hw accelerated transcode (h264_cuvid -> h264_nvenc with -hwaccel cuvid) was
broken after the filtergraph initialization was changed to intialize decoder
first followed by encoder (commit af1761f7b5).
During initialzing encoder with bframes, local buffers are allocated
internally in encoder which fails since no cuda context is available. Now
pushing the correct cuda context before encoder initialization fixes the issue.
Also adding push/pop cuda ctx during create/destroy/map/unmap resources and
destroy encoder session.
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
This patch aims to reduce the number of input/output surfaces
NVENC allocates per session. Previous default sets allocated surfaces to 32
(unless there is user specified param or lookahead involved). Having large
number of surfaces consumes extra video memory (esp for higher resolution
encoding). The patch changes the surfaces calculation for default, B-frames,
lookahead scenario respectively.
The other change involves surface selection. Previously, if a session
allocates x surfaces, only x-1 surfaces are used (due to combination
of output delay and lock toggle logic). To prevent unused surfaces,
changing surface rotation to using predefined fifo.
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
qmin and qmax are not necessary for nvenc vbr.
Enforcing this constraint, doesn't allow user to use vbr 2 pass mode without explicity setting the qmin and qmax options
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
Round qpIntra and qpInter calculation instead of old floor behavior.
Adopted from vaapi_encode_h264.c
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
When input surfaces are cuda frames, we will not know what the actual
underlying format (nv12, p010, etc) is at surface allocation time.
On the other hand, we will know when the input frames are actually
registered and associated with a surface.
So, let's delay format discovery until registration time, which is
actually how we handle other frame properties, such as dimensions.
By itself, this change doesn't allow for transcoding of 10bit
content from cuvid, but it reduces the problem to the hardcoding of
the sw format in ffmpeg_cuvid.c
Signed-off-by: Philip Langdale <philipl@overt.org>
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
This dubious behaviour in nvenc was finally removed by nvidia, and
as we refuse to run on anything older than 7.0, we don't need to
keep it around for old versions.
User selectable surfaces are not working correctly, if you set number of
surfaces on cmdline, it will always use minimum 32 or 48 depends on
selected resolution, but in nvenc it is not necessary to use so many
surfaces.
So from now you can define as low as 1 surface and nvenc will still
work, it will ofcourse lower GPU memory usage by 95% and async_delay to zero
That was the easy part, now littlebit more...
Next part of this patch is to always prefer rc_lookahead to be more
important for number of surfaces, than user defined surfaces value.
Maximum rc_lookahead from nvidia documentation is 32, but could increase
in future generations so there is no limit for this yet. Value
async_depth is still accepted and prefered over rc_lookahead.
There were also bug when you request more than rc_lookahead > 31, it
will always set maximum 31, because surface numbers recalculation was
after setting lookahead, which is now fixed.
Results:
If you set -rc_lookahead 32 and -bf 3 it will now use only 40 surfaces
and lower GPU memory usage by 20%, also it will now increase PSNR by 0.012dB
Two more comments:
1. from my internal test, i don't understand addition of 4 more surfaces
when lookahead is calculated, i didn't used this and everything works as
with those 4 more extra surfaces, does anybody know what is going on
there? I looks like it was used for B frames which are calculated
separately, because B frames maximum is 4.
2. rc_lookahead is defined default to -1, but in test condition if
(ctx->rc_lookahead) which sets lookahead it will be always true, i don't
know if this is intended behavior, so in default behavior is lookahead
always on!
This is default condition when rc_lokkahead is -1 (not defined on
cmdline), whis is maybe something that is not intended:
ctx->encode_config.rcParams.enableLookahead = 1;
ctx->encode_config.rcParams.lookaheadDepth = 0;
ctx->encode_config.rcParams.disableIadapt = 0;
ctx->encode_config.rcParams.disableBadapt = 0;
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
And use a macro to reduce the boilerplate.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
nvenc still encodes as yuv, but does the conversion internally which
brings some performance gains.
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
Ganapathy Kasi
Timo Rothenpieler
Sumit Agarwal
Ben Chang
Clément Bœsch
Konda Raju
Luca Barbato
Miroslav Slugen
Ruta Gadkari
Philip Langdale
Matt Oliver
Sven C. Dack
Yogender Gupta
James Almer