Shoutcast charset (sc_trans)

[naitsirch]

Hi,

we are using sc_trans and shoutcast server. But the Problem is, if the MP3 files have special chars in the ID3 tags, they are removed.

For example: "schöner leben" will be displayed as "schner leben".

What could I do in this case? Which charset should I use?

Thanks for any help and best regards
______________
Christian

[DrO]

what version of the transcoder and shoutcast server are you using and also what OS are you running things on?

-daz

[naitsirch]

We are running both on Linux/Debian.

sc_trans version: 2.0 beta 5
Shoutcast version: DNAS/Linux v1.9.8 (Feb 28 2007)

Thanks for your fast reply

[DrO]

had a feeling the v1 DNAS was being used as i've found out it doesn't do things nicely from another user as you're seeing.

the only way to basically resolve the issue is to upgrade to use a v2 DNAS from what i've found from testing feedback. however i only fixed the v2 DNAS at the end of last week so there's no public version available but i have a test build i could let you use if you're able to upgrade to the newer version.

-daz

[naitsirch]

It's a pity, but we cannot upgrade to v2 DNAS, because it is still in beta

We'll use a workaround to solve this issue. We will transform special chars like "ö" to "oe".

But one of our customers said that the special chars are correctly displayed if the are streaming live with SAM Broadcaster. How is it possible?

[DrO]

It's a pity, but we cannot upgrade to v2 DNAS, because it is still in beta

not even in a test setup? as it would make sense to see if it will be ok when v2 isn't classified as beta (and i'd be happier to see if things are ok with the fix on other setups before the next build is released).

We'll use a workaround to solve this issue. We will transform special chars like "ö" to "oe".

that really isn't a solution and is just masking issues in the underlying tool.

But one of our customers said that the special chars are correctly displayed if the are streaming live with SAM Broadcaster. How is it possible?

all i know is that using a legacy source, the v1 dnas will most likely attempt to remove characters it cannot understand and different sources can send title information in different encoding (as there's no real rule on how it's meant to work) which most likely explains some of the difference.

with the v2 change i've made, it'll basically attempt to convert the legacy source title to utf8 which then should work ok (seems so from the tests done using the same characters as you've posted).

-daz