Added unintentionally with IWYW but the older versions of
libavcoded didn't have the header (and had it included directly in
libavcodec/avcodec.h that is included later).
Until now, in verbose mode was printed the dynamically selected port
pair. But the knowledge should be useful also to ensure the ports where
it doesn't need to be clear, eg. in server mode or for receiver to see
that 5004 is indeed bound.
It didn't seem that R10k worked for anyone before 18th Sep (commit
3c9e2602) because bit shuffling is needed (r10k_to_sdl2()).
Also the negative meaning (=no), which has been added just in the last
week commit can be replaced with '--param decoder-use-codec='!R10k'`.
But set the padding as the maximum of AV_INPUT_BUFFER_PADDING_SIZE
(if there is include) and MAX_PADDING (both are currently 64).
Use __has_include() instead of HAVE_LIBAVCODEC_H.
Create inner loop with fixed amount of iterations (16). This will allow
the compiler to unroll the inner loop and vectorize (16 iterations per
4 bytes is 512b allowing up to 512b instructions).
The eventual rest (%16 != 0) is computed per pixel as it used to be..
In order to the initialize the option `driver=` successfully the video
subsystem must be called with the driver name (SDL_VideoInit), not with
SDL_Init[SubSystem].
This allows eg. `-d sdl:driver=KMSDRM` to be used.
Also use the driver in show_help (for `-d sdl:driver=KMSDRM:help`),
otherwise will SDL_Init() in help complained that "x11 not available".
refers to GH-412
This is a hack to check whether given invalid_argument message belongs
to stoi/stod. It is usually used to catch non-numeric user input where
a number was expected.
The GPU conversion should be better choice in general. The situation
where it can be useful is when the GPU is working hard to manage the
compression alone so that deploying on CPU saves some performance for
the encoder.
Set the pixfmt conversion CUDA kernel as cmpto_j2k_enc preprocessor,
not run directly.
This also eliminates to need to have the conversion kernel if conversion
is needed - CPU conversion will be sufficient. Currently not effective,
only R12L is converted for which there is the kernel.
refer to GH-406
Initialize the encoder stuff in configure_with, not _init.
It will allow the encoder parameter changes during the
reconfiguration. This will, however, require additional syncrhonization
and cleanup on the reconf.
There is no functional change in this commit but it changes the internal
behavior, namely the codec state is destroyed on reconfiguration and
newly created.
The data_len var used to compute packet_count may be lesser than its
initial value (tx->mtu - hdrs_len) because it might have been decreased
to number of remaining bytes of the last packet.
Returned back data_len to the do-loop and compute the value as
max_len.
For unsompreesed audio, where it has only the advantage, set default
R-S values to k=160 n=240 (50 % redundancy).
For uncompressed audio, the number of packets per frmae start at 3
(for 16-bit 30p), for which the value is appropriate. Not for the
audio captured at 60 Hz, however. Here it, is 1600 B with 16-bit samples,
with the R-S FEC this yields 2 packets only.
Actually the original value should have triggered the warning almost
always when there is RGB->YCbCr or YCbCr->RGB conversion. Seemingly it
doesn't happen often.
This reverts commit 4143900605.
Actaully, the transcoding reflector doesn't set the export module,
so that it will crash with this. Crasing command:
```
hd-rum-transcode 8M 5004 -c lavc -P 6004 100::
```
