[FFmpeg-user] processing a vumeter (levelmeter) as video for audio inputs

[Kuban Altan]

Thanks Dave,
But there is a minor issue with ebur128 Momentary level. It is defined for
a 0.4 second time range. Which is very long compared to 0.04 second for
each frame in PAL standard. Do you think that there would be a possible
workaround to hijack ebur128 Momentary (0.4s) vumeter, to use 0.04 second?

kuban altan

On Fri, Nov 7, 2014 at 5:21 PM, Dave Rice <dave at dericed.com> wrote:

> Hi,
>
> > On Nov 7, 2014, at 8:27 AM, Kuban Altan <kuban.altan at gmail.com> wrote:
> >
> > Hi,
> > I am testing *ebur128* filter, which can produce a video output. But I
> need
> > a simple *vumeter* display per audio channel, which does not seem to
> exist.
> > Any idea to accomplish this?
> >
> > So far, I have only been able to *volumedetect* per frame, mentioning a
> -ss
> > time and a duration with -t 0.04 (for 25fps)... So I can get a peak sound
> > level as text output using volumedetect. But this is not very efficient.
> > All I need is a video and 2 rectangular bars, which have their lengths
> > connected to input audio level.
>
> There is a momentarily ebur128 level in the ebur128 video output so you
> could crop that out and reuse it noting that the concept that it measures
> is differently than a typical vu meter. Here’s an example where I split a
> stereo input and have two side by side momentary ebur128 meters:
>
> ffplay -f lavfi
> "amovie=stereo.mp3,channelsplit[a][b];[a]ebur128=video=1:meter=18[a1][out0];[b]ebur128=video=1:meter=18[b1][out1];[a1]crop=20:432:612:40[a2];[b1]crop=20:432:612:40[b2];[a2][b2]framepack[out2]”
>
> From here you could use drawbox and/or drawtext to add labelling.
>
> Dave Rice
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>