/bin/sox DBUG sox: start-up time = 0.001017
Overall Left Right
DC offset -0.000003 -0.000003 -0.000002
Min level -0.976959 -0.960907 -0.976959
Max level 0.977509 0.881134 0.977509
Pk lev dB -0.20 -0.35 -0.20
RMS lev dB -17.72 -18.13 -17.35
RMS Pk dB -7.53 -8.05 -7.53
RMS Tr dB -59.37 -59.37 -50.78
Crest factor - 7.75 7.21
Flat factor 0.00 0.00 0.00
Pk count 2 2 2
Bit-depth 16/16 16/16 16/16
Num samples 28.6M
Length s 649.320
Scale max 1.000000
Window s 0.050

Yuck, the column formatting got munched...

 

[ddude003]

Input File : '/Users/dd/Desktop/multichan/AAC5.1.flac'
Channels : 6
Sample Rate : 44100
Precision : 24-bit
Duration : 00:00:46.63 = 2056192 samples = 3496.93 CDDA sectors
File Size : 9.05M
Bit Rate : 1.55M
Sample Encoding: 24-bit FLAC
Endian Type : little
Reverse Nibbles: no
Reverse Bits : no

Output File : '' (null)
Channels : 6
Sample Rate : 44100
Precision : 24-bit
Duration : 00:00:46.63 = 2056192 samples = 3496.93 CDDA sectors

sox INFO sox: effects chain: input 44100Hz 6 channels
sox INFO sox: effects chain: stats 44100Hz 6 channels
sox INFO sox: effects chain: output 44100Hz 6 channels
Overall Ch1 Ch2 Ch3 Ch4 Ch5 Ch6
DC offset -0.000059 -0.000059 -0.000044 -0.000037 0.000021 -0.000045 -0.000036
Min level -0.764507 -0.740280 -0.636481 -0.764507 -0.132807 -0.664265 -0.712370
Max level 0.516816 0.516816 0.411199 0.506089 0.131884 0.463755 0.491864
Pk lev dB -2.33 -2.61 -3.92 -2.33 -17.54 -3.55 -2.95
RMS lev dB -26.41 -25.90 -25.27 -27.32 -26.13 -26.84 -27.42
RMS Pk dB -13.47 -13.87 -14.60 -13.77 -20.87 -14.45 -13.47
RMS Tr dB -133.03 -114.45 -113.90 -114.18 -133.03 -113.83 -113.85
Crest factor - 14.60 11.68 17.76 2.69 14.60 16.73
Flat factor 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Pk count 2 2 2 2 2 2 2
Bit-depth 24/24 24/24 24/24 24/24 22/24 24/24 24/24
Num samples 2.06M
Length s 46.626
Scale max 1.000000
Window s 0.050

SOX did not like the original mp4 container... So I put it in a FLAC container that SOX liked...

 

[Aburguy]

On a recent @Bright Side HT podcast, show guest Paul Hurt, a sound technician that mixes both music and TV material, joined host DJ Briggs with an interesting spin on reference volume levels.

Just a bit of background, but the reference levels we set our processors and AVRs to are typically either 75dB or 85dB, depending on whether you're following THX or Dolby standards. Those numbers are derived from being either -30dB or -20dB from a max SPL of 105dB, which gives sound engineers a wide amount of dynamic range to master everything from wispy wind and dialog to window-rattling explosions.

The idea is to calibrate your system so that your speakers are channel leveled to either 75 or 85dB at the primary listening positing in a room. Thus, when you turn your volume knob to "0" on a relative volume scale, you're operating at reference level, meaning that you're hearing the content - from the whisper quiet to the loud - at a loudness level that a mixer intends you to experience.

Short end of the story is that Paul says reserach in psychoacoustics in small rooms says that a loud sound (say, 75dB) sounds louder in a small room than a large room (like a theater). Thus, mixers have found that reference minus 6dB is the best level for reference mixing. Thus, if you want a reference experience, you should be gunning for -6dB on your receiver or processor.

DJ can chime in here. Hopefully we can Paul to join the thread and talk shop about this.

In the mean time, curious to hear thoughts!

 

[Marc Lombardi]

Forgive me if I missed this if it was previously discussed in this thread ... but what about Fletcher-Munson? If someone mixes at 75db or 85db reference, then the aspects of the perception curve will be baked into what they heard, right? And if we listen at a different reference then we may or may not need more or less compensation ... and this affects our overall perception of how loud it all is ... am I close on this idea?

This also speaks to the propensity for people to use house curves, boosting the bass ...

 

[VJM]

Forgive me if I missed this if it was previously discussed in this thread ... but what about Fletcher-Munson? If someone mixes at 75db or 85db reference, then the aspects of the perception curve will be baked into what they heard, right? And if we listen at a different reference then we may or may not need more or less compensation ... and this affects our overall perception of how loud it all is ... am I close on this idea?

This also speaks to the propensity for people to use house curves, boosting the bass ...

Hi Marc, I don't believe that's been discussed but I think you're on the right track, if listening at reference or -6 for a small room then Fletcher-Munson wouldn't come into play. That would come into play when listening at a lower level, for example, -20, for these people the house curve would help with the loss of bass perception

 

[VJM]

I'd like to add, I've been using the BEQCatalogue database for my reference offset and it's perfect.

A few that I watched and what volume level (with my room size I use -7 as my reference):

Tropic Thunder - DTS-HD MA - no offset -7
Open Range - DTS-HD MA - offset +2, my listening level -5
Total Recall - Atmos - offset +2.5, my listening level -4.5
The Martian - Atmos - offset +3.5, my listening level -3.5
Twister - TrueHD 5.1 - offset -2.5 my listening level -9.5

Anyone else can try this and see what you think with these offsets.

 

[Paul Hurt]

Hi again VJ!

I think you nailed it.

We did touch on Fletcher-Munson in the podcast… mainly in relation to Audyssey.

The difference between the equal loudness curves at listening levels of 79dBC and 85dBC is very small, so I don’t think F-M effects are going to affect how movies and home entertainment mixes come out.

Can’t say I’ve ever felt any need to do that bass boost at low volumes. I guess I’m rarely watching movies at low volume… because of the dynamic range, there’s not a big range of playback volumes that even work (because you don’t have to turn down very far from reference before the dialog becomes difficult to hear). But maybe if I was using Dolby DRC or some other dynamic range compression.

Re house curves… I probably should look into this a little more, but the use of a house curve in a mixing environment designed for doing HE mixes doesn’t make much sense to me - since the aim is to simulate the environment that the product will be viewed in, and not many people implement any kind of eq curve for their system at home… unless they’ve run Audyssey and left it at its default with the gentle HF attenuation.

But maybe there’s a house curve they aim for to try and get their mixes to translate between different systems within a post facility. That doesn’t mean tho that you need to apply a “house curve” at home.

Dolby do specify a target curve that they would like your monitoring system to hit when mixing music in Atmos tho.

The idea of house curves is quite a deep topic, but not something most of us need to be concerned about. “Flat” is fine IMO. Was good enough for the greatest hifi systems ever built, don’t see why it wouldn’t be good enough for HT.

 

[Todd Anderson]

I like starting a movie knowing I’m less likely to have to keep tweaking the volume throughout the film. Means I can just settle back and watch it.

I like this comment. It's especially true for folks like us... I find it hard enough not to pay attention to technical side of things!

 

[VJM]

A couple more that I watched and what volume level (with my room size I use -7 as my reference):

Superman II: The Richard Donner Cut - Atmos - No Offset, my listening level -7
Harry Potter and the Deathly Hallows: Part 1 - DTS:X - Offset +2.5, my listening level -4.5

 

[NBPK402]

I have found that listening in a 20k cf room @ say 85db atthe mlp sounds quieter than listening at the same spl in say a 5k cf room.. I always thought it was because in a smaller room all the reflections might be making it less clear, and that that would make it sound louderl

 

[ddude003]

Because of the law of inverse square spl drops as the distance increases between speakers and mlp... For each doubling of distance from a point source, the sound pressure level decreases by approximately 6 dB... I am sure there is another formula for room volume that would compound the complexity, as well as absorption and reflections... Bad room... Lol...

 

[Paul Hurt]

Of course, inverse square law only holds true in a free-field eg outdoors. Once you’re running a speaker in a room, the way sound waves are contained / reflected means you don’t get that expected drop in SPL as you increase distance from the speaker.

But the thing about reference level is you’re measuring the SPL at the listening position, so if you calibrate to the same reference SPL, a big room or small room shouldn’t influence how loud things sound. Yet it does. Which is weird.

One of my theories is that it might be to do with the arrival time of the first reflections. In a big room those first reflections arrive many milliseconds after the direct sound… in a small room they arrive at your ears very quickly. I wonder if your brain somehow interprets this as louder, even tho the SPL is the same.

[which I’ve just realised is exactly what NBPK was saying!…]