Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Replaygain offset (Read 4206 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Replaygain offset

Hi,

I'm calculating the average RMS value of a CD (mostly between -8 dB and -14 dB) but I'd like to compare that with what replaygain calculates (without their reference signal), so that 2 results should be pretty close. Can anyone tell me how I can change the replaygain result into the same scale as my average RMS?

Robin

Replaygain offset

Reply #1 –
You probably won't be able to get much correlation between the two. What replaygain calculates is vastly different than average RMS.

Replaygain offset

Reply #2 –
I agree, that's what I'd like to show people, that replaygain will give a better correlation to the ear (I'm dealing with audio engineers).

But I'd like to know how I can get both onto the same scale?

Thanks

Robin

Replaygain offset

Reply #3 –
Why don't you just replaygain a file then calculate the average RMS of the replaygained file. This will give you the difference in dB between the two methods on this one file.

Replaygain offset

Reply #4 –
That's not really true as the replaygain calculation uses a reference level (http://replaygain.hydrogenaudio.org/calibration.html) and the replaygain will be the difference between the reference (-20 dBFS pink noise) and the calculated file, where as the average RMS is calculated with a reference of full scale. Any way, I need to display the average RMS (which I'm doing without a problem) and the replaygain BASED ON THE SAME SCALE. I'd like to know how I can do this?

Robin

Replaygain offset

Reply #5 –
ReplayGain is frequency dependent, RMS is not. Therefore you can tie them together for a given frequency sine wave - but it'll be wrong for all other sine waves, and any multi-frequency signal.

To tie them together, subtract the ReplayGain values from a chosen conversion reference level.

If you want to tie them together based on the loudest frequency RG "hears", use -22-RG value.

If you want to tie them together based on pink noise (a better choice), use -11.86-RG.

e.g.
RG=0
RMS = -11.86=-11.86

RG=-6
RMS = -11.86--6=-5.86

RG=+6
RMS = -14-6=-17.86

This gives RMS relative to 0dB FS (for a full scale sine wave) for the chosen conversion reference level.

Hope this helps.

Cheers,
David.

Replaygain offset

Reply #6 –
Hi, thanks for the information. I'm trying it with pink noise (-20 dBFS unweigthed),

RMS: -19.91 dB
Replay gain: 5.19 dB

If I do: replaygain = -11.86 - 5.19 dB, I get -17.05 dB which seems close, but when I try it with other recordings it seems that when the RMS goes down, the replaygain goes up. Am I doing it right now, or is there still something wrong?

Thanks

Robin

Replaygain offset

Reply #7 –
Without wishing to sound insulting, did you remember that minus a minus is a plus?

Other than that, the correlation between RMS and ReplayGain really isn't very good. However, if you take a track and drop it by 6dB then the ReplayGain should increase by 6dB, and your "RMS-like ReplayGain equivalent" should drop by 6dB.

Cheers,
David.

Replaygain offset

Reply #8 –
Ow I know that replaygain works like that, but I'd like to use it different. Now I use:

-14.73 - analyzeResult ( A, sizeof(A)/sizeof(*A) );

Which gives both 19.92 dB RMS and Replaygain for my pinknoise signal. Although signals seem to have a higher RG, but now they're equivalent (RMS drops = RG drops).

That's what I wanted to see, I just hope that I'm doing it right!

Kind regards,

Robin