AfterDawn | News | Guides | Software downloads | Tech Support | Forums | HIGH.FI
AfterDawn.com

Version history for SubRip

<<Back to software description

Changes for v1.55.1 - v1.55.2

  • * (Small bugfix release)
  • - OCR engine: added '/' (slash) and '' (Spanish inverted exclamation mark) to the set of glyphs requiring extra-strong (temporary) char-diff-sensibility, to reduce possible confusion between slash (/) and italics-'l', and between 'i' and '' glyphs.
  • - Post-OCR I/l correction: added another filter for lower-case chars followed by an 'I'.
  • - Preview Options: removed .jpg file extension from the background image file requester (unsupported), avoiding further error messages down the road.
  • - Matrix files: removed all 'weak' colon (:) glyphs (with only the upper dot in the glyph image) from matrix # 4, 8, 11, 24, 25, 28, 29, 34, 38, 41, 42, 47, 50, 51, 60, 65, 66, 68, 73, 80, 88, 91 and 93, as they might be confused with the left (incomplete) part of a French '' (an 'i' with diaeresis). Thanks to Sven Bent for finding such a case.
  • Joined matrix #67 into #8.
  • Added 5 new matrix files and updated a few others.



Changes for v1.53.0 - v1.55.1

  • * (Small bugfix release)
  • - Post-OCR I/l Correction: fixed a few cases where an 'I' follows a hyphen (-) and is followed by certain other chars (affecting Czech, English, German, Polish and Spanish subtitles).
  • - Updated several matrix files, including:
  • - matrix #041: fixed two mistaken glyphs (a lower-case 'i' mistaken for a
  • Spanish inverted exclamation mark and another one with a wrong accent).
  • - joined 5 redundant matrix files and added 2 new ones.
  • I recommend to replace the entire collection, as some of the matrix numbers have changed to fill the gaps.



Changes for v1.52.0 - v1.53.0

  • + /AUTOTEXT CLI cmd: added --sortmatrix option to specify if and how to sort the matrix file before saving, if changed.
  • + Added /SORTMATRIX CLI cmd for [batch-]sorting existing/imported matrix files.
  • - CLI.txt: /FINDMATRIX: added --sortmatrix option to example script #2 and also fixed the script to correctly handle spaces within subtitle filenames.
  • - "Open VOB" window: fixed a minor bug where the FindMatrix status wasn't properly cleared after unsuccessfully trying to run a FindMatrix search with no .vob/.sub file(s) selected.
  • - "About" window: fixed a few layout issues. Used to look fine on WindowsXP as well as Wine but rather strange on Windows7 with certain settings/themes. Hopefully looking better now...



Changes for v1.51.1 - v1.52.0

  • + "Open VOB" window: added FindMatrix support, as already avaliable in CLI mode via /FINDMATRIX cmd. Essentially, this allows to use multiple small Matrix files (one Matrix per font), as an alternative to a large all-in-one Matrix.
  • + "Options" menu: added "FindMatrix Options" window.
  • + "Global Options" window: added an option to automatically save the character matrix after saving a subtitle, saving another mouse click.
  • + Added a few Help buttons here and there.
  • - punct.dic: fixed an issue where spaces after an opening parenthesis or bracket weren't filtered anymore (since 1.50b7). Oops..
  • - /FINDMATRIX: fixed a path problem for cases where was specified exactly as "..", where the result filename(s) were missing the leading Windows drive letter (although they did contain a fully resolved root path).
  • - SortMatrix: fixed a minor bug where deleted characters were showing up in the listview again after sorting (although they were still marked as deleted and were therefor still excluded from saving).
  • - "View/Edit Matrix" window: strongly raised the font size for the text edit widget, so one can actually recognize the typed character(s) a little better.



Changes for v1.50 Beta 7 - v1.51.1

  • 2014/12/16 SubRip 1.51.1 (c) Bloody
  • "Save As": fixed a bug where the Unicode output settings weren't always recognized correctly, which could also cause the "AutoSave after Post-OCR" function to hang (and other side effects).
  • 2014/12/15 SubRip 1.51.0 (c) Bloody
  • Text output files are now saved in UTF-8 format by default, which can be configured in the "Global Options" window, under "Text Output Options".
  • This was already the default for the /AUTOTEXT CLI command; now the GUI interface is doing the same.
  • "Subtitles" window: added support for loading UTF-8 encoded subtitles. The program will also no longer ask for a post-load conversion if the input file was encoded as UTF-16LE (or UTF-8 with BOM, for that matter).
  • "Global Options" window: added options to specify which output directory to use and how to proceed after saving (to save a couple of mouse clicks).
  • Note that these options must be manually enabled because having them enabled by default might confuse some people.
  • "Fast LoadMatrix" is now the general standard for CLI and GUI mode. The point is that as long as you don't need to view/edit the matrix, building the glyph list can be avoided/postponed until the "View/Edit Matrix" window is opened.
  • "View/Edit Matrix" window: added support for Matrix Sorting, used to optimize a matrix for higher subtitle conversion speeds, especially if the matrix is very large. For example, converting a subtitle with a matrix of 15000 glyphs can get twice as fast on average after sorting.
  • When conducting speed tests, see that the text auto-scrolldown checkbox in the Subtitles window is unchecked (generally faster if disabled).
  • Note: matrix sorting is focusing primarily on Latin characters (7-bit ASCII). For other languages (like Arabic, Chinese, Russian etc.), the speed gain is either little, none or even negative, depending on things like if the matrix also contains many italic-style glyphs, multi-char glyphs etc.
  • "Add new character(s)" window: added a checkbox for tagging certain glyphs as temporary 'junk' chars. This helps when sometimes only a part/fragment of a character is recognized, where "extend-right" sometimes won't solve the problem. Marking a character as temporary 'junk' allows to proceed as usual with that 'thing' to complete the subtitle, while also keeping such chars from 'contaminating' the matrix file when it is eventually saved to disk.
  • "Add new character(s)" window: added a 'quick' Paste button to insert text from the clipboard, making things a little easier when running SubRip together with an external Unicode character utility (like e.g. BabelMap, gucharmap, KCharSelect etc.). A checkbox near the Paste button can be used to enable Auto-OK after paste, saving another mouse click.
  • The Paste button will also filter the inserted text by removing all codes <=31 and 127 (stuff like Tabs, line breaks, Backspaces etc.) as well as any leading/trailing Spaces.
  • "Post-OCR Correction" window: added a checkbox to auto-open the Post-OCR window after subtitle conversion so it can't be forgotten, eliminating a nasty source of human error. Also saves another mouse click.
  • Also added a checkbox to proceed directly to the "Save As.." dialog after Post-OCR Correction, saving another mouse click.
  • /FINDMATRIX CLI command: added --tiny-match-percent option for dealing with very short subtitles, e.g. from short 'bonus' clips or tiny 'Forced' subtitles. This means that your scripts can now be simplified as they won't need to check for exit code 2 anymore (only used by older (beta) versions). This new option now takes care of such cases.
  • CLI.txt: /FINDMATRIX: enhanced example script 2 and added a few more comments regarding huge matrix files.
  • Application version scheme changed to "SubRip-{Major}.{Minor}.{Revision}".



Changes for v1.50 Beta 6 - v1.50 Beta 7

  • Fixed a bug where sometimes output files could not be saved (where the "save as" button did nothing in that case).
  • "View/Edit Matrix" window: fixed background colors for panels on the right side so they won't be influenced by certain Windows themes anymore.
  • "Save as Unicode": when using the GUI, the output filename in the file requester will now be pre-built in the same way as in the 'regular' requester ("Save as ANSI"). Avoids a bit of filename typing there.
  • Post-OCR I/l Correction: fixed a few more cases, like a lone I/l followed by a hyphen '-' or any I/l after a left parenthesis '(' or a left bracket '[' in certain cases.
  • punct.dic: added more fixes for the percent (%) character issue, for combinations of Slash/Degree chars where one of the two is marked as italic while the other is not, according to matching entries in the current character matrix.
  • CLI.txt: added a better usage example for finding 'double' matrix files using the /FINDMATRIX command, for people who prefer to go directly to the "Examples:" section without further reading about the available options.
  • "Wake me up": raised the timer interval from 10/10 to 15/20 seconds.



Changes for v1.50 Beta 4 - v1.50 Beta 6

  • 2014/28/11 SubRip 1.50 beta 6 (c) Bloody
  • This version is based on v1.50b5 by darkfrogger at: http://forum.doom9.org/showthread.php?t=160932
  • ..which is based on v1.50b4 by the original author(s) at: http://zuggy.wz.cz/
  • Install/update: note that the Dict/ folder now contains subdirs for internal and user-side .dic files, simply to draw a distinction between those two.
  • Added /AUTOTEXT command for auto-conversion from .bup|.idx|.ifo|.sub|.vob to .srt|.ssa|.ttxt without user interaction other than adding new glyphs to the character matrix.
  • For a quick usage example, see also: CLI.txt (under "Examples:").
  • Added /FINDMATRIX command for using/maintaining multiple matrix files.
  • For more information, see also: CLI.txt (under "/FINDMATRIX").
  • Improved Post-OCR Correction speed.
  • Fixed several issues with the Post-OCR I/l Correction. Although i had to tighten a few of these filters to eliminate certain false positives, the overall results have mostly improved.
  • Post-OCR Punctuation Correction: fixed an issue with mistakenly removed spaces before a dot (.) character that actually belongs to a numeric expression, like "A.30 calibre gun" or "only.5 millimeters".
  • punct.dic: added fixes for the Percent (%) character issue. Just enter what you see (Degree|Slash|Degree) and the punctuation filter will do the rest. If asked about a 2/3 part of a Percent char (Degree+Slash or Slash+Degree), then simply enter a % char.
  • Moved the orthography Post-OCR fix to the very end of the chain (this has nothing to do with "punct.dic" which is always executed during punctuation correction). The external (user-defined) dict filter should be the last resort if all other corrections didn't fix the problem. It also makes sense to apply fixes like "Format Whole Words Only" before applying the external dict, otherwise it might fail to recognize words that require fixing.
  • SubRip.ini: added missing item for Post-OCR: "Format Whole Words Only".
  • Fixed crashes caused by the Char Matrix Debugger and also disabled Post-OCR Correction if Char Matrix Debugging is enabled (makes no sense in that case). The output text file can still be saved in order to search for problems using a text editor.
  • Added all recent language files/updates posted on http://zuggy.wz.cz/ between 2006/11/07 and 2014/03/10.
  • Changed the main window timer interval from 0.5 to 0.05 seconds. This reduces the start/end delay for CLI processing as well as the delay between switching VOB files.
  • /AUTOTEXT: added debug options for some Post-OCR Correction filters.
  • "New Character(s)" window: fixed a few layout issues.
  • About: fixed mailto: syntax for mouse clicks on email links to correctly add the subject & body strings.
  • Added a *nix Makefile and a Make-Release script for building & packaging SubRip on Linux under Wine.
  • Moved Doc/gpl.txt to License.txt which is now part of both binary and src distributions.



<<Back to software description