2013-09-30 19:37:13 -07:00
|
|
|
/******************************************************************************
|
|
|
|
|
Copyright (C) 2013 by Hugh Bailey <obs.jim@gmail.com>
|
|
|
|
|
|
|
|
|
|
This program is free software: you can redistribute it and/or modify
|
|
|
|
|
it under the terms of the GNU General Public License as published by
|
2013-12-02 21:24:38 -08:00
|
|
|
the Free Software Foundation, either version 2 of the License, or
|
2013-09-30 19:37:13 -07:00
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
|
GNU General Public License for more details.
|
|
|
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
|
|
|
along with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
|
******************************************************************************/
|
|
|
|
|
|
2013-10-14 04:21:15 -07:00
|
|
|
#pragma once
|
2013-09-30 19:37:13 -07:00
|
|
|
|
2014-02-14 14:13:36 -08:00
|
|
|
#include "media-io-defs.h"
|
2013-09-30 19:37:13 -07:00
|
|
|
|
|
|
|
|
#ifdef __cplusplus
|
|
|
|
|
extern "C" {
|
|
|
|
|
#endif
|
|
|
|
|
|
libobs: Redesign/optimize frame encoding handling
Previously, the design for the interaction between the encoder thread
and the graphics thread was that the encoder thread would signal to the
graphics thread when to start drawing each frame. The original idea
behind this was to prevent mutually cascading stalls of encoding or
graphics rendering (i.e., if rendering took too long, then encoding
would have to catch up, then rendering would have to catch up again, and
so on, cascading upon each other). The ultimate goal was to prevent
encoding from impacting graphics and vise versa.
However, eventually it was realized that there were some fundamental
flaws with this design.
1. Stray frame duplication. You could not guarantee that a frame would
render on time, so sometimes frames would unintentionally be lost if
there was any sort of minor hiccup or if the thread took too long to
be scheduled I'm guessing.
2. Frame timing in the rendering thread was less accurate. The only
place where frame timing was accurate was in the encoder thread, and
the graphics thread was at the whim of thread scheduling. On higher
end computers it was typically fine, but it was just generally not
guaranteed that a frame would be rendered when it was supposed to be
rendered.
So the solution (originally proposed by r1ch and paibox) is to instead
keep the encoding and graphics threads separate as usual, but instead of
the encoder thread controlling the graphics thread, the graphics thread
now controls the encoder thread. The encoder thread keeps a limited
cache of frames, then the graphics thread copies frames in to the cache
and increments a semaphore to schedule the encoder thread to encode that
data.
In the cache, each frame has an encode counter. If the frame cache is
full (e.g., the encoder taking too long to return frames), it will not
cache a new frame, but instead will just increment the counter on the
last frame in the cache to schedule that frame to encode again, ensuring
that frames are on time and reducing CPU usage by lowering video
complexity. If the graphics thread takes too long to render a frame,
then it will add that frame with the count value set to the total amount
of frames that were missed (actual legitimately duplicated frames).
Because the cache gives many frames of breathing room for the encoder to
encode frames, this design helps improve results especially when using
encoding presets that have higher complexity and CPU usage, minimizing
the risk of needlessly skipped or duplicated frames.
I also managed to sneak in what should be a bit of an optimization to
reduce copying of frame data, though how much of an optimization it
ultimately ends up being is debatable.
So to sum it up, this commit increases accuracy of frame timing,
completely removes stray frame duplication, gives better results for
higher complexity encoding presets, and potentially optimizes the frame
pipeline a tiny bit.
2014-12-31 01:53:13 -08:00
|
|
|
struct video_frame;
|
|
|
|
|
|
2014-02-18 12:37:56 -08:00
|
|
|
/* Base video output component. Use this to create a video output track. */
|
2013-09-30 19:37:13 -07:00
|
|
|
|
|
|
|
|
struct video_output;
|
2014-09-25 17:44:05 -07:00
|
|
|
typedef struct video_output video_t;
|
2013-09-30 19:37:13 -07:00
|
|
|
|
2013-10-31 10:28:47 -07:00
|
|
|
enum video_format {
|
2014-01-19 02:16:41 -08:00
|
|
|
VIDEO_FORMAT_NONE,
|
2013-10-25 10:25:28 -07:00
|
|
|
|
|
|
|
|
/* planar 420 format */
|
2013-10-26 14:32:06 -07:00
|
|
|
VIDEO_FORMAT_I420, /* three-plane */
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
VIDEO_FORMAT_NV12, /* two-plane, luma and packed chroma */
|
2013-10-25 10:25:28 -07:00
|
|
|
|
|
|
|
|
/* packed 422 formats */
|
|
|
|
|
VIDEO_FORMAT_YVYU,
|
|
|
|
|
VIDEO_FORMAT_YUY2, /* YUYV */
|
|
|
|
|
VIDEO_FORMAT_UYVY,
|
|
|
|
|
|
|
|
|
|
/* packed uncompressed formats */
|
2013-10-24 00:57:55 -07:00
|
|
|
VIDEO_FORMAT_RGBA,
|
|
|
|
|
VIDEO_FORMAT_BGRA,
|
|
|
|
|
VIDEO_FORMAT_BGRX,
|
|
|
|
|
};
|
|
|
|
|
|
2014-12-11 19:47:51 -08:00
|
|
|
enum video_colorspace {
|
|
|
|
|
VIDEO_CS_DEFAULT,
|
|
|
|
|
VIDEO_CS_601,
|
|
|
|
|
VIDEO_CS_709,
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
enum video_range_type {
|
|
|
|
|
VIDEO_RANGE_DEFAULT,
|
|
|
|
|
VIDEO_RANGE_PARTIAL,
|
|
|
|
|
VIDEO_RANGE_FULL
|
|
|
|
|
};
|
|
|
|
|
|
2014-02-18 12:37:56 -08:00
|
|
|
struct video_data {
|
Implement encoder interface (still preliminary)
- Implement OBS encoder interface. It was previously incomplete, but
now is reaching some level of completion, though probably should
still be considered preliminary.
I had originally implemented it so that encoders only have a 'reset'
function to reset their parameters, but I felt that having both a
'start' and 'stop' function would be useful.
Encoders are now assigned to a specific video/audio media output each
rather than implicitely assigned to the main obs video/audio
contexts. This allows separate encoder contexts that aren't
necessarily assigned to the main video/audio context (which is useful
for things such as recording specific sources). Will probably have
to do this for regular obs outputs as well.
When creating an encoder, you must now explicitely state whether that
encoder is an audio or video encoder.
Audio and video can optionally be automatically converted depending
on what the encoder specifies.
When something 'attaches' to an encoder, the first attachment starts
the encoder, and the encoder automatically attaches to the media
output context associated with it. Subsequent attachments won't have
the same effect, they will just start receiving the same encoder data
when the next keyframe plays (along with SEI if any). When detaching
from the encoder, the last detachment will fully stop the encoder and
detach the encoder from the media output context associated with the
encoder.
SEI must actually be exported separately; because new encoder
attachments may not always be at the beginning of the stream, the
first keyframe they get must have that SEI data in it. If the
encoder has SEI data, it needs only add one small function to simply
query that SEI data, and then that data will be handled automatically
by libobs for all subsequent encoder attachments.
- Implement x264 encoder plugin, move x264 files to separate plugin to
separate necessary dependencies.
- Change video/audio frame output structures to not use const
qualifiers to prevent issues with non-const function usage elsewhere.
This was an issue when writing the x264 encoder, as the x264 encoder
expects non-const frame data.
Change stagesurf_map to return a non-const data type to prevent this
as well.
- Change full range parameter of video scaler to be an enum rather than
boolean
2014-03-16 16:21:34 -07:00
|
|
|
uint8_t *data[MAX_AV_PLANES];
|
2014-02-14 14:13:36 -08:00
|
|
|
uint32_t linesize[MAX_AV_PLANES];
|
2013-10-31 10:28:47 -07:00
|
|
|
uint64_t timestamp;
|
2013-09-30 19:37:13 -07:00
|
|
|
};
|
|
|
|
|
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
struct video_output_info {
|
2013-10-31 10:28:47 -07:00
|
|
|
const char *name;
|
|
|
|
|
|
2014-01-19 02:16:41 -08:00
|
|
|
enum video_format format;
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
uint32_t fps_num;
|
|
|
|
|
uint32_t fps_den;
|
2013-10-31 10:28:47 -07:00
|
|
|
uint32_t width;
|
|
|
|
|
uint32_t height;
|
libobs: Redesign/optimize frame encoding handling
Previously, the design for the interaction between the encoder thread
and the graphics thread was that the encoder thread would signal to the
graphics thread when to start drawing each frame. The original idea
behind this was to prevent mutually cascading stalls of encoding or
graphics rendering (i.e., if rendering took too long, then encoding
would have to catch up, then rendering would have to catch up again, and
so on, cascading upon each other). The ultimate goal was to prevent
encoding from impacting graphics and vise versa.
However, eventually it was realized that there were some fundamental
flaws with this design.
1. Stray frame duplication. You could not guarantee that a frame would
render on time, so sometimes frames would unintentionally be lost if
there was any sort of minor hiccup or if the thread took too long to
be scheduled I'm guessing.
2. Frame timing in the rendering thread was less accurate. The only
place where frame timing was accurate was in the encoder thread, and
the graphics thread was at the whim of thread scheduling. On higher
end computers it was typically fine, but it was just generally not
guaranteed that a frame would be rendered when it was supposed to be
rendered.
So the solution (originally proposed by r1ch and paibox) is to instead
keep the encoding and graphics threads separate as usual, but instead of
the encoder thread controlling the graphics thread, the graphics thread
now controls the encoder thread. The encoder thread keeps a limited
cache of frames, then the graphics thread copies frames in to the cache
and increments a semaphore to schedule the encoder thread to encode that
data.
In the cache, each frame has an encode counter. If the frame cache is
full (e.g., the encoder taking too long to return frames), it will not
cache a new frame, but instead will just increment the counter on the
last frame in the cache to schedule that frame to encode again, ensuring
that frames are on time and reducing CPU usage by lowering video
complexity. If the graphics thread takes too long to render a frame,
then it will add that frame with the count value set to the total amount
of frames that were missed (actual legitimately duplicated frames).
Because the cache gives many frames of breathing room for the encoder to
encode frames, this design helps improve results especially when using
encoding presets that have higher complexity and CPU usage, minimizing
the risk of needlessly skipped or duplicated frames.
I also managed to sneak in what should be a bit of an optimization to
reduce copying of frame data, though how much of an optimization it
ultimately ends up being is debatable.
So to sum it up, this commit increases accuracy of frame timing,
completely removes stray frame duplication, gives better results for
higher complexity encoding presets, and potentially optimizes the frame
pipeline a tiny bit.
2014-12-31 01:53:13 -08:00
|
|
|
size_t cache_size;
|
2014-12-11 19:47:51 -08:00
|
|
|
|
|
|
|
|
enum video_colorspace colorspace;
|
|
|
|
|
enum video_range_type range;
|
2013-09-30 19:37:13 -07:00
|
|
|
};
|
|
|
|
|
|
2014-02-09 04:51:06 -08:00
|
|
|
static inline bool format_is_yuv(enum video_format format)
|
|
|
|
|
{
|
|
|
|
|
switch (format) {
|
|
|
|
|
case VIDEO_FORMAT_I420:
|
|
|
|
|
case VIDEO_FORMAT_NV12:
|
|
|
|
|
case VIDEO_FORMAT_YVYU:
|
|
|
|
|
case VIDEO_FORMAT_YUY2:
|
|
|
|
|
case VIDEO_FORMAT_UYVY:
|
|
|
|
|
return true;
|
|
|
|
|
case VIDEO_FORMAT_NONE:
|
|
|
|
|
case VIDEO_FORMAT_RGBA:
|
|
|
|
|
case VIDEO_FORMAT_BGRA:
|
|
|
|
|
case VIDEO_FORMAT_BGRX:
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2014-02-18 12:50:09 -08:00
|
|
|
enum video_scale_type {
|
Implement encoder interface (still preliminary)
- Implement OBS encoder interface. It was previously incomplete, but
now is reaching some level of completion, though probably should
still be considered preliminary.
I had originally implemented it so that encoders only have a 'reset'
function to reset their parameters, but I felt that having both a
'start' and 'stop' function would be useful.
Encoders are now assigned to a specific video/audio media output each
rather than implicitely assigned to the main obs video/audio
contexts. This allows separate encoder contexts that aren't
necessarily assigned to the main video/audio context (which is useful
for things such as recording specific sources). Will probably have
to do this for regular obs outputs as well.
When creating an encoder, you must now explicitely state whether that
encoder is an audio or video encoder.
Audio and video can optionally be automatically converted depending
on what the encoder specifies.
When something 'attaches' to an encoder, the first attachment starts
the encoder, and the encoder automatically attaches to the media
output context associated with it. Subsequent attachments won't have
the same effect, they will just start receiving the same encoder data
when the next keyframe plays (along with SEI if any). When detaching
from the encoder, the last detachment will fully stop the encoder and
detach the encoder from the media output context associated with the
encoder.
SEI must actually be exported separately; because new encoder
attachments may not always be at the beginning of the stream, the
first keyframe they get must have that SEI data in it. If the
encoder has SEI data, it needs only add one small function to simply
query that SEI data, and then that data will be handled automatically
by libobs for all subsequent encoder attachments.
- Implement x264 encoder plugin, move x264 files to separate plugin to
separate necessary dependencies.
- Change video/audio frame output structures to not use const
qualifiers to prevent issues with non-const function usage elsewhere.
This was an issue when writing the x264 encoder, as the x264 encoder
expects non-const frame data.
Change stagesurf_map to return a non-const data type to prevent this
as well.
- Change full range parameter of video scaler to be an enum rather than
boolean
2014-03-16 16:21:34 -07:00
|
|
|
VIDEO_SCALE_DEFAULT,
|
|
|
|
|
VIDEO_SCALE_POINT,
|
|
|
|
|
VIDEO_SCALE_FAST_BILINEAR,
|
|
|
|
|
VIDEO_SCALE_BILINEAR,
|
|
|
|
|
VIDEO_SCALE_BICUBIC,
|
2014-02-18 12:50:09 -08:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
struct video_scale_info {
|
|
|
|
|
enum video_format format;
|
|
|
|
|
uint32_t width;
|
|
|
|
|
uint32_t height;
|
Implement encoder interface (still preliminary)
- Implement OBS encoder interface. It was previously incomplete, but
now is reaching some level of completion, though probably should
still be considered preliminary.
I had originally implemented it so that encoders only have a 'reset'
function to reset their parameters, but I felt that having both a
'start' and 'stop' function would be useful.
Encoders are now assigned to a specific video/audio media output each
rather than implicitely assigned to the main obs video/audio
contexts. This allows separate encoder contexts that aren't
necessarily assigned to the main video/audio context (which is useful
for things such as recording specific sources). Will probably have
to do this for regular obs outputs as well.
When creating an encoder, you must now explicitely state whether that
encoder is an audio or video encoder.
Audio and video can optionally be automatically converted depending
on what the encoder specifies.
When something 'attaches' to an encoder, the first attachment starts
the encoder, and the encoder automatically attaches to the media
output context associated with it. Subsequent attachments won't have
the same effect, they will just start receiving the same encoder data
when the next keyframe plays (along with SEI if any). When detaching
from the encoder, the last detachment will fully stop the encoder and
detach the encoder from the media output context associated with the
encoder.
SEI must actually be exported separately; because new encoder
attachments may not always be at the beginning of the stream, the
first keyframe they get must have that SEI data in it. If the
encoder has SEI data, it needs only add one small function to simply
query that SEI data, and then that data will be handled automatically
by libobs for all subsequent encoder attachments.
- Implement x264 encoder plugin, move x264 files to separate plugin to
separate necessary dependencies.
- Change video/audio frame output structures to not use const
qualifiers to prevent issues with non-const function usage elsewhere.
This was an issue when writing the x264 encoder, as the x264 encoder
expects non-const frame data.
Change stagesurf_map to return a non-const data type to prevent this
as well.
- Change full range parameter of video scaler to be an enum rather than
boolean
2014-03-16 16:21:34 -07:00
|
|
|
enum video_range_type range;
|
2014-02-18 12:50:09 -08:00
|
|
|
enum video_colorspace colorspace;
|
|
|
|
|
};
|
|
|
|
|
|
2014-04-15 08:23:38 -07:00
|
|
|
EXPORT enum video_format video_format_from_fourcc(uint32_t fourcc);
|
|
|
|
|
|
2014-04-23 15:25:39 -07:00
|
|
|
EXPORT bool video_format_get_parameters(enum video_colorspace color_space,
|
|
|
|
|
enum video_range_type range, float matrix[16],
|
|
|
|
|
float min_range[3], float max_range[3]);
|
|
|
|
|
|
2013-09-30 19:37:13 -07:00
|
|
|
#define VIDEO_OUTPUT_SUCCESS 0
|
|
|
|
|
#define VIDEO_OUTPUT_INVALIDPARAM -1
|
|
|
|
|
#define VIDEO_OUTPUT_FAIL -2
|
|
|
|
|
|
2014-09-25 17:44:05 -07:00
|
|
|
EXPORT int video_output_open(video_t **video, struct video_output_info *info);
|
|
|
|
|
EXPORT void video_output_close(video_t *video);
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
|
2014-09-25 17:44:05 -07:00
|
|
|
EXPORT bool video_output_connect(video_t *video,
|
2014-02-27 22:14:03 -08:00
|
|
|
const struct video_scale_info *conversion,
|
Implement encoder interface (still preliminary)
- Implement OBS encoder interface. It was previously incomplete, but
now is reaching some level of completion, though probably should
still be considered preliminary.
I had originally implemented it so that encoders only have a 'reset'
function to reset their parameters, but I felt that having both a
'start' and 'stop' function would be useful.
Encoders are now assigned to a specific video/audio media output each
rather than implicitely assigned to the main obs video/audio
contexts. This allows separate encoder contexts that aren't
necessarily assigned to the main video/audio context (which is useful
for things such as recording specific sources). Will probably have
to do this for regular obs outputs as well.
When creating an encoder, you must now explicitely state whether that
encoder is an audio or video encoder.
Audio and video can optionally be automatically converted depending
on what the encoder specifies.
When something 'attaches' to an encoder, the first attachment starts
the encoder, and the encoder automatically attaches to the media
output context associated with it. Subsequent attachments won't have
the same effect, they will just start receiving the same encoder data
when the next keyframe plays (along with SEI if any). When detaching
from the encoder, the last detachment will fully stop the encoder and
detach the encoder from the media output context associated with the
encoder.
SEI must actually be exported separately; because new encoder
attachments may not always be at the beginning of the stream, the
first keyframe they get must have that SEI data in it. If the
encoder has SEI data, it needs only add one small function to simply
query that SEI data, and then that data will be handled automatically
by libobs for all subsequent encoder attachments.
- Implement x264 encoder plugin, move x264 files to separate plugin to
separate necessary dependencies.
- Change video/audio frame output structures to not use const
qualifiers to prevent issues with non-const function usage elsewhere.
This was an issue when writing the x264 encoder, as the x264 encoder
expects non-const frame data.
Change stagesurf_map to return a non-const data type to prevent this
as well.
- Change full range parameter of video scaler to be an enum rather than
boolean
2014-03-16 16:21:34 -07:00
|
|
|
void (*callback)(void *param, struct video_data *frame),
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
void *param);
|
2014-09-25 17:44:05 -07:00
|
|
|
EXPORT void video_output_disconnect(video_t *video,
|
Implement encoder interface (still preliminary)
- Implement OBS encoder interface. It was previously incomplete, but
now is reaching some level of completion, though probably should
still be considered preliminary.
I had originally implemented it so that encoders only have a 'reset'
function to reset their parameters, but I felt that having both a
'start' and 'stop' function would be useful.
Encoders are now assigned to a specific video/audio media output each
rather than implicitely assigned to the main obs video/audio
contexts. This allows separate encoder contexts that aren't
necessarily assigned to the main video/audio context (which is useful
for things such as recording specific sources). Will probably have
to do this for regular obs outputs as well.
When creating an encoder, you must now explicitely state whether that
encoder is an audio or video encoder.
Audio and video can optionally be automatically converted depending
on what the encoder specifies.
When something 'attaches' to an encoder, the first attachment starts
the encoder, and the encoder automatically attaches to the media
output context associated with it. Subsequent attachments won't have
the same effect, they will just start receiving the same encoder data
when the next keyframe plays (along with SEI if any). When detaching
from the encoder, the last detachment will fully stop the encoder and
detach the encoder from the media output context associated with the
encoder.
SEI must actually be exported separately; because new encoder
attachments may not always be at the beginning of the stream, the
first keyframe they get must have that SEI data in it. If the
encoder has SEI data, it needs only add one small function to simply
query that SEI data, and then that data will be handled automatically
by libobs for all subsequent encoder attachments.
- Implement x264 encoder plugin, move x264 files to separate plugin to
separate necessary dependencies.
- Change video/audio frame output structures to not use const
qualifiers to prevent issues with non-const function usage elsewhere.
This was an issue when writing the x264 encoder, as the x264 encoder
expects non-const frame data.
Change stagesurf_map to return a non-const data type to prevent this
as well.
- Change full range parameter of video scaler to be an enum rather than
boolean
2014-03-16 16:21:34 -07:00
|
|
|
void (*callback)(void *param, struct video_data *frame),
|
Simplify media i/o interfaces
Completely revamped the entire media i/o data and handlers. The
original idea was to have a system that would have connecting media
inputs and outputs, but at a certain point I realized that this was an
unnecessary complexity for what we wanted to do. (Also, it reminded me
of directshow filters, and I HATE directshow with a passion, and
wouldn't wish it upon my greatest enemy)
Now, audio/video outputs are connected to directly, with better callback
handlers, and will eventually have the ability to automatically handle
conversions such as 4:4:4 to 4:2:0 when connecting to an input that uses
them. Doing this will allow the video/audio i/o handlers to also
prevent duplicate conversion, as well as make it easier/simple to use.
My true goal for this is to make output and encoder plugins as simple to
create as possible. I want to be able to be able to create an output
plugin with almost no real hassle of having to worry about image
conversions, media inputs/outputs, etc. A plugin developer shouldn't
have to handle that sort of stuff when he/she doesn't really need to.
Plugins will be able to simply create a callback via obs_video() and/or
obs_audio(), and they will automatically receive the audio/video data in
the formats requested via a simple callback, without needing to do
almost anything else at all.
2014-01-14 00:58:47 -08:00
|
|
|
void *param);
|
|
|
|
|
|
2014-09-26 15:25:59 -07:00
|
|
|
EXPORT bool video_output_active(const video_t *video);
|
2014-02-22 19:14:19 -08:00
|
|
|
|
2014-09-26 15:25:59 -07:00
|
|
|
EXPORT const struct video_output_info *video_output_get_info(
|
|
|
|
|
const video_t *video);
|
libobs: Redesign/optimize frame encoding handling
Previously, the design for the interaction between the encoder thread
and the graphics thread was that the encoder thread would signal to the
graphics thread when to start drawing each frame. The original idea
behind this was to prevent mutually cascading stalls of encoding or
graphics rendering (i.e., if rendering took too long, then encoding
would have to catch up, then rendering would have to catch up again, and
so on, cascading upon each other). The ultimate goal was to prevent
encoding from impacting graphics and vise versa.
However, eventually it was realized that there were some fundamental
flaws with this design.
1. Stray frame duplication. You could not guarantee that a frame would
render on time, so sometimes frames would unintentionally be lost if
there was any sort of minor hiccup or if the thread took too long to
be scheduled I'm guessing.
2. Frame timing in the rendering thread was less accurate. The only
place where frame timing was accurate was in the encoder thread, and
the graphics thread was at the whim of thread scheduling. On higher
end computers it was typically fine, but it was just generally not
guaranteed that a frame would be rendered when it was supposed to be
rendered.
So the solution (originally proposed by r1ch and paibox) is to instead
keep the encoding and graphics threads separate as usual, but instead of
the encoder thread controlling the graphics thread, the graphics thread
now controls the encoder thread. The encoder thread keeps a limited
cache of frames, then the graphics thread copies frames in to the cache
and increments a semaphore to schedule the encoder thread to encode that
data.
In the cache, each frame has an encode counter. If the frame cache is
full (e.g., the encoder taking too long to return frames), it will not
cache a new frame, but instead will just increment the counter on the
last frame in the cache to schedule that frame to encode again, ensuring
that frames are on time and reducing CPU usage by lowering video
complexity. If the graphics thread takes too long to render a frame,
then it will add that frame with the count value set to the total amount
of frames that were missed (actual legitimately duplicated frames).
Because the cache gives many frames of breathing room for the encoder to
encode frames, this design helps improve results especially when using
encoding presets that have higher complexity and CPU usage, minimizing
the risk of needlessly skipped or duplicated frames.
I also managed to sneak in what should be a bit of an optimization to
reduce copying of frame data, though how much of an optimization it
ultimately ends up being is debatable.
So to sum it up, this commit increases accuracy of frame timing,
completely removes stray frame duplication, gives better results for
higher complexity encoding presets, and potentially optimizes the frame
pipeline a tiny bit.
2014-12-31 01:53:13 -08:00
|
|
|
EXPORT bool video_output_lock_frame(video_t *video, struct video_frame *frame,
|
|
|
|
|
int count, uint64_t timestamp);
|
|
|
|
|
EXPORT void video_output_unlock_frame(video_t *video);
|
2014-09-26 15:25:59 -07:00
|
|
|
EXPORT uint64_t video_output_get_frame_time(const video_t *video);
|
2014-09-25 17:44:05 -07:00
|
|
|
EXPORT void video_output_stop(video_t *video);
|
libobs: Redesign/optimize frame encoding handling
Previously, the design for the interaction between the encoder thread
and the graphics thread was that the encoder thread would signal to the
graphics thread when to start drawing each frame. The original idea
behind this was to prevent mutually cascading stalls of encoding or
graphics rendering (i.e., if rendering took too long, then encoding
would have to catch up, then rendering would have to catch up again, and
so on, cascading upon each other). The ultimate goal was to prevent
encoding from impacting graphics and vise versa.
However, eventually it was realized that there were some fundamental
flaws with this design.
1. Stray frame duplication. You could not guarantee that a frame would
render on time, so sometimes frames would unintentionally be lost if
there was any sort of minor hiccup or if the thread took too long to
be scheduled I'm guessing.
2. Frame timing in the rendering thread was less accurate. The only
place where frame timing was accurate was in the encoder thread, and
the graphics thread was at the whim of thread scheduling. On higher
end computers it was typically fine, but it was just generally not
guaranteed that a frame would be rendered when it was supposed to be
rendered.
So the solution (originally proposed by r1ch and paibox) is to instead
keep the encoding and graphics threads separate as usual, but instead of
the encoder thread controlling the graphics thread, the graphics thread
now controls the encoder thread. The encoder thread keeps a limited
cache of frames, then the graphics thread copies frames in to the cache
and increments a semaphore to schedule the encoder thread to encode that
data.
In the cache, each frame has an encode counter. If the frame cache is
full (e.g., the encoder taking too long to return frames), it will not
cache a new frame, but instead will just increment the counter on the
last frame in the cache to schedule that frame to encode again, ensuring
that frames are on time and reducing CPU usage by lowering video
complexity. If the graphics thread takes too long to render a frame,
then it will add that frame with the count value set to the total amount
of frames that were missed (actual legitimately duplicated frames).
Because the cache gives many frames of breathing room for the encoder to
encode frames, this design helps improve results especially when using
encoding presets that have higher complexity and CPU usage, minimizing
the risk of needlessly skipped or duplicated frames.
I also managed to sneak in what should be a bit of an optimization to
reduce copying of frame data, though how much of an optimization it
ultimately ends up being is debatable.
So to sum it up, this commit increases accuracy of frame timing,
completely removes stray frame duplication, gives better results for
higher complexity encoding presets, and potentially optimizes the frame
pipeline a tiny bit.
2014-12-31 01:53:13 -08:00
|
|
|
EXPORT bool video_output_stopped(video_t *video);
|
2013-09-30 19:37:13 -07:00
|
|
|
|
2014-09-26 15:25:59 -07:00
|
|
|
EXPORT enum video_format video_output_get_format(const video_t *video);
|
|
|
|
|
EXPORT uint32_t video_output_get_width(const video_t *video);
|
|
|
|
|
EXPORT uint32_t video_output_get_height(const video_t *video);
|
|
|
|
|
EXPORT double video_output_get_frame_rate(const video_t *video);
|
2014-04-01 11:55:18 -07:00
|
|
|
|
2014-09-26 15:25:59 -07:00
|
|
|
EXPORT uint32_t video_output_get_skipped_frames(const video_t *video);
|
|
|
|
|
EXPORT uint32_t video_output_get_total_frames(const video_t *video);
|
2014-07-01 11:01:22 -07:00
|
|
|
|
2014-04-01 11:55:18 -07:00
|
|
|
|
2013-09-30 19:37:13 -07:00
|
|
|
#ifdef __cplusplus
|
|
|
|
|
}
|
|
|
|
|
#endif
|
