On 11/1/18, Fred Gleason wrote:
> On Oct 31, 2018, at 15:43, Gregory Avedissian wrote:
>
>> Is it safe to do something as simple as having the script go to /var/snd
>> and run:
>> for i in *.wav ; do sox --norm=-5 "$i" /tmp/"$i" && cp /tmp/"$i" "$i" ;
>> done
>>
>> Oh yeah, there will be rm
There's a project called bs1770gain which normalizes audio to a set
lufs, loudness.
Marius
On 11/2/18 4:27 PM, Andy Higginson wrote:
I seem to remember some discussion on using EBU R128 loudness
measuring for normalization of audio. It was mentioned a couple of
years ago and I think someone
I seem to remember some discussion on using EBU R128 loudness measuring for
normalization of audio. It was mentioned a couple of years ago and I think
someone on this list had found some open source code to enable this monitoring.
Has anything further happened on this? On Thu, 01 Nov
On Nov 1, 2018, at 10:56, Rob Landry <41001...@interpring.com> wrote:
> My understanding of "normalize" is that it means multiplying every sample in
> a file by some coefficient calculated to set the largest sample in the file,
> i.e. the peak amplitude, to a specific point, such as -13 dBfs.
>
On Thu, 1 Nov 2018 11:06:31 +1300
Robert Jeffares wrote:
> In digital audio, overshoots cause lots of instant distortion.
Almost right.
"Overshoot" has a specific technical meaning, ( involving bass ) and
isn't as intuitive as one might think, but I think I know what you mean.
In analog
On Thu, 1 Nov 2018 07:45:26 -0400
Fred Gleason wrote:
> and rewriting audio can cause quality degradation (especially if MPEG Layer
> II format is
> involved).
Can cause ?
I would argue that rewriting audio in any format other than what it already is,
absolutely *will* cause some degree of
My understanding of "normalize" is that it means multiplying every sample
in a file by some coefficient calculated to set the largest sample in the
file, i.e. the peak amplitude, to a specific point, such as -13 dBfs.
However, lately I've seen "normalize" used to mean adjusting the *average*
On Oct 31, 2018, at 15:43, Gregory Avedissian wrote:
> Is it safe to do something as simple as having the script go to /var/snd
> and run:
> for i in *.wav ; do sox --norm=-5 "$i" /tmp/"$i" && cp /tmp/"$i" "$i" ; done
>
> Oh yeah, there will be rm /tmp/"$i" in there, too. Just noticed that.
>
Gregory,
it will work
in that your /var/snd/ files will all be normalised by sox to -5.
However the -13 level is a better option than bumping the audio file
level up.
For a number of reasons the -13 level was chosen to allow 'headroom' for
the bits of audio that overshoot 0vu.
In digital
We've had a problem with broadcast volume levels, and we've settled on
using a normalization level of -5 when importing into the library. The
engineer has tweaked the transmission as best he can without bothering the
nearby airport.
There are still a bunch of carts that were imported at -13.
10 matches
Mail list logo
