| |
V 2.0.30, May 5 , 2003
- back
to old
video data handling, to fix endless "Cutting video"-Bug on
certain stations
V 2.0.29, April 20 , 2003
- BUG
fixed in marks.vdr handling, Version 2.0.28 withdrawn
V 2.0.28, April 20 , 2003
- modified
video data handling
- BUG
fixed concerning autocut with marks.vdr and multi-file input
V 2.0.27, April 14 , 2003
- added
"autocut" of vdr-recordings (using the cuelist from
marks.vdr)
- fixed
problem with dreambox recordings from "new" Premiere Direkt
channels
V 2.0.26, February 3 , 2003
- Win2K-problem
("unable to open audio file") fixed
- fixed
problem with certain dreambox recordings
V 2.0.25, December 28 , 2002
- BUGFIX
: fixed a bug in video buffer handling, which could cause
PVAStrumento to crash (thanks to Linda for pointig this out)
- IMPROVEMENT
in audio handling: in certain combinations of video and audio format,
audio frames could be "lost" (_periodic_ "missing audio
frames").
V 2.0.24.1, December 23 , 2002
- really only minor tweaking
in TS-handling:
with this sub-version PVAStrumento handles DreamBox TS files from Premiere
Direkt recordings, as well as transport streams recorded with the
"Topfield" recorder (tested with one recording which was kindly
given to me)
V 2.0.24, December 20 , 2002
- added
DVB-MPEG2-Transportstream (TS) to the input Formats. Tested with transport
streams from the "DreamBox", there is a good chance that it will
work with other TS also - please inform me about success (or failure)
- found the problem with
certain Linux-VDR recorded streams, that led PVAStrumento to believe that
it had an "illegal stream"
- changed the handling of an
audio format change (stereo to mono and vice versa, as can be observerd on
the ZDV.vision channels): A GOP in which the audio change occurs is now
completely thrown away. This should solve the problem with mono-recordings
having the stereo packets in the beginning (led to "mickey
mouse" sound with certain programs).
- NOTE : PVAStrumento
still extracts only the major audio from a recording. Though VDR
and certain hard-disk-receiver/recorders are able to record multiple
audio, this feature is not yet supported by PVAStrumento (working on it).
- NOTE to
DreamBox users : There
have been reports that PVAStrumento 2.0.24 will not work with recordings
from Premiere Direkt channels. As I do not have a DreamBox myself,
it might take a few days to confirm the reports and have the problem
fixed.
- NOTE
: There have been reports that PVAStrumento has problems
writing output on Win2K operating systems.
V 2.0.23a, October 13 , 2002
- BUGFIX :
cPVAS crashes on
WinXP fixed ; crash with certain input files on WinXP fixed
- NOTE : PVAStrumento
not yet tested on WinXP SP1.
V 2.0.23, September 27 , 2002
- NOTE : Dynamically
linked versions of PVAStrumento are no longer distributed, only statically
linked binaries (including all library routines) are available, because
older C-runtime libraries on some systems led to unexpected behaviour of
PVAStrumento.
- audio input handling has
been modified ; audio misalignements will no longer be reported in the
logfile
- PVACut had a problem with
PVAStrumento generated PVA files > 2 GB (seems like some kind of buffer
problem in PVACut) ; PVAStrumento output has been modified, so that PVACut
will now accepts generated PVA > 2GB
- file I/O has been changed
from raw I/O to windows functions
- individual info messages
for GOPs that are not DVD-compliant are suppressed, insted a summary is
given at the end
- when <demux>ing the
logfile directory defaulted to the input directory - which could lead to
problems when the input directory was read-only; this has been changed to
the output directory, as was already the case with <make ps> and
<make pva>
NOTE: since there is no output
directory when <scan>ning, PVAStrumento will try to write the
logfile in the input directory
- NOTE:
checked compatibility with DVD2SVCD ver. 1.0.9 build 2 and DVX
2.29b conversion GUIs ; both work with PVAStrumento 2.0.23
- BUGFIX: Some
self-generated PVA-files were not accepted by PVAStrumento
- BUGS, KNOWN:
- Still
no solution to the "RTL-Problem", see V 2.0.21; Anyone having the
same "effect" with other recordings ?
- Still no overlapping output on multi-file input.
V 2.0.22, September 9 , 2002
- BUGFIX:
Thanks to an alert user the input-overwriting bug was dicovered fast (see
BUG, SERIOUS in Version 2.0.21). This bug has been fixed. You may again
savely use jobfiles.
- added a safeguard against
inadvertently overwriting the input, in case a user enters the same name
for the output as for the input by mistake.
- tested PVAStrumento V
2.0.22 to run with DVD2SVCD. No need to use older versions with DVD2SVCD
anymore
- added an info message when
PVAStrumento detects a GOP where the GOP size does not conform to DVD
specs (i.e. more than 15 pictures per GOP). To see this message you have
to turn logging level at least down to info.
- BUGS, KNOWN: When
splitting the output in a muli-file input situation output overlapping
will not work. (Multi-file input splitting works fine without overlap
and overlap works fine with single file input).
V 2.0.21, September 6 , 2002
- Multi-file input: Multiple
input files handled as one input - comes handy when recordings have been
autosplit
- Extracting audio from
audio-only recordings through "demux": auto-detecting of radio
recordings
- The first GOP at file
start as well as GOPs dropped after a missing video frame, are now
"closed" by removing the B-frames referring to the previos GOP.
Therefore DVD2AVI will now report (correctly !) 0ms audio delay and the cuts on
dropped frames are now "clean" (i.e. no more pixelization)
- The job-file syntax has
been extended to cover multi-file input and the documentation has been
brought up-to-date
- The default for sequence
scanning is now "short scan", there should be no use for long
scan anymore in normal conditions.
- BUGS, KNOWN, SERIOUS !: When
generating a jobfile for conversion to MPEG2 PS this version generates a
job that overwrites the input ! Do not use jobfiles from this version !
- BUGS, KNOWN: RTL-Problem:
Process recordings from RTL to generate a (supposedly fixed) .pva or .mpg
output. Process this output a second time and although this output should
be fixed, PVAStrumento report missing frames. Strangely this seems to
happen only with recordings from the RTL group ?? To be investigated.
- BUGS, UNKNOWN: Probably
a lot.
V 2.0.20, July 17, 2002
- BUGFIX:
cPVAS would crash on WinXP. Fixed.
- Added a "scan"
command in the GUI version: Scans a file and reports errors, does not
produce output.
- Added a "run
batch" command to GUI version, so GUI can now run pre-created batches
- Additional JOBs can now
appended to a batch file
- Output working directory
is now remembered between sessions
- BUGS,
KNOWN: Helpfile
needs updating, SCR calculation needs to be re-done to be able to directly
write an SVCD !
- BUGS,
UNKNOWN: Probably
a lot.
V 2.0.19, April 19, 2002
-
BUGFIX: In PVA generation, file splitting would not work.
Fixed.
-
BUGFIX: The change in SCR calculation would cause
"stuttering" on Software DVD-Players. SCRs are now contantly set
to Zero, which is wrong and probably causes stand-alone DVD-Players to
choke. But at least MPEG2 PSes run smoothly with Software DVD-Players
again. The SCR calculation funnily enough seems to be a problem- againl: do not expect PVAStrumento generated
MPEG2 PS to run directly on your DVD-Player !
-
BUGFIX: Fixed a bug whereby an EOF found during a long
sequence scan could cause PVAStrumento to terminate the file early.
V 2.0.18, April 15, 2002
- URGENT
BUGFIX Release: there was an ugly bug in PVA-generation, which
had to be fixed quickly !
- Changed the SCR
calculation in MPEG2 PS generation. But it is still
not quite right, so still: do not expect PVAStrumento generated
MPEG2 PS to run directly on your DVD-Player !
V 2.0.17, April 14, 2002
- PVAStrumento can now
generate PVA files as output (from either MPEG or PVA input). So you can
use the great PVACut program even if your original file was an MPEG2 PS.
Just run it once through PVAStrumento, generate a PVA, cut this with PVACut
and let the generated PVA run through PVAStrumento again for demuxing or
generating a sync MPEG2 PS.
- There is no option to set
the GOP PTS anymore. DVD2AVI will again announce a (negative) audio delay.
When muxing the original audio and video that has been processed through
DVD2AVI you have to honour that audio delay (due to the way DVD2AVI
handels un-closed GOPs).
- Again tried to enhance
44.1 kHz audio handling
- NOT FIXED: Still haven't
found the reason why the command shell version will crash on certain
installations
- POSSIBLY NOT FIXED:
Handling of MD8.1b recordings, tried a different approach on analysing the
stream data, but not sure, if it works, because I have no test file
available.
- NOT FIXED: The
documentation generally still reflects version 2.0.16!
V 2.0.16, Feb 24, 2002
- changed syncing behaviour
in case a GOP would not cover an integer number of audio frames,
"long scan" option not "really" nedded anymore. This
results in a much better behaviour with 44.1 kHz audio.
- found and fixed a bug when
the first byte of an audio header would be the last byte of a packet,
leading to a better behaviour with streams (MPEG2 PS) captured with
Hauppauge PVR PCI and PVR PCI II boards.
- fixed a bug in the
"split on audio format change" handling
- added an option to set the
GOP PTS to either the first I-Frame (will give an DVD2AVI audio delay) or
to the first frame in time (DVD2AVI will not report an audio delay). This
option does not change the picture PTSes and will have no effect on
demuxing and on the data contents of the stream. It just has an effect
on what DVD2AVI will believe the audio offset is.
- updated the documentation
to reflect the current version of the program
V 2.0.15, Feb 5, 2002
- internal use only, not
published
V 2.0.14, January 31, 2002
- added an option to split
the output when a change in audio format occurs ("Audio tab" -
not yet documented)
- corrected the handling of
NTSC-recordings (framerate 29.97 fps) and improved handling of unusual
GOPs ("No integer A/V frame match" situation).
- checked PVAStrumento usage
for Hauppauge's WinTV PVR recordings - which have the usual problems
keeping synch. PVAStrumento does its duty here as well.
V 2.0.13, January 25, 2002
- checked handling of MPEG2
PS generated with Hauppauge WinTV PVR - tweaked MPEG2 multiplexing
("make ps") a bit to better handle these MPEG2 PS. Checked that
PVAStrumento works as synchronising tool in the toolchain WinTV PVR ->
PVAStrumento -> Ulead DVD factory
V 2.0.12, January 20, 2001
- bugfix in audio frame
detection, where PVAStrumento would incorrectly detect "short audio
frames" on some recordings from ARD channels
V 2.0.11, November 20, 2001
- fixed problem, when
satellite stations would change audio format during transmission (aka
"ZDF-Problem")
V 2.0.10, date September 14,
2001
The manual has not
yet been updated with the new options !
- fixed an audio bug that
was triggered by a change of streaming in the ARD channels - this bug was
introduced in Version 2.0.5. PVAStrumento should now also better handle
certain MPEG PS files.
- added an option in the
"experts" tab and a command to cPVAS jobfiles to drop GOPs with
a wrong timestructure or to keep them. The dropping of GOPs with a
corrupted time structure had been introduced "silently" in
version 2.0.9 in order to make streams compliant with certain DVD
authoring software.
- added an option to the
"video" tab a command to cPVAS jobfiles to adjust the GOP
timecodes to correspond to the PTS times, this had been a feature of the
1.x-Versions of PVAStrumento and was thought to be not needed. But
experience tells otherwise.
- checked execution of
PVAStrumento 2.0.10 on Windows XP, RC1
V 2.0.9, date August 14, 2001
- changes in MPEG2 PS input
handling so that PVAStrumento now can not only digest, correct, and demux
DVB-S files but also used to transform MPEG2 recordings with DAZZLE
DVCII (also known as FAST DVD.master) to something useful
- minor bugfixes:
changed the initial setting (on first start of PVAStrumento) of "PTS
jitter" to 1000 microseconds, to conform to "Set Defaults"
fixed the background bitmap which somehow got junked in Version 2.0.8
loglevel Parameter was ignored in cPVAS (commandshell version) - this is
now fixed
- I checked that output from
PVAStrumento 2.0.8 and 2.0.9 can be loaded into TMPGENC Beta12i
V 2.0.8, date August 10, 2001
- Version 2.0.7 had an error
which caused it to crash with certain streams
- at least one TV station
(SAT.1) now transmits audio sequences where the presentation time for
audio frames varies non-linear with the number of frames. For these
streams V 2.0.7 introduced an echo effect in the output. This has been
fixed in Version 2.0.8
V 2.0.7, date August 9,
2001
- more finetuning of audio
frame detection/correction: fixing a problem with new Technotrend Software
(2.06) that caused spurios "audio misalignment" errors
- added an option (in
"expert" tab) to do a less stringent audio PTS checking. This is
necessary for DVD.master generated files and also for some DVB files. Due
to "unorthodox" timestamps in audio, you would encounter
"missing audio frams" error, where there actually were no audio
frames missing.
V 2.0.6, date August 1,
2001
- fixed the bug when playing
generated MPEG PS through DirectShow (MediaPlayer etc. would only play
audio)
- when file splitting, you
can now specify to split (almost) evenly
V 2.0.5, date July 5,
2001
- in 2.03, 2.04 and
-finally- 2.05 the adio frame detection has been improved
- when file splitting, you
can now have an optional overlap
- added MPEG PS input (in
addition to .PVA input)
- numerous bugfixes
V 2.0.2, date Mar. 22,
2001
- found another
buffer-overflow that would crash the program on Win2K
- added an option to write
packets of unconstrained size in MPEG PS. That way, if you only use the PS
on the computer, you gain about 10% in file size (due to the missing
padding stream)
- added an option to set the
size of a long video scan
- the thread priority can be
changed
- input- and output-buffer
sizes can be changed through the registry
V 2.0.1, date Mar. 20,
2001
- fixed a bug that caused
PVAStrumento to crash with certain streams on Win2K
- changed the audio frame
handling a bit, should result in less "misligned audio" alerts
- changed file-I/O to raw
read/write. This slows down the program (by about 5%, measured with my
setup), but should allow for files > 4GB on Win2K now.
V 2.0.0 initial release
|
|
|
|