PDA

View Full Version : FLAC Cue Sheet Support



Stephen Sweeney
2004-10-13, 12:24
I'm able to get FLAC Cue Sheets working - basically flac disc image and an
external cue sheet. Excellent work, developers!

I'm wondering why the Cue Sheets embedded in the flac file aren't used. I
understand why an external cue sheet might take priority over an embedded cue
sheet. But when the external cue sheet is absent, why wouldn't the embedded
cue sheet doesn't seem to be used?

Jason Holtzapple
2004-10-13, 13:12
The last time I looked at this, there was a bit of a conflict between how
the server handles file types. It expects a file to be a playlist or a
music file, but not both.

Someone has contributed code to parse the internal cuesheet already, some
clever hacking remains to get those tracks presented to the user. Then
there's the fairly hairy area of internal cuesheet metadata. None of this
is insurmountable though. :)

Stephen Sweeney wrote:
> I'm able to get FLAC Cue Sheets working - basically flac disc image and an
> external cue sheet. Excellent work, developers!
>
> I'm wondering why the Cue Sheets embedded in the flac file aren't used. I
> understand why an external cue sheet might take priority over an embedded cue
> sheet. But when the external cue sheet is absent, why wouldn't the embedded
> cue sheet doesn't seem to be used?

michael
2004-10-13, 15:41
With recent releases of slimserver, the embedded cuesheet should be
properly read (in the absence of an external cuesheet). While there's
still some work to do in this area, the basic functionality is in
place. If that's not working right for some reason, I'd like to hear
more about it.

As Jason mentions, metadata for the contained tracks is not
standardised, but there are a few workable systems in place.
How are your flacs tagged?

If you include a bit more information, I might be able to better
help. Which version of slimserver are you using? How does it react to
a flac with an embedded cuesheet? (Does it not show up at all? Does it
show up as one big long song? etc..)

Thanks,

-michael

Jason Holtzapple <jasonholtzapple (AT) yahoo (DOT) com> writes:
> The last time I looked at this, there was a bit of a conflict between how
> the server handles file types. It expects a file to be a playlist or a
> music file, but not both.
>
> Someone has contributed code to parse the internal cuesheet already, some
> clever hacking remains to get those tracks presented to the user. Then
> there's the fairly hairy area of internal cuesheet metadata. None of this
> is insurmountable though. :)
>
> Stephen Sweeney wrote:
>> I'm able to get FLAC Cue Sheets working - basically flac disc image
>> and an external cue sheet. Excellent work, developers!
>> I'm wondering why the Cue Sheets embedded in the flac file aren't
>> used. I understand why an external cue sheet might take priority
>> over an embedded cue sheet. But when the external cue sheet is
>> absent, why wouldn't the embedded cue sheet doesn't seem to be used?

Stephen Sweeney
2004-10-13, 21:58
michael <michael@...> writes:
>
> With recent releases of slimserver, the embedded cuesheet should be
> properly read (in the absence of an external cuesheet). While there's
> still some work to do in this area, the basic functionality is in
> place. If that's not working right for some reason, I'd like to hear
> more about it.
>
> As Jason mentions, metadata for the contained tracks is not
> standardised, but there are a few workable systems in place.
> How are your flacs tagged?
>
> If you include a bit more information, I might be able to better
> help. Which version of slimserver are you using? How does it react to
> a flac with an embedded cuesheet? (Does it not show up at all? Does it
> show up as one big long song? etc..)
>
> Thanks,
>
> -michael
>
> Jason Holtzapple <jasonholtzapple@...> writes:
> > The last time I looked at this, there was a bit of a conflict between how
> > the server handles file types. It expects a file to be a playlist or a
> > music file, but not both.
> >
> > Someone has contributed code to parse the internal cuesheet already, some
> > clever hacking remains to get those tracks presented to the user. Then
> > there's the fairly hairy area of internal cuesheet metadata. None of this
> > is insurmountable though. :)
> >
> > Stephen Sweeney wrote:
> >> I'm able to get FLAC Cue Sheets working - basically flac disc image
> >> and an external cue sheet. Excellent work, developers!
> >> I'm wondering why the Cue Sheets embedded in the flac file aren't
> >> used. I understand why an external cue sheet might take priority
> >> over an embedded cue sheet. But when the external cue sheet is
> >> absent, why wouldn't the embedded cue sheet doesn't seem to be used?
>

Thanks for the offer to help. I'm running SlimServer 5.3.1 for Windows -
10/13 build. I use flac 1.1.1 and EAC 0.95PB5.

I rip the CDs with EAC using Copy Image & Create CUE Sheet.

When I encode the .wav to .flac through EAC external compression and leave the
cue sheet external, it seems that SlimServer does not correctly recognize the
last track in the EAC CUE sheet. Otherwise, everything seems to work fine.
When browsing the folders through SlimServer, the final album track shows the
name of the second track, which seems to be default when there is no cue
sheet. When browsing artists or albums, the final track does not show up at
all.

When I use the same method of encoding but embed the EAC CUE sheet (at the
time of encoding using the --cuesheet="CDimage.flac.cue" flag), the track info
doesn't show up in the metadata.

So perhaps I haven't found the best way to embed the EAC cue sheet to the FLAC
file. Does EAC generate a non-compliant CUE sheet when using the Copy Image &
Create CUE Sheet command?

michael
2004-10-14, 14:31
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:
> Thanks for the offer to help. I'm running SlimServer 5.3.1 for Windows -
> 10/13 build. I use flac 1.1.1 and EAC 0.95PB5.

It sounds like you've got the right collection of recent versions.

> I rip the CDs with EAC using Copy Image & Create CUE Sheet.
>
> When I encode the .wav to .flac through EAC external compression and leave the
> cue sheet external, it seems that SlimServer does not correctly recognize the
> last track in the EAC CUE sheet. Otherwise, everything seems to work fine.
> When browsing the folders through SlimServer, the final album track shows the
> name of the second track, which seems to be default when there is no cue
> sheet.

Unfortunately, browsing folders of files that depend on cuesheets is
one of those areas that still needs some improvement.

> When browsing artists or albums, the final track does not show up at
> all.

I used to have that problem with embedded cuesheets, where the last
track would either disappear, or would become the entire album. But
it's been fixed for awhile now. I don't use external cuesheets much,
other than for testing, so it's possible that this slipped
though. I'll see if I can replicate this on my end.

> When I use the same method of encoding but embed the EAC CUE sheet (at the
> time of encoding using the --cuesheet="CDimage.flac.cue" flag), the track info
> doesn't show up in the metadata.
>
> So perhaps I haven't found the best way to embed the EAC cue sheet to the FLAC
> file.

So, there's (at least) one thing that isn't really obvious without
scouring the documentation. Embedding a cuesheet into a flac file
includes all the info on track indexes, but doesn't include CD-TEXT.
http://flac.sourceforge.net/faq.html#general__no_cuesheet_tags
Unfortunately, most programs that parse external cuesheets use CD-TEXT
as a way to store metadata, even if there was no CD-TEXT on the
original disc. So, when you embed that cuesheet into your flac file,
you'll still need to tag the files metadata.
I don't know if this is part of your problem or not, but it's worth
looking at.

This brings us into the whole debate about how to tag a file with
multiple tracks, an issue that the flac developers have intentionally
avoided. A program like foobar2000 http://foobar2000.org/ will stuff a
cddb record into the metadata field of the flac file, and slimserver
will parse this. Alternately, if you're going to enter the metadata by
hand, there are cleaner (i.e. easier to parse) formats that slimserver
will understand. One example is...

ARTIST=foo
ALBUM=bar
TRACKNUMBER[1]=1
TITLE[1]=Popular Song
TRACKNUMBER[2]=2
TITLE[2]=Radio Hit
....

There are other formats that will work as well, just make sure you've
got one of them in place.

> Does EAC generate a non-compliant CUE sheet when using the Copy Image &
> Create CUE Sheet command?

If you send me a copy of one of the generated cuesheets, I'll make
sure that slimserver can parse it, especially in light of the above
issue about losing the last track. (eac is popular enough that even if
it's non-compliant, it's probably worth making slimserver parse it as
correctly as possible.)

I hope this helps some.

-michael

Stephen Sweeney
2004-10-15, 15:02
michael <michael@...> writes:
>
> Stephen Sweeney <ssweeney@...> writes:
> > When browsing the folders through SlimServer, the final album track shows
the
> > name of the second track, which seems to be default when there is no cue
> > sheet.
>
> Unfortunately, browsing folders of files that depend on cuesheets is
> one of those areas that still needs some improvement.
>
Understood.

> > When browsing artists or albums, the final track does not show up at
> > all.
>
> I used to have that problem with embedded cuesheets, where the last
> track would either disappear, or would become the entire album. But
> it's been fixed for awhile now. I don't use external cuesheets much,
> other than for testing, so it's possible that this slipped
> though. I'll see if I can replicate this on my end.
>
Interested whether you found anything.

> >
> > So perhaps I haven't found the best way to embed the EAC cue sheet to the
FLAC
> > file.
>
> This brings us into the whole debate about how to tag a file with
> multiple tracks, an issue that the flac developers have intentionally
> avoided. A program like foobar2000 http://foobar2000.org/ will stuff a
> cddb record into the metadata field of the flac file, and slimserver
> will parse this. Alternately, if you're going to enter the metadata by
> hand, there are cleaner (i.e. easier to parse) formats that slimserver
> will understand. One example is...
>
Where does foobar put the text? Vorbis, ID or both?
>
> > Does EAC generate a non-compliant CUE sheet when using the Copy Image &
> > Create CUE Sheet command?
>
> If you send me a copy of one of the generated cuesheets, I'll make
> sure that slimserver can parse it, especially in light of the above
> issue about losing the last track. (eac is popular enough that even if
> it's non-compliant, it's probably worth making slimserver parse it as
> correctly as possible.)
>
I emailed a cue sheet that doesn't get processed as I'd expect. Don't know if
it is the parser or the file.

It was generated from EAC using the standard option, which I think defaults to
the "non-compliant" layout. SlimServer shows the last track (11) with the
name of the 2nd track.

I appreciate your help.

BTW - this is the vorbis metadata block from the flac file. It is encoded
this way by default from the image .wav. Funny.

METADATA block #3
type: 4 (VORBIS_COMMENT)
is last: false
length: 172
vendor string: reference libFLAC 1.1.1 20041001
comments: 6
comment[0]: artist=Harry Connick Jr.
comment[1]: title=Love Is Here To Stay
comment[2]: album=When Harry Met Sally
comment[3]: date=1989
comment[4]: tracknumber=2
comment[5]: genre=Jazz

Stephen Sweeney
2004-10-15, 21:12
>
> > When browsing artists or albums, the final track does not show up at
> > all.
>
> I used to have that problem with embedded cuesheets, where the last
> track would either disappear, or would become the entire album. But
> it's been fixed for awhile now. I don't use external cuesheets much,
> other than for testing, so it's possible that this slipped
> though. I'll see if I can replicate this on my end.
>
Also, when there is ONLY an external cue sheet - no internal one - SlimServer
correctly processes the track listing correctly. However, the entire album
does show up in the track listing as a separate entry. Is this the expected /
desired behavior?

So it seems that the problem I was seeing was with the presence of an external
cue sheet (I sent it to you), an internal cue sheet (with no track names, as
flac doesn't support that), and a single set of vorbis comments (below). The
vorbis contains a track title and number, which must be part of the problem.

METADATA block #3
type: 4 (VORBIS_COMMENT)
is last: false
length: 172
vendor string: reference libFLAC 1.1.1 20041001
comments: 6
comment[0]: artist=Harry Connick Jr.
comment[1]: title=Love Is Here To Stay
comment[2]: album=When Harry Met Sally
comment[3]: date=1989
comment[4]: tracknumber=2
comment[5]: genre=Jazz

michael
2004-10-16, 20:25
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:

> Also, when there is ONLY an external cue sheet - no internal one - SlimServer
> correctly processes the track listing correctly.

Having both an external cuesheet and an internal cuesheet for the same
file is bound to confuse slimservers parsing of that file. For now,
it's probably best to stick with one or the other (for a given file).

> However, the entire album
> does show up in the track listing as a separate entry. Is this the
> expected / desired behavior?

It's certainly not the desired behaviour.

> So it seems that the problem I was seeing was with the presence of an external
> cue sheet (I sent it to you), an internal cue sheet (with no track names, as
> flac doesn't support that), and a single set of vorbis comments (below). The
> vorbis contains a track title and number, which must be part of the problem.

For an external cuesheet, I'd suggest keeping the vorbis comments
clear, so you don't have conflicts between what the cuesheet is
claiming and what the vorbis comments claim. This is probably the
issue you're seeing with track 2 being the only one in the vorbis
comments. (Why a piece of software would pick track 2 as the only one
to include is beyond me.)

For an internal cuesheet, You'll want to setup the vorbis tags to
represent all of your metadata.

Mixing and matching (for the same file) is best avoided until
slimserver handles the ambiguity a bit better.

That said, the cuesheet you passed along seems to parse ok.
I tried it as an external cuesheet (pointing to a flac with completely
different contents, since I don't own that album).
I then tried it instead as an internal cuesheet (with hand crafted
vorbis comments) and that seemed to work ok as well.
I'm not sure without going over the spec what makes it
"non-compliant", but it seems to work ok with slimserver.

> Where does foobar put the text? Vorbis, ID or both?

If I understand it correctly, foobar2000 stuffs the content of a cddb
record into the vorbis comments. I can't run foobar2k myself, so I
could be understanding this wrong.

I'm still trying to track down what triggers showing the entire album
as a separate track entry. I'm not seeing this behaviour currently,
but I have seen it before, and if you're still experiencing it, I may
bug you some more for details about your setup.

I hope this helps.

-michael

Stephen Sweeney
2004-10-17, 18:31
michael <michael@...> writes:

>
> Stephen Sweeney <ssweeney@...> writes:
> >
> > However, the entire album
> > does show up in the track listing as a separate entry. Is this the
> > expected / desired behavior?
>
> It's certainly not the desired behaviour.
>
I figured this out. I encoded the entire album with artist and album vorbis
tags. SlimServer picks these and lists the .flac it as a playable entry in
browse by artist or browse by album. Removing these tags "fixes" the
problem.

I think we should be able to encode the album and artist in vorbis with an
external cue sheet. Otherwise, only the filename identifies the album, which
is a little scary. What do you think?
>
> For an internal cuesheet, You'll want to setup the vorbis tags to
> represent all of your metadata.

I'm going to stick with the external cue sheet. I'm not comfortable embedding
metadata if there is no clear standard. I guess I'll wait.

>
> I'm still trying to track down what triggers showing the entire album
> as a separate track entry. I'm not seeing this behaviour currently,
> but I have seen it before, and if you're still experiencing it, I may
> bug you some more for details about your setup.

Let me know what else you need. I think the above explanation is it.
>
> I hope this helps.
>
> -michael
>
Most definitely. Thanks a million.

stephen

michael
2004-10-17, 20:51
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:
....
> I figured this out. I encoded the entire album with artist and album vorbis
> tags. SlimServer picks these and lists the .flac it as a playable entry in
> browse by artist or browse by album. Removing these tags "fixes" the
> problem.

Good to know. Certainly having both internal and external references
pointing to the same file is an area that needs work (not just in
regard to flac).

> I think we should be able to encode the album and artist in vorbis with an
> external cue sheet. Otherwise, only the filename identifies the album, which
> is a little scary. What do you think?

That's one reason I stick with internal cuesheets. No way for them to
get separated from their codependent file. Others have chosen to do
the opposite, and as we've discovered, the grey areas in between are
still in need of some polishing.

I certainly intend to work out some of those issues with external and
internal data, and eventually even allow multiple sources to reference
the same file (or parts of it). But some of that may only work it's way
into the future db backed slimserver.

> I'm going to stick with the external cue sheet. I'm not comfortable embedding
> metadata if there is no clear standard. I guess I'll wait.

It may be awhile before various vendors settle on a default standard
(remember that xiph is specifically avoiding being a part of
this). Most programs don't really even know how to handle multiple
tracks in a single file yet. Right now, as far as I can tell, the
closest thing to a standard would be the cddb thing that foobar
does. It's supported by two programs (foobar2000 and slimserver).

Glad to hear you got it working though.

-michael

Stephen Sweeney
2004-10-18, 08:41
michael <michael@...> writes:
>
> Stephen Sweeney <ssweeney@...> writes:
>
> > I think we should be able to encode the album and artist in vorbis with an
> > external cue sheet. Otherwise, only the filename identifies the album,
> > which is a little scary. What do you think?
>
> That's one reason I stick with internal cuesheets. No way for them to
> get separated from their codependent file. Others have chosen to do
> the opposite, and as we've discovered, the grey areas in between are
> still in need of some polishing.
>
Getting the cue sheet into .flac is no problem. It's getting the metadata
tags into vorbis. All (or almost all) the info is written by EAC to the cue
sheet. So if there were a script to parse the cue sheet into the format that
could be loaded into vorbis tags, I'd do it. Metaflac can import the tags
from an input file. Have you ever seen a script that does this? Or is there
an easier way?

Thanks so much for your help, btw. Much appreciated! SS

michael
2004-10-19, 15:37
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:
> Getting the cue sheet into .flac is no problem. It's getting the metadata
> tags into vorbis. All (or almost all) the info is written by EAC to the cue
> sheet. So if there were a script to parse the cue sheet into the format that
> could be loaded into vorbis tags, I'd do it. Metaflac can import the tags
> from an input file. Have you ever seen a script that does this? Or is there
> an easier way?

I'm sure it would be fairly easy to throw together a bit of perl that
creates a vorbis comments file from the eac cue sheet. Let me see what
I can come up with...

-michael

David Lutz
2004-10-19, 21:51
I have a very simple python script that I use. I always seem to
need to hand edit what CDDB returns, so this script may not be
complete enough for you, but it should be a start.

D.L.


On Mon, Oct 18, 2004 at 03:41:46PM +0000, Stephen Sweeney wrote:
> michael <michael@...> writes:
> >
> > Stephen Sweeney <ssweeney@...> writes:
> >
> > > I think we should be able to encode the album and artist in vorbis with an
> > > external cue sheet. Otherwise, only the filename identifies the album,
> > > which is a little scary. What do you think?
> >
> > That's one reason I stick with internal cuesheets. No way for them to
> > get separated from their codependent file. Others have chosen to do
> > the opposite, and as we've discovered, the grey areas in between are
> > still in need of some polishing.
> >
> Getting the cue sheet into .flac is no problem. It's getting the metadata
> tags into vorbis. All (or almost all) the info is written by EAC to the cue
> sheet. So if there were a script to parse the cue sheet into the format that
> could be loaded into vorbis tags, I'd do it. Metaflac can import the tags
> from an input file. Have you ever seen a script that does this? Or is there
> an easier way?
>
> Thanks so much for your help, btw. Much appreciated! SS
>
>
>

michael
2004-10-20, 12:48
If you have a cddb record handy, you can just strip the comments out
and load the thing directly into the vorbis comments block. Does eac
have an option to spit out the raw cddb record? If not, parsing their
cuesheet into a vorbis comment block might not be a bad alternative.

Alternately, I personally use one of musicbrainz command line tools,
that I've modified the output to match what metaflac is expecting to
import. I personally think that musicbrainz generally has better
quality data than cddb, but this is a step separate from whatever
you're using to rip the disc, and I don't know if there's even a
windows equivalent to this utility.

-michael

David Lutz <lutzd (AT) pobox (DOT) com> writes:
> I have a very simple python script that I use. I always seem to
> need to hand edit what CDDB returns, so this script may not be
> complete enough for you, but it should be a start.
>
> D.L.
>
>
> On Mon, Oct 18, 2004 at 03:41:46PM +0000, Stephen Sweeney wrote:
>> michael <michael@...> writes:
>> >
>> > Stephen Sweeney <ssweeney@...> writes:
>> >
>> > > I think we should be able to encode the album and artist in vorbis with an
>> > > external cue sheet. Otherwise, only the filename identifies the album,
>> > > which is a little scary. What do you think?
>> >
>> > That's one reason I stick with internal cuesheets. No way for them to
>> > get separated from their codependent file. Others have chosen to do
>> > the opposite, and as we've discovered, the grey areas in between are
>> > still in need of some polishing.
>> >
>> Getting the cue sheet into .flac is no problem. It's getting the metadata
>> tags into vorbis. All (or almost all) the info is written by EAC to the cue
>> sheet. So if there were a script to parse the cue sheet into the format that
>> could be loaded into vorbis tags, I'd do it. Metaflac can import the tags
>> from an input file. Have you ever seen a script that does this? Or is there
>> an easier way?

Stephen Sweeney
2004-10-21, 12:49
michael <michael@...> writes:

>
> If you have a cddb record handy, you can just strip the comments out
> and load the thing directly into the vorbis comments block. Does eac
> have an option to spit out the raw cddb record?

I don't know but I'll check. I see that you can save cd information in a
local freedb. I'm not sure though how to get the info out as a raw cddb
record.

> If not, parsing their
> cuesheet into a vorbis comment block might not be a bad alternative.

Foobar2000 is able to parse a vorbis comment that contains the entire EAC cue
sheet. Pretty slick. So the cue sheet is in a single tag named "cuesheet".
If slimserver could parse this correctly then I'd be golden.
>
> Alternately, I personally use one of musicbrainz command line tools,
> that I've modified the output to match what metaflac is expecting to
> import.

What is / are the layout(s) that slimserver can handle? Which perl module
handles the parsing so I can look deeper?

michael
2004-10-21, 16:18
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:

> Foobar2000 is able to parse a vorbis comment that contains the entire EAC cue
> sheet. Pretty slick. So the cue sheet is in a single tag named "cuesheet".
> If slimserver could parse this correctly then I'd be golden.

Can you send me one?
Part of the current slimserver parsing is supposed to handle that, but
since I don't have a windows machine to run foobar2k on, I haven't
really had a good example to test with.

-Thanks,

-michael

michael
2004-10-21, 16:24
Stephen Sweeney <ssweeney (AT) houston (DOT) rr.com> writes:
....
> What is / are the layout(s) that slimserver can handle? Which perl module
> handles the parsing so I can look deeper?

There are currently four formats that it handles, and they're found
towards the end of slim/server/Slim/Formats/FLAC.pm

If anyone finds a program that deals with this in a way slim doesn't
already handle, please let me know.

Thanks,

-michael

Stephen Sweeney
2004-10-22, 09:16
michael <michael@...> writes:

>
> Stephen Sweeney <ssweeney@...> writes:
>
> > Foobar2000 is able to parse a vorbis comment that contains the entire EAC
cue
> > sheet. Pretty slick. So the cue sheet is in a single tag
named "cuesheet".
> > If slimserver could parse this correctly then I'd be golden.
>
> Can you send me one?
> Part of the current slimserver parsing is supposed to handle that, but
> since I don't have a windows machine to run foobar2k on, I haven't
> really had a good example to test with.
>
I've sent you two files directly. If they aren't what you want/need just let
me know.

One is the cue sheet from EAC (.cue). The other is a tag export (using
metaflac --export-tags-to=) to show how the cuesheet is being stored as a
multi-line tag. Metaflac doesn't currently support importing multi-line
tags. But there is a version of tag.exe that does. http://synthetic-
soul.mysite.wanadoo-members.co.uk/tag.html

I looked through the code. It doesn't appear that anything would handle a
cuesheet as a vorbis tag. But it doesn't look to be difficult. Much of the
logic is already in slimserver routine for parsing external cue sheets. I
think the trick is getting the single cuesheet tag into an array. The line
breaks WITHIN the tag show as a ^M through VIM - Win32.

Ed Atlee
2004-10-22, 09:40
"Stephen Sweeney" <ssweeney (AT) houston (DOT) rr.com> wrote
in message news:loom.20041018T031923-979 (AT) post (DOT) gmane.org...
> michael <michael@...> writes:
> > Stephen Sweeney <ssweeney@...> writes:
> > >
> > > However, the entire album
> > > does show up in the track listing as a separate entry. Is this the
> > > expected / desired behavior?
> >
> > It's certainly not the desired behaviour.
> >
> I figured this out. I encoded the entire album with artist and album
vorbis
> tags. SlimServer picks these and lists the .flac it as a playable entry
in
> browse by artist or browse by album. Removing these tags "fixes" the
> problem.

I have tried this, and I can't seem to make the extra (whole album) entry go
away, even after removing all tags from the .flac with foobar2k and
confirming with metaflac --list, then rescanning. Is there anything else
you did that I might be missing?

My other observations seem to match what you've said further on in this
thread: from what I've seen, SlimServer ignores an internal cuesheet in a
Vorbis comment block. It works well with an external cue sheet, except for
the extra listing that appears for the .flac. The only thing I haven't
tested is an internal cue via FLAC tags (flac --cuesheet option), but since
that does not include song titles, I would think that would be a less than
optimal solution - I don't think I'd want the Squeezebox displaying "Track
01" instead of the song title - unless SlimServer would somehow grab song
titles from other tags.

In the long term, I think the ideal would be if SlimServer would be able to
work with external or internal-via-Vorbis-tag cuesheets, but also to have an
option to ignore one or the other, to avoid duplicate listings. That seems
to be what you've acheived now (though which one is used is not
controllable), but I haven't been able to duplicate your results.

Ed Atlee
x84HurstOlds at hotmail dot com

Ed Atlee
2004-10-25, 14:01
"Ed Atlee" <x84HurstOlds (AT) hotmail (DOT) com> wrote in
message news:clbdc6$sgu$1 (AT) sea (DOT) gmane.org...
> I have tried this, and I can't seem to make the extra (whole album) entry
go
> away, even after removing all tags from the .flac with foobar2k and
> confirming with metaflac --list, then rescanning. Is there anything else
> you did that I might be missing?

I have figured this out, mostly. First, I got the wife to take care of the
teething infant, which greatly helped my focus ;-) Really, the next time I
started SlimServer, I tried browsing albums and things came up all screwy;
there would be 3 randon tracks from the album listed, along with 5 entries
that just gave the album name. I then did yet another rescan, and
everything's been fine sense. If I browse by album or artist (or genre)
now, I get the listings of the tracks from the ext. cuesheet under each
album. The entries for the flac files themselves go into a folder entitled
"No Album" or "No Artist" or whatever, which is good enough for me. IIRC,
using "Browse My Music Folder", I still get the double entries but I can
live with that. Not sure why this didn't work before, maybe some residual
screwiness from my earlier attempts to stick the year in the folder names -
without parens, they were getting interpreted as track numbers - DOH!

Of course I'd prefer to be able to have the flacs tagged and somehow tell
SlimServer to use *either* the tags *or* the cuesheet, but I'm not much good
with Perl so I'll take what I can get :-)

Ed Atlee
x84HurstOlds at hotmail dot com

JayNYC
2005-11-02, 23:40
is there a way to force the .flac file to NOT show in squeezebox, and instead only show the .cue?

or is there a way to say if .cue exists, don't show .flac?