How to adjust audio dialog normalization in Sonic Soft Encode ?

[TeNSoR]

Hello!

I'd like to ask about how to adjust dialog normalization value in "Soft Encode"? Thanks in advance!

[Skiller]

You're supposed to adjust it according to average dialog volume.

That's easier said than done however.

I usually import the center channel WAV (because that's where the dialog is) in WaveLab and measure the average volume level (RMS) of the entire file and in addition I look for a dialog of a few seconds with as few background sounds as possible and measure that as well.

Most of the time it's around -27 dB anyways and having this setting off by just one, two or three dB isn't going to do any harm. In fact the only situation in which a dial norm setting way off could do harm is when you are using DRC during playback. Otherwise it's almost irrelevant and the worst it could do is lower the overall volume.

[TeNSoR]

You're supposed to adjust it according to average dialog volume.

That's easier said than done however.

I usually import the center channel WAV (because that's where the dialog is) in WaveLab and measure the average volume level (RMS) of the entire file and in addition I look for a dialog of a few seconds with as few background sounds as possible and measure that as well.

Most of the time it's around -27 dB anyways and having this setting off by just one, two or three dB isn't going to do any harm. In fact the only situation in which a dial norm setting way off could do harm is when you are using DRC during playback. Otherwise it's almost irrelevant and the worst it could do is lower the overall volume.

Thanks for your reply, Skiller. Can you link me to this WaveLab?

[Skiller]

It's called Steinberg WaveLab, but it's not free.

There might be free ways to measure RMS, but I just don't know of any. Maybe somebody has an idea?

[TeNSoR]

I could find the demo version from the program, Skiller. Will you show me how to measure RMS of the entire file?

[Skiller]

Press ctrl + A to mark the entire waveform.

Then just press Y to bring up the Global Analyze tools.

Then select the Volume tab.

Increase the threshold to -70 or more (well, less actually, since it's negative).

Click "Analyze" and the average RMS volume will be displayed after it's done scanning where it says Average.

In my example (sorry it's the German version of WaveLab) it's -28.45 dB, hence I would set the Dialog Normalization setting for AC3 encoding to -28 dB.

[TeNSoR]

Press ctrl + A to mark the entire waveform.

Then just press Y to bring up the Global Analyze tools.

Then select the Volume tab.

Increase the threshold to -70 or more (well, less actually, since it's negative).

Click "Analyze" and the average RMS volume will be displayed after it's done scanning where it says Average.

In my example (sorry it's the German version of WaveLab) it's -28.45 dB, hence I would set the Dialog Normalization setting for AC3 encoding to -28 dB.


[Attachment 39191 - Click to enlarge]

Thank you so much! unfortunately, I used BeSweet to create six mono wave files
then I encoded in Sonic Soft Encode but the AC3 file was so low comparing with the original and I don't know why?! anyway, thanks again, Skiller for your help

[Skiller]

You are correct in making 6 mono WAV files!

Regarding the AC3 having a lower volume compared to the original, what was the original file format?

Also, what Dialog Normalization value did you come up with? If it's far off from -27 dB then there must be something wrong or your source is kinda odd.

During playback, do you have DRC (Dynamic Range Compression) enabled or not? DRC should not affect the volume of dialogs (due to the hopefully correctly set Dial Norm value) but it does make louder parts quiter and quiter parts louder to reduce the differnces in volume between loud and quiet parts.

Whenever you switch between DRC=on and DRC=off during playback the volume level of a standard dialog has to stay exactly the same. If it does change then the Dial Norm has an incorrect value.

[TeNSoR]

You are correct in making 6 mono WAV files!

Regarding the AC3 having a lower volume compared to the original, what was the original file format?

Also, what Dialog Normalization value did you come up with? If it's far off from -27 dB then there must be something wrong or your source is kinda odd.

During playback, do you have DRC (Dynamic Range Compression) enabled or not? DRC should not affect the volume of dialogs (due to the hopefully correctly set Dial Norm value) but it does make louder parts quiter and quiter parts louder to reduce the differnces in volume between loud and quiet parts.

Whenever you switch between DRC=on and DRC=off during playback the volume level of a standard dialog has to stay exactly the same. If it does change then the Dial Norm has an incorrect value.

Thanks again for reply I ran the analysis like you showed me and got this :



About DRC in Soft Encode this option in like this :

[Skiller]

Try if increasing the threshold to -100 changes the average volume by a few dB.

-18 dB is almost certainly not correct, especially since you already mentioned the AC3 becomes very quiet.

You should also look for a dialog scene of a few seconds (less than 10 seconds) with no loud background noises and measure that part the same way. Just mark it with the mouse and then analyze. Typically, Dial Norm should be in the range of -24 to -29 dB.

[TeNSoR]

Try if increasing the threshold to -100 changes the average volume by a few dB.

-18 dB is almost certainly not correct, especially since you already mentioned the AC3 becomes very quiet.

You should also look for a dialog scene of a few seconds (less than 10 seconds) with no loud background noises and measure that part the same way. Just mark it with the mouse and then analyze. Typically, Dial Norm should be in the range of -24 to -29 dB.

I did so and gave me the same value (-18db) and you're right the Dvd source is odd (Old Hindi) even I tried AAC in MeGUI and the same problem, Thanks so much Skiller for your help