Re: Re: Speaker-test improvements

[Takashi Iwai <tiwai@suse.de>]

At Thu, 14 Apr 2005 13:41:47 -0700 (PDT),
Tom Watson wrote:
> 
> Måns Rullgård <mru@inprovide.com> Writes:
> 
> 
> > Takashi Iwai <tiwai@suse.de> writes:
> >
> >> At Wed, 13 Apr 2005 21:02:31 +0100,
> >> James Courtier-Dutton wrote:
> >>> 
> >>> Hi,
> >>> 
> >>> I have got some voice samples saying things like "Front Right" etc. that
> >>> I want the speaker-test program to output instead of the loud sine wave
> >>> it currently does. The sine wave will stay there as an option as I find
> >>> it useful for some things.
> >>> 
> >>> The samples are 10 .wav files in 48khz mono S16_LE format.
> >>> Where abouts should I put them in the CVS, and where abouts should they
> >>> get installed to, so that speaker-test can find them?
> >>
> >> That's great.  (Is the voice female or male?  We might need both to be
> >> PC ;)
> >
> >Not to mention all the languages of the world.
> >
> >-- 
> >Måns Rullgård
> >mru@inprovide.com
> 
> 
> I made up a sound file similar to this and it was quite helpful.  One of the
> files had individual voices (one at a time) to identify the speakers, and being
> with the proper 4-channel specifications for broadcast they were identified as
> "left front", "right front", "left BACK" and "right BACK".  In doing so it
> eliminated the abbreviation of "R" being both "rear" and "right" (not very
> clear).
> 
> Another test was having all of the channels identified "simultaneously",
> usually in a loop.  This allowed me to play with mixer controls which always
> get mixed up and get the correct things in the correct places.  When you hook
> up sound cards and amplifiers and other external gear, it helps to have the
> channel identified.  Some mixer controls on some cards fold the "back" channels
> into the "front" channels, and then you need to poke around in your mixer
> program ('alsamixer' is but one) to turn on/off "mutes" to get it right. 

This sounds like a good idea.

> Having the speaker test "loop" with spoken words was VERY helpful to me.
> 
> The only "problem" I see is having the samples at one speed.  Sometimes it is
> helpful to have the program send the sounds out at a specific speed to the ALSA
> api, where it does further operations (another re-sampling).

It should be fine when you specify the PCM with "plug" layer, e.g.

	% speaker-test -Dplug:surround51 -c6

> Another useful option on a sequential test would be to have the channel
> identified on standard output at the beginning of the spoken words.
> 
> I really don't see a problem with the "voice" being male and English.  I
> suspect that since it is ONLY A TEST, we don't need to be PC at all.

Don't take my comment with a smiley so serious :)

> Perhaps the sound files for the sounds (mono for each, merged in the program)
> can be put in such a place that they can be altered to suit the user.  It may
> be that "LFE" may not be the desired name for the particular output.  There may
> be 8 or more channels to identify, and maybe having them labeled "one"... would
> be "better".  Given that more than one sound card might be used, and all slaved
> thru plugins, it could go higher.  Who knows.

Yes, that would be helpful, too.


Takahsi


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_ide95&alloc_id\x14396&op=click