[Ltru] Tagging of silent films

Karen_Broome at spe.sony.com Karen_Broome at spe.sony.com
Wed Sep 28 20:09:32 CEST 2005

I actually agree with you. I've admitted that this is not a pure use of 
the language tag. Silent is not a language in any traditional sense -- 
hence the private tag. But for various reasons, we feel we need this as a 
language parameter in our industry. I am not pushing to register this tag 
and do not think it should be added to any language registry. 

We use the undetermined tag to indicate films where the language is truly 
undetermined. When we say silent, we are making a distinct classification 
that may have rights implications. And it is true that a lip-reader could 
determine the mute language that is being spoken.

We're still debating this, but we may even want separate values for "no 
linguistic content" or "no narrative language" vs. silent which has a 
specific meaning -- an actor mouthing words in advance of title cards to 
indicate what that person is saying. (These films were localized so it is 
possible that the title cards could be in a different language than what 
the mute actor is speaking. So we could potentially have two versions of 
the same film -- one that is "Silent-audio/French-text" and one that is 

Randy raised the question about what home video software would do upon 
finding the qsi tag. It's a point well-taken and I'll give it some 
thought, but I'm wondering what the software would need to do if the film 
is silent. Or is the suggestion that because it doesn't recognize the tag 
it would get a headache and crash?

Karen Broome

"Kent Karlsson" <kentk at cs.chalmers.se>
Sent by: ltru-bounces at ietf.org
09/28/2005 09:35 AM

        To:     ltru at ietf.org
        Subject:        Re: [Ltru] Tagging of silent films

> Well, this is a metadata standard, so part of the documentation of the
> standard
> would be "This field contains an RFC 3066 tag; the private-use tag 'qsi'

Ok, so far.

> means 'silent film'."

qsi is supposedly a *language* tag. "Silent", "film", "silent film"
does not really fit in there. If qsi in that metadata standard is
to have any language relation with respect to this, interpreting it
as "no lingustic content" (for an audio stream; or in some cases for
the images stream) would be far better. Or are we to have a language
subtag (variant) for films that does not show anyone speaking (or any
imaged text) but there is a narrative? (Like for nature/wildlife films
and the like.) I think not.

Using language tags for 'silent film', 'film with narrative' and
such would not be a suitable use of *language* tags.

Using 'und' or 'qsi' (within that context) for "no [determinable]
linguistic content" (for text, audio and imagages separate) would be
fine. And I don't think 'und' is all that wrong for that. Maybe the
sound of the talk is so faint that you didn't hear it ;-). If the
metadata standard allows for a langauge tag to be optional, leaving
the language tag out for saying "no [determinable] linguistic content"
would be fine too. If you want to distinguish "no-one is heard speaking"
from "jibberish-A" (and from "jibberish-B" etc.), which may or may not
be in some language unknown to the taggers (or just artistic, like
the "Swedish chef" in the Muppet Show), would that be worthwhile?
I'm not so sure.

   /kent k

Ltru mailing list
Ltru at ietf.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.alvestrand.no/pipermail/ietf-languages/attachments/20050928/5710ac0f/attachment.html

More information about the Ietf-languages mailing list