alter the position of SRT subtitles by a dummy character on a new line

[zdoe]

yes, i'm aware of:

position tag1 - {\pos(xxx,yyy)}
this is NG because the positions are not proportional so that when used with different image sizes the titles get thrown off

position tag2
{\an7} {\an8} {\an9}
{\an4} {\an5} {\an6}
{\an1} {\an2} {\an3}
this is better, but still NG as the raster it provides is too coarse

so i would like to do the same by doing something like this:

-----------------------------------------------
1
00:01:55,824 --> 00:02:01,663
How are you today, Dave?
- Can't complain
dummyChar
dummyChar
dummyChar

2
00:02:12,966 --> 00:02:15,761
The weather is gorgeous today

-----------------------------------------------

thus making the to-be-lifted -text block to be a 5-line one, instead of the standard 2.

regular spaces get ignored with most players i used, so i took an M-space from word to try this. which works on most everything, but on media player classic it displays garbage, and on pot-player (my current favorite) it, and the lines that contain it by itself, get ignored.

who knows what would be a dummy character that would work across all players out there?

[Baldrick]

Wouldn't be better to use ssa/ass subtitles that actually support positions/etc?

[zdoe]

well - perhaps - but i have no experience on how do those behave inside an mp4 or MKV container. and given i never dealt with ssa/ass, i suspect that universal support across different players has ways to go...

[El Heggunte]

...........

regular spaces get ignored with most players i used, so i took an M-space from word to try this.

........

who knows what would be a dummy character that would work across all players out there?

Please, define what is "an M-space from word"

Depending on your answer, I could suggest you an alternative (or not).

[zdoe]

microsoft word:
insert > symbols > special characters > M-space

and from word then saved as plain text UTF-8 txt and renamed to srt.

[El Heggunte]

microsoft word:
insert > symbols > special characters > M-space

and from word then saved as plain text UTF-8 txt and renamed to srt.

Thanks for answering

So my suggestion is, try the "well-known" NO-BREAK SPACE
(ALT + 0160 on the numeric keyboard, aka "00A0" in UNICODE)

http://en.wikipedia.org/wiki/Whitespace_character

It used to work well with VSFilter, at least.

Off-Topic: and IMHO, M$ Word is overkill for writing *plain text* There are many many many replacements for Notepad --- EmEditor, EditPlus, NoteTab, UltraEdit, Notepad++, etc etc etc.

[zdoe]

ok - your suggestion is better than what i'd done. with the NO-BREAK SPACE ("00A0" in UNICODE) character the positioning works on kmplayer / vlc player. on potplayer & media player classic those rows get ignored, but at least no garbage characters show up, as it did with the M-space on media player classic.

so - basically the quest continues to find a dummy that would work across all players, but maybe that's too tall an order!

word - sure, but overkill doesn't hurt, and i'm used to working with that. i liked notepad+++ with this stuff for a brief moment, but files out of it seemed to sometimes have lost the right character encoding information so i went back to word.

cheers!

[El Heggunte]

EDIT: I've just run a more-detailed test with VSFilter, and indeed you are right,
IF all of the dummy lines contain only the nbsp character, then VSFilter ignores them all
At least the last of the dummy lines must contain a "printable" character, in order to "fool" VSFilter.

[El Heggunte]

This workaround does work with xy-VSFilter ,

not sure whether it works in VLC or KMPlayer though

http://en.wikipedia.org/wiki/Zero-width_non-breaking_space

Requires the file to be in Unicode --- if Big-Endian, then ZWNBSP = "FEFF"
( I used a hex editor for tweaking, I don't want to check whether M$ Word can do the same )

Example:

1
00:00:02,500 --> 00:00:05,500
Line 1
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0xFEFF

2
00:00:08,048 --> 00:00:11,518
blah-blah-blah

EDIT:

1) it's safe to replace 0x00A0 with 0xFEFF

2) EditPlus can copy and paste the ZWNBSP character

[zdoe]

good stuff - cheers.

how do you get the above kind of view of the hex-text? i'm having trouble navigating the file in a hex-view, both in editplus, and another hex-editor i have installed.

yes - all my SRT' stuff is UTF-8, need it for the foreign characters. on this project umlauts and simplified chinese.

[El Heggunte]

how do you get the above kind of view of the hex-text? i'm having trouble navigating the file in a hex-view, both in editplus, and another hex-editor i have installed.

yes - all my SRT' stuff is UTF-8, need it for the foreign characters. on this project umlauts and simplified chinese.

If you mean the PSEUDO-code:

1
00:00:02,500 --> 00:00:05,500
Line 1
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0x00A0
0xFEFF

2
00:00:08,048 --> 00:00:11,518
blah-blah-blah

then you misunderstood me , apologies for not being clear enough ,
hopefully the image below explains it all =^.^=

[zdoe]

yes - i did mean the "pseudo" code in your previous post. - i'm not sure i'll be able to manage editing characters on a view like your last .PNG without messing it up.

but perhaps i CAN try to verify that i got the right ones if i get them there with word or a text editor.

[vhelp]

i'm still not sure what being tried. anyway. it you are "inserting" these codes, you should be able to do that part in excel.
paste the list in column c, write the filter function in column b, and filter/copy in column a, give or take whatever works for you.

[El Heggunte]

I've just tested Winword 2002 SP3, and yes, it correctly copies and pastes the sequence

"0xFEFF" + "CR&LF"