PDA

View Full Version : flac files - time wrong, causes continuous refresh



John Moody
2003-12-27, 05:34
Version 5.0.1 did not play my flac files.
I tried several versions from the nightly builds, they
all play them.

Here is the problem.
The song time runs about 4 times too fast.
When displaying remaining time; the time counts down
to -00:00 then jumps to -00:59, when it hits 00:00
again it displays a double minus (--1:59).

When the double minus appears, the browser refresh
timing goes to zero causing continuous refresh. The
cpu load causes dropouts.

Any ideas? (I'm using a dual cpu windows box)


__________________________________
Do you Yahoo!?
New Yahoo! Photos - easier uploading and sharing.
http://photos.yahoo.com/

Dan Sully
2003-12-27, 10:15
* John Moody <moodymz3 (AT) yahoo (DOT) com> shaped the electrons to say...

> Here is the problem.
> The song time runs about 4 times too fast.
> When displaying remaining time; the time counts down
> to -00:00 then jumps to -00:59, when it hits 00:00
> again it displays a double minus (--1:59).
>
> When the double minus appears, the browser refresh
> timing goes to zero causing continuous refresh. The
> cpu load causes dropouts.

These are known problems - and are being worked on.

-D
--
If it wasn't for my horse, I wouldn't have spent that year in college.

Dan Sully
2003-12-28, 23:07
* Dan Sully <daniel (AT) electricrain (DOT) com> shaped the electrons to say...

> > When the double minus appears, the browser refresh
> > timing goes to zero causing continuous refresh. The
> > cpu load causes dropouts.
>
> These are known problems - and are being worked on.

Can you try the cvs build, or tonight's nightly build?

-D
--
Down, not across.

Caleb Epstein
2003-12-29, 06:14
On Sun, Dec 28, 2003 at 10:07:54PM -0800, Dan Sully wrote:

> * Dan Sully <daniel (AT) electricrain (DOT) com> shaped the electrons to say...
>
> > > When the double minus appears, the browser refresh
> > > timing goes to zero causing continuous refresh. The
> > > cpu load causes dropouts.
> >
> > These are known problems - and are being worked on.
>
> Can you try the cvs build, or tonight's nightly build?

I'm unfortunately not at my squeeze to check this, but the
comments and code indicate that wall clock time will be used
for the counters. What happens when you pause and resume?

--
Caleb Epstein | bklyn . org |
cae at | Brooklyn Dust | Good day to avoid cops. Crawl to school.
bklyn dot org | Bunny Mfg. |

Dan Sully
2003-12-29, 09:51
* Caleb Epstein <cae (AT) bklyn (DOT) org> shaped the electrons to say...

> > Can you try the cvs build, or tonight's nightly build?
>
> I'm unfortunately not at my squeeze to check this, but the
> comments and code indicate that wall clock time will be used
> for the counters. What happens when you pause and resume?

Exactly what you'd think would happen.

One bug fixed, another one created. *sigh*.

-D
--
<iNoah> kernel's original recipe: 11 secret args and switches

Caleb Epstein
2003-12-29, 20:59
On Mon, Dec 29, 2003 at 08:51:16AM -0800, Dan Sully wrote:

> * Caleb Epstein <cae (AT) bklyn (DOT) org> shaped the electrons to say...
>
> > > Can you try the cvs build, or tonight's nightly build?
> >
> > I'm unfortunately not at my squeeze to check this, but the
> > comments and code indicate that wall clock time will be used
> > for the counters. What happens when you pause and resume?
>

> Exactly what you'd think would happen. > One bug fixed,
> another one created. *sigh*.

There's also a long-standing problem I have with pause/resume
not working for FLACs anyway. Maybe half the time I try
pausing a FLAC file during playback and then resume playing, I
get nothing but static from the Squeezebox. It seems to be
more frequent if the pause is short.

What exactly is the linkage between the FLAC > PCM pipeline
that is driven by a line from convert.conf and the packets
that get sent to the Squeezebox, and is it possible for the
code thats driving the elapsed/remaining time display to know
more about the state of the Squeeze and keep tabs on where it
really is in the stream? I think I need to look at the code
more closely. I assume these bugs would be present for any
audio that was being streamed via convert.conf -> pcm?

--
Caleb Epstein | bklyn . org | If your life was a horse, you'd have to
cae at | Brooklyn Dust | shoot it.
bklyn dot org | Bunny Mfg. |

Dan Sully
2003-12-29, 21:49
* Caleb Epstein <cae (AT) bklyn (DOT) org> shaped the electrons to say...

> What exactly is the linkage between the FLAC > PCM pipeline
> that is driven by a line from convert.conf and the packets
> that get sent to the Squeezebox, and is it possible for the
> code thats driving the elapsed/remaining time display to know
> more about the state of the Squeeze and keep tabs on where it
> really is in the stream? I think I need to look at the code
> more closely. I assume these bugs would be present for any
> audio that was being streamed via convert.conf -> pcm?

That is correct - see Slim::Player::Source::songTime() and
Slim::Buttons::Playlist::songTime();

-D
--
God, root, what is difference?