I play a video project including 2 tracks set at -2.3 dB. I get a volume peak measured by the "Master Bus" module of something like -0.1 dB for left channel and -0.2 dB for right channel.
Now, I see both tracks to -2.2 dB, I now get volume peak of +1.2 dB for both channels.
Also, when I play it again, not changing anything, I may get different results.
Is there something going with dB that I didn't get right or is this measurement just not reliable?
Lossy or lossless audio codec?
Media sample rate/project sample rate match or mismatch?
+ added to first post: Also, when I play it again, not changing anything, I may get different results.
Project: 48 kHz, 24 bit
Media 1 (FLAC from CD): 44.1 kHz, 16 bit
Media 2 (sound from a video recording in a Sony .mts file, in this project it's from a Sony RX100 M5) : 48 kHz (not sure about bit depth)
Use the True Peak meters in the Loudness Meters window and see what you get, and understand that intersample peaks can give inconsistent metering results.
It may be irrelevant but my Canon camera mts files give me volume problems in VMS 16 and7. Possibly it's the same in Pro. It's the ac3 encoded with Dolby that is the issue. Vegas nowadays uses Windows components for ac3. A Vegas rendered AVCHD file (not using Dolby) with ac3 is perfect.
and understand that intersample peaks can give inconsistent metering results.
Thank you. Very confusing to me as I thought that the computation of the same data multiple times should always give the same results, but thanks to your comment I spotted other references mentioning it. (I won't have time to dive into those right now.)
I had already noticed some inconsistencies, but never as big as in this current project...
It may be irrelevant but my Canon camera mts files give me volume problems in VMS 16 and7. Possibly it's the same in Pro. It's the ac3 encoded with Dolby that is the issue. Vegas nowadays uses Windows components for ac3. A Vegas rendered AVCHD file (not using Dolby) with ac3 is perfect.
Thank you. I have not noticed real volume issues in this type of projects... Though I don't need them to be super precise and I can afford some imprecision. Also, I'm not talking about a rendered file but really only about the peak measure as displayed (in "Master Bus") while editing
Consider that Vegas, by default, adds Track fX, Track Noise Gate, Track EQ and Track Compressor. It might be worth removing them to see if you get results that you are expecting.
@EricLNZ said:
"Vegas nowadays uses Windows components for ac3. A Vegas rendered AVCHD file (not using Dolby) with ac3 is perfect."
I've been curious about the volume effects on AVCHD (with AC3) projects that I created in Vegas 13 and earlier when opened and rendered in 14 through 19 because of the different decoding methods.
I created identical projects in Vegas 13-453 and Vegas 19-424 with AVCHD (AVC video and AC3 audio in an MTS wrapper) from a Sony camcorder. I removed all track fX and rendered to 48kHz, 16 bit WAV files. Opening each of the rendered files in Sound Forge 9 gave me the following Statistics:
I created identical projects in Vegas 13-453 and Vegas 19-424 with video shot with a Canon camera that recorded AVC video and PCM audio in a MOV wrapper. The video was of the same event at the same time and Quicktime was not required for these files by either version of Vegas. I removed all the default track fX and rendered to 48kHz, 16 bit WAV files. Opening each of the rendered files in Sound Forge 9 gave me the following Statistics:
I observed that there is variation in rendered output between Vegas versions for AC3 audio while the is none for PCM. You decide whether the variation is meaningful.
For submissions, (broadcast. streaming, ect) all use the loudness measurement these days. This is totally different than full scale peak, average RMS, ect. The most important factor in the loudness measurement is the LUFS integrated factor, true peak dBTP is important as well but most causes of rejections in the 'integrated' loudness factor, which is measured over a period of time.. Whilst RMS ts closer to LU loudness than peak, all will have a different reading. In the editing stage, full scale peak (dBFS) is the important factor. Any file rendered over 0.0 dB is clipped. Loudness is usually fine tuned in the mastering stage.. if there is one.