On Mon, Sep 11, 2017 at 11:10 AM, David Noble <[email protected]> wrote:
> > I agree with this, the video operator is the one who can easily check >> the levels being received and so should be aware of the nominal levels > > > This might be a deviation, but I think it might also be a good way of > improving the audio quality and consistency significantly. > > With regards to this specifically, I suggest that some sort of R128 > monitoring be implemented in Voctomix. A really useful example of this is > in ebumeter(1): http://kokkinizita.linuxaudio.org/linuxaudio/ebumeter-doc/ > quickguide.html - it might be interesting to look at how it behaves when > some previously recorded material is played through it, to get an idea of > the readings it gives. > > The I-value shows the deviation from target loudness - and the operator > should ideally aim for a value of zero. This is quite a lot easier than > trying to explain the concepts of loudness in relation to PPM or VU meters. > If the value is positive, gradually reduce the signal level, if the value > is negative, increase it. > > The remaining risk beyond this is clipping - this can be reduced by using > a limiter, or a soft-knee compressor before the signal is input into the > computer. > > Can you add this as an ER? this seems to be the place: https://github.com/voc/voctomix/issues/157 > For what it's worth, Nageru implements this form of loudness monitoring > already. > Can you point to the code that implements it? > > David > > _______________________________________________ > Debconf-video mailing list > [email protected] > http://lists.debconf.org/mailman/listinfo/debconf-video > >
_______________________________________________ Debconf-video mailing list [email protected] http://lists.debconf.org/mailman/listinfo/debconf-video
