Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[DarkSpace]

Greetings,
it is time for me to, once again, ask a question which will probably be easily answered:

As I understand (from its thread), the goal of Loudness is to make Audio Volume changes based on how human hearing works rather than how signal works.
Further, the way I understand it, R128 Volume Leveling is supposed to achieve a Reference Volume by automatically adjusting the playback volume, basically like having it adjust the Volume Slider for you. I may confuse some things here with what ReplayGain aimed to do, though.
So, if both of those understandings of mine are correct, I think it might make sense to apply Loudness, if it is selected, to Volume Leveling as well as to the Volume Slider.

I realize that this is not a full reasoning (examples: Does Loudness work when increasing the Volume? I've seen a few - very few - cases where R128 required an increase in Volume; How does it correspond to Reference Volume settings, that make Loudness work only when going below a certain threshold?; How does that work in combination with Adaptive Volume?), but I believe at at least the last two examples could be solved by placing a checkbox somewhere in the DSP's settings that specifies whether to use Loudness when adjusting Volume or not.

Also, I couldn't find anything about it (maybe I searched for the wrong keywords?), but it makes sense that someone else already had this idea*, and that, if that is so, it was rejected (since, you know, it's not implemented). Can someone please explain why, or give me a link to the corresponding thread?

* I'm not that special.

[mwillems]

Greetings,
it is time for me to, once again, ask a question which will probably be easily answered:

As I understand (from its thread), the goal of Loudness is to make Audio Volume changes based on how human hearing works rather than how signal works.
Further, the way I understand it, R128 Volume Leveling is supposed to achieve a Reference Volume by automatically adjusting the playback volume, basically like having it adjust the Volume Slider for you. I may confuse some things here with what ReplayGain aimed to do, though.
So, if both of those understandings of mine are correct, I think it might make sense to apply Loudness, if it is selected, to Volume Leveling as well as to the Volume Slider.

I think you might've misunderstood how "loudness" works.  Compensation applied by loudness is designed to adjust for how human hearing reacts to frequency response at different volumes, based on internationally standardized charts (that are refined versions of the old fletcher-munson curves).  See https://en.wikipedia.org/wiki/Equal-loudness_contour.  

The key point to understand is that the equal loudness contours were measured from a known real-world starting volume, and they don't scale linearly downward.  That is to say that humans (on average) hear 80dB sound differently than 60dB sound, but also the difference between how we hear 80dB and 60dB sound is not the same as the difference between how we hear 60dB sound and 40dB sound. So absolute volume level matters.

For those reasons the loudness correction only works optimally when starting from a known volume.  That's why it's recommended that you calibrate your system volume to a reference level (a -20dBFS pink noise signal measuring 83dB) in order to get the most out of Loudness correction. http://wiki.jriver.com/index.php/Volume#Reference_Level_Calibration

But calibrating your system is only half the battle because music is mastered to widely varying average volume levels.  Volume levelling assists in making this work because it "re-centers" the volume of music so that it aligns with that 83dB reference level.   http://wiki.jriver.com/index.php/Volume_Leveling

So to summarize, in order for loudness to work correctly, your system needs to be calibrated to a certain actual volume level, and your program material needs to have its volume recentered to that same volume level.  So the system is already designed for volume leveling and loudness correction to work perfectly together.  Applying loudness as part of volume leveling would effectively result in applying the loudness effect twice (which would be too much), and would apply vastly different loudness corrections to different music regardless of the actual real world volume (which wouldn't help).

[blgentry]

Very, very well said mwillems. 

Sound playback is complex!

Brian.

[DarkSpace]

Well, that was certainly educational (and quick), thank you.

The short description (I'm going to read the links in detail this evening) reminds me of the differences between perceptual and linear color spaces in video. I should have considered a possibility like that.