Loudness Equalization via software

[XMonarchY]

I enable NVidia Audio Loudness Equalization for my HDTV connected to my rig with HDMI. It works well.

If I don't use that feature, voices tend to be overly quiet in films. I think it's because my TV is old (2009!) with very limited dynamic audio range. I do NOT select "Full Range" when I set up my HDTV NVidia Audio device, but that's not enough and neither is setting Digital Range Compression (DRC) to 100% in LAV Audio settings. To get good audio and be able to easily distinguish voices, I have to have all 3:
- Full Range UN-selected
- DRC set to 100% in LAV Audio Filter settings
- Loudness Equalization enabled

LatencyMon reports increased NVidia driver latency whenever I enable any NVidia Audio Enhancement, including Loudness Equalization.

I want to reduce my NVidia Audio driver latency and for that I need to keep all NVidia enhancements disabled. That, again, will mess up voices in films. How can I mitigate that issue? Are there some other LAV Audio parameters to check out? I only use Stereo, no up/down-conversion to/from 5.1/7.1. In LAV Audio, I also select (as per some guru's recommendation) "Enable Mixing", "Output Speaker Config = Stereo", enabled "Don't Mix Stereo Sources", and enabled "Clipping Protection".

[huhn]

http://www.ac3filter.net/wiki/Download_AC3Filter

LatencyMon reports increased NVidia driver latency whenever I enable any NVidia Audio Enhancement, including Loudness Equalization.
when activating or continuously?

the range of speaker is only about bass and nothing else it pretty much asking if your speaker can create bass on it own (full range) or not.

[smok3]

Use a player with imilar effect, which can hopefully keep the perfect sync (mpv).

[hello_hello]

My method requires ffdshow and I've no idea if it increases latency. PotPlayer might be able to load WinAmp plugins on it's own. Thinking about it, I'm pretty sure it can, and probably has some not completely crappy compression built in, but I use MPC-HC and LAV only for decoding. ffdshow downmixes and compresses the decoded audio. I have AC3's DRC disabled as RockSteady compresses everything. Well.... it doesn't "compress" as such. It increases the volume of the quiet bits rather than reduce the volume of the loud parts, which is a better way to do it (much like the Dynamic Audio Normaliser (https://www.videohelp.com/software/Dynamic-Audio-Normalizer)).

RockSteady only supports stereo (or mono) and 16 bit.

I've settled on the settings below. It compresses enough without resulting in horrible volume pumping, like AC3's DRC or the "night mode" in most TV audio setups. It pays to reduce the volume with ffdshow's volume filter by around 6dB (or use a normalised matrix for downmixing) and don't include the LFE channel because it increases the potential for volume pumping. I increase the volume of the centre channel a little when downmixing, but if I'm honest, after compression that's probably just a placebo.

RockSteady (https://winampheritage.com/plugin/rocksteady-2-1/1099)

https://i.postimg.cc/jd9XXpD7/ffdshow.gif

https://i.postimg.cc/PrNy1X5H/Rock-Steady.gif