RE: [Ltru] Tagging of silent films (was: Re: el truth?)

At the risk of straying off topic...

If you are distinguishing between audio tracks and subtitles, then I
would think you could have a film in which there is no audio track, but
there are subtitles. For such a film, <audio lang="no linguistic
content"> and <subtitle lang="English"> are plausibly valid choices. But
I guess the case of silent films is special since there is no spoken
linguistic content in the audio track while there is written linguistic
content, only not in subtitles but in the video content itself.

So, in such a case, it would seem to me that that written content is in
some language such as English that should be specified - e.g., "The Gold
Rush" is definitely silent but also definitely in English - and that
separately you might have a *separate* metadata categorization to
indicate "silent" as a particular category of film, with implications
both wrt genre and date as well as wrt how dialogue and narration are
represented.

What do you do in modern cases in which a film presents prologue or
epilogue content in writing, such as "In a galaxy far, far away..."?

Peter

________________________________

From: ltru-bounces@ietf.org [mailto:ltru-bounces@ietf.org] On Behalf Of
Karen_Broome@spe.sony.com
Sent: Tuesday, September 27, 2005 10:29 AM
To: Harald Tveit Alvestrand
Cc: ltru-bounces@ietf.org; ltru@ietf.org
Subject: Re: [Ltru] Tagging of silent films (was: Re: el truth?)

I discussed whether to request a formal tag for this usage with the MPA.
We figured if we registered it, we would need to dilute the meaning to
make it more global and define it, as Peter suggests, as "no linguistic
content" or "no language present." 

But this is semantically different from "silent," which is very specific
to our industry. Consider the difference between a silent film --
defined as a film with an actor mouthing words that features title cards
to indicate what the actor is saying -- to films like Baraka,
Koyaanisqatsi, and the Triplets of Belleville, which feature no language
at all. (Belleville does have songs in French, but that's slightly
different.) Also, the Library of Congress genre list defines "actuality"
films as those early films that wowed viewers by merely showing a train
entering a station. These latter examples would fall under the "no
language present" classification, but not the "silent" classification. 

In other words, it's possible we may need both tags -- silent AND no
language present. 

I don't consider silent to be a modal classification -- spoken, written,
and signed are modes. And I'll even admit this is a somewhat impure use
of language data. (Another reason why I don't think this should be
registered.) Ideally, this metadata should be a boolean value. Silent or
not silent. But within our industry, for compliance with external
standards and product definition, we must define this as a language. 

I think und-silent is an oxymoron. If it's undetermined, how do you know
it's silent? You've determined that the film features an actor mouthing
the words, and you may have defined the title cards as being in French,
English, or Sio. That doesn't seem undetermined to me. (If you don't
mind, I'd prefer to gloss over the lipreaders issue.) 

At present, we have no plans to register this tag and prefer to use
"qsi" over "x-silent." I do think if the private use tags cannot be used
for this purpose, what good are they? Do you really want me to request
formal tags for both "silent" and "no language present"? That's my
potential business need. That could set a precedent for other weird tags
from other industries to enter the registry. Silent seems very
industry-specific to me and this seemed to be an appropriate use case
for the private use tag. 

Karen Broome
Metadata Systems Designer
Sony Pictures Entertainment
310.244.4384 

Harald Tveit Alvestrand <harald@alvestrand.no> 
Sent by: ltru-bounces@ietf.org 
09/27/2005 06:53 AM 

        To:        "John.Cowan" <jcowan@reutershealth.com> 
        cc:        ltru@ietf.org 
        Subject:        Re: [Ltru] Tagging of silent films (was: Re: el
truth?)

--On tirsdag, september 27, 2005 09:25:39 -0400 "John.Cowan" 
<jcowan@reutershealth.com> wrote:

> Harald Tveit Alvestrand scripsit:
>
>> However, they ARE guaranteed to conflict with other private use of
those
>> codes.
>>
>> Don't standardize private-use codes - not even in an "industry-wide
>> agreement". It hurts.
>
> I think that use of private tags by a group is precisely what they are
> for.  If private tags can't even be used by private agreement (between
> film people), they are useless.

That logically follows, of course - consider what happens when the film 
people adopt "qsi" for silent movies, the car people adopt "qsi" for a 
"test" voice on the car guidance system, and you insert a disk into the
car 
entertainment system.....

Outside of use inside a single system/database, private tags are
useless, 
IMHO.

>> i-silent is no longer available for public registration, methinks.
>
> Technically it still is, since we are in the RFC 3066 regime.  In the
> RFC 3066bis regime, the equivalent would be simply "silent", a non-639
> language subtag.

I certainly wouldn't encourage creating more interim  registrations.....

>> und-silent would be a valid code, if "silent" is registered as a
variant
>> subtag with "und" as its prefix.
>
> Yes, that is registerable in both regimes.  It's a little paradoxical,
> though, since "und" represents ignorance on the part of the tagger, to
> qualify ignorance with something specific.  OTOH, people in silent
films
> are generally speaking in some language, but viewers don't know or
care
> what it is.

I'm told that deaf people often laugh in the most bizarre spots when 
watching silent movies - they are used to lip-reading the dialogue :-)

_______________________________________________
Ltru mailing list
Ltru@ietf.org
https://www1.ietf.org/mailman/listinfo/ltru