Deciding between image (png) and text (srt) subtitles in VF

[Adridulte]

Hello,

I'm a colorist, and often doing deliveries. Lately I have been scratching my head on the issue of subtitle types in VF DCPs.
I'm an advocate for image based subtitles delivered as VF DCP, which enables me to precisely control formatting, colors, positioning… If I were to choose I would always deliver one DCP with an additional VF with image based subtitles + a DCP with burned in subtitles. But sometimes for time or money reasons I'm only able to do the VF.
Ever since I'm doing this I've never had a problem and delivered to multiple large European film festivals. But not so long ago I stumbled across the technical guide of the Cannes Film festival which explicitly prohibit "sub-picture" image based subtitles.

https://www.marchedufilm.com/wp-content ... 23-eng.pdf

Since Cannes is undoubtedly a major actor in Film festivals I'm debating switching to only deliver text based subtitles from now on, from a convenience and compatibility standpoint. The other option would be delivering both formats (image and text) but then I don't know how I would differentiate the 2 VFs.

What are you thoughts on the matter?

Thanks,
Adrien

[IoannisSyrogiannis]

« ...but then I don't know how I would differentiate the 2 VFs. »

That is a question of naming the DCP. Often, one may find a version file numbered like so: _VF-1, _VF-2, etc.
Similarly, the feature (or other) is numbered. _FTR-3_, and so on.

Now, you know.

If I was you and was using an established subtitling program, I would put trust on it and export text. If and when problem was to rise, I would then go to .png. But then, again, whatever makes us sleep carefree is the best to do. (This is not, in any way, an incitation to eradicate the neighborhood's pet animals and wild life.)

Edit: When I am writing "text", I don't mean .srt file to turn into DCI .xml. I mean to export directly to DCI .xml, retaining all positioning info.

[Carsten]

In my opinion, the only reason to supply PNGs as subs is if you need to support glyphs that the timed text approach does not support properly. I know that e.g. japanese subtitles are known to cause issues for years, so that in Japan, they are usually supplied as PNG. Also, right-to-left textflow may cause issues.

There may be other reasons, e.g. when for creative reasons, a 'crafted' caption that needs to comply with the image underneath is needed.


If you actually start to supply timed text, burn-in, and PNG simultaneously you are making it worse instead of better.

If your avialble fonts supplies all the glyphs you need, stick to timed-text.

[Adridulte]

Thanks for the answer.
I’m using subtitle edit to export my PNGs so I will look into it for exporting timed text directly as xml. And still use PNGs when it is required.

For the naming VF-1 and VF-2 feels like a good way to being confused even with a written notice, once it’s in the hand of the client, projectionist.

[Carsten]

Well, ideally, even if they mix them up - they should both show more or less identical subtitles


Are you subtitling for a specific country/language?

[Adridulte]

French, Italian, English, German… I’ve used png subtitles for color changes and custom formatting to translate static credit plates.

[IoannisSyrogiannis]

[...]
For the naming VF-1 and VF-2 feels like a good way to being confused even with a written notice, once it’s in the hand of the client, projectionist.

First of all, if you are to have two version files for the same subtitle language, you don't need to deliver both, do you? If yes, then I see no big gain in having the two versions.

As far as confusion goes, I would give some more credit to the projectionists or whomever is in charge of putting the show on. And there is always the choice to create one of the two packages on another day, or with different version numbering where the format goes and so on.

I understand that you want to make everything fool proof, and I am for that, but up to a point. After all, what Douglas Adams wrote is correct:

A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools.

[Adridulte]

I mean, I’m not a distributor so I’m usually not the one handing things to the projectionist. Like you said I only need one version at a time, so if I’m creating multiples I have to justify and explain to the filmmaker/production the technicalities and reasoning behind the different versions.

I probably want to make things foolproof, but what I especially want is not having to go back and fixing things on the last minute because my DCP got rejected in a festival even though it works fine.