Date |
Date |
Build |
Notes |
v2.52 b01 |
15
Sept
2004 |
040915 |
Reworked parts of the MPEG-4 parsing routine. It was
failing to even identify some files as MPEG-4
altogether, and when it did it sometimes provided
erroneous information (the BVOP indicator in
particular). |
12
Sept
2004 |
040912 |
¤ |
Fixed exception violation ("crash") issues exhibited
by, in particular, v2.51b2, which often overran an
internal data buffer. The current version has been
"road-tested" with over 2000 media files1 |
¤ |
Added 16:9 anamorphic support (for target and/or
source) in the resizing information function. Also
added ability to select different target sizes for
"AVI" (i.e. 1:1 PAR) targets. |
¤ |
Reorganized the GUI layout. See
screenshot. |
¤ |
Added support for parsing information from the
optional MPEG-2 "sequence display extension". When
available, the "display" height and width are shown,
as well as the color space mapping matrix (which may
be useful only to a few video techno-geek users out
there ;) |
¤ |
Added support for identifying so-called "packed
bitstreams", an MPEG-4 construct sometimes used DivX
5.x and other MPEG-4 streams contained in an AVI
file. The information is displayed as text in the
"container" section; additionally a new indicator
labeled "NVOP", as these constructs are informally
called, with the other MPEG-4 indicators. |
¤ |
Finally "restored" the ability to find the details
of the codecs and other filters, as well as the
mediatypes involved, when the "MS" (Microsoft
DirectShow) render function is used (see
Figure 1 below). The word "restored" is in
reference to the fact that older v2.21 had similar
capability, albeit more limited than the information
now available in this version. |
.1
No exception violations occurred while examining any
of the test files, nor on the render tests performed
on a subset of them, except for this outstanding
issue: an exception appears to occur upon exiting
GSpot, if 1) a "GSpot video render" test was
previously performed, and 2) the test was
successful, and 3) the test involved a VFW type
codec. The exception is caused by an invalid pointer
supplied by the O/S, so it's not a simple fix; the
issue is currently being evaluated. |