PULLDOWN v0.99d - April 14, 2001

Features added:

-drop_frame [true, false] : Set drop_frame_flag for NTSC streams
-aspect_ratio [1:1, 4:3, 16:9, 2.21:1] : Change aspect ratio

The purpose of the aspect ratio flag should fairly obvious and hopefully useful to some.

The drop_frame_flag is not used during playpack; however, I have discovered that DVDMaestro and Scenarist use this flag to synchronize subtitles in 29.97fps streams.  If you set this flag to true then you can mux your subtitle streams from MaestroSBT or ScriptConverter without problems.  This flag is required for proper subtitle sync on progressive, hybrid or interlaced 29.97fps streams.  PAL, of course, does not have this problem.

Example usage for a 29.97 int or hybrid source stream:

pulldown source.m2v target.m2v -nopulldown -drop_frame true

Example usage for a 23.976fps source:

pulldown source.m2v target.m2v -drop_frame true

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

PULLDOWN v0.99c - April 17, 2001

Features added:

PULLDOWN now reports the max bitrate as well as calculating the average bitrate.  Please not that that this is the TRUE average bitrate - ie, it takes into account whether or not pulldown is being performed.

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

PULLDOWN v0.99b - April 14, 2001

Features added:

-nopulldown : pulldown will not insert the necessary flags for 2:3 pulldown
-framerate [23.976, 25, 29.97] (defaults to 29.97) : changes framerate
-norff : strips out all RFF flags
-tff [odd, even] : changes the field order
-notimecodes : does not recalculate timecodes for the new framerate (timecodes are recalculated by default)
-prog_seq [p, i] : marks the entire MPEG stream as progressive or interlaced (by default, PULLDOWN creates an interlaced sequence)
-prog_frames [p, i] : marks each frame as progressive or interalced (by default, PULLDOWN marks all frames with an RFF as interlaced)

Here are some examples of how these new flags might be useful:

To strip out all pulldown information and return a stream to its original 23.976fps progressive format:

pulldown.exe source.m2v target.m2v -nopulldown -norff -framerate 23.976 -tff odd -prog_seq p -prog_frames p

To change the field order to odd on a stream that you accidentally encoded as even:

pulldown.exe source.m2v target.m2v -nopulldown -tff odd

To perform 2:3 pulldown without setting the frame and stream types to interlaced (use this if you're using a Hollywood+ and you are having problems playing the newer streams generated by PULLDOWN):

pulldown.exe source.m2v target.m2v -prog_seq p -prog_frames p

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

PULLDOWN v0.2e - April 10, 2001

Features added:

- You can now specify the output filename (hi, Nihilo).  Doing PULLDOWN.EXE <filename.m2v> without an output filename will still produce pulldown.m2v, however.

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

PULLDOWN v0.2d - April 9, 2001

Fixed:

- Stupid bug that caused pulldown.exe to write files in 32k chunks.  This would cause certain muxing problems to fail in certain circumstances.

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

PULLDOWN v0.2c - Mar 6, 2001

Quite a bit has changed in this release.  PULLDOWN is now a lot more active in changing the stream's encoding settings.  As a result, you can probably royally screw up the runtime encoding settings in TMPGEnc/bbMPEG/CCESP/Whatever and PULLDOWN will adjust the stream to the proper framerate and encoding types.  Because of this, you can re-run PULLDOWN on your old m2v's that previous versions of PULLDOWN may have already screwed up. I made these changes because I believe that PULLDOWN was introducing subtle video/audio sync problems.  I have not had the time to test the release as much as I would like so if you find problems (or even if you don't) I'd appreciate it if you could post the results to Doom9's (www.doom9.net or go.to/doom9) CCESP discussion forum.  No, you can't e-mail me; I get enough email as it is.

Fixed:

- PULLDOWN frames with a temporal value of 0 would cause the pulldown sequence to shift by one.
- no more PTS/DTS underflows while muxing (use bbMPEG 1.24b17a for best results)

Features added:

- Made PULLDOWN act more like bbMPEG (hopefully this is a good thing).  Specifically:
- Time codes are now calculated (thanks, Brent!)
- Start sequences are flagged as interlaced
- Chroma 420 types are now calculated
- Fields are set as Interlaced or Progressive appropriately

Possibly still needed:

- VBV calculations (not sure if this is really necessary)
- Bitrate values (again, not sure if this is necessary or how to calculate it)