[MPlayer-dev-eng] [PATCH] af_pan: hard clipping, and more than 100% of source channels

[Guillaume POIRIER]

Hi,

On 9/5/05, The Wanderer <inverseparadox at comcast.net> wrote:
> Since comment was re-requested, addressing the one thing I can really be
> sure about:
> 
> Corey Hickey wrote:
> 
> >      * remove_logo filter
> >      * lavcresample now used by default (-af-adv force=0 gives old behavior)
> >      * vf_expand and vf_dsize now support aspect and round parameters
> > +    * af_pan can mix channels in portions greater than 100%
> 
> I'm not entirely sure this is quite right... it's not inaccurate, it
> just seems like a strange choice of words. The only thing I can think of
> off the top of my head is that maybe you mean "proportions" instead, and
> I'm not completely sure that that means the same thing.
> 
> >  .IPs <lij>
> > -How much of input channel j is mixed into output channel i (0\-1).
> > +The percentage (as a decimal, 0.5=50%, 2=200%, etc.) that input channel
> > +j is to be mixed into output channel i (0\-512). That is 0% \- 51200%.
> 
> I'm not sure it's helpful to talk about this as a percentage; in
> addition to making the phrasing more difficult to handle without
> mangling the language, it also doesn't seem particularly clear, since
> the arguments to the option are not given as percentages. I might say
> something like "The multiple of input channel j which is to be mixed
> into output channel i"; that's not necessarily entirely clear, but it
> does avoid the additional level of obfuscation. (It took me a few
> seconds to figure out exactly what was being described here, with the
> current phrasing, and that's coming right after having read the much
> better description of the concept of mixing in the advanced audio
> features guide.)
> 
> Making the suggested change would arguably eliminate the need for the
> final above sentence entirely, but if you decide to keep it, it needs
> rejiggering. It refers back by pronoun ("that") to something inside
> parentheses in the previous sentence, the available range; the only way
> I can think of to use the current phrasing and be grammatically correct
> would be to say "(0-512, that is, 0% - 51200%).", which is inconsistent
> with the fact that the parenthetical bit is supposed to be simply a
> notation of the valid arguments for the function. I don't really have a
> good alternate suggestion; I haven't had enough sleep right now, but I'm
> up and reading E-mail because I couldn't get back to sleep for some
> reason.
> 
> > +Be careful when using values higher than 1; higher values can be quite
> > +loud and may exceed the sample range of your sound card. If you hear any
> > +painful pops or clicks, either decrease the values or follow pan with
> > +",volume" so the volume filter will clip out-of-range values. Clipping
> > +damages sound quality so use it sparingly.
> 
> I thought I noticed several minor grammatical issues here, but looking
> at it now I only see one: you need a comma after "sound quality".

Hopefully the attached patch should feature The Wanderer's
suggestions. Please comment on it, I feels like committing it soon.

Guillaume
-- 
Reading doesn't hurt, really!
  -- Dominik 'Rathann' Mierzejewski
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: af_pan.diff
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20050911/da42f2f2/attachment.asc>