Commit Graph

703 Commits (703e0e28a8e86de6a410b13cf32ff6f92162d5fb)

Author SHA1 Message Date
John Bradley 54fdd6555e mac-syphon: Fix NSURL selector compatibility 10.8
Use NSURL path.fileSystemRepresentation to be
compatibile with the 10.8 SDK. This appears to be
a convenience selector and should operate the same.
2015-03-25 19:01:47 -05:00
jp9000 96353332f3 win-capture: Fix game capture acquiring issue
If capture starts too quickly, the file mapping will return 2, which
means file not found, and it would then reset the capture and try again.
Sometimes this would result in long intervals where it wouldn't capture.
This fixes the issue by simply making game capture retry if file mapping
returns error number 2.
2015-03-25 14:14:43 -07:00
jp9000 2f6ba39b41 obs-filters: Add color key filter
Masks out a specific color range in RGB color space.  Commonly used with
solid image data or animated video rather than live video.
2015-03-25 14:14:39 -07:00
jp9000 adb392e08a obs-filters: Add chroma key filter
This allows applying a mask based upon the chroma value (U/V) of a
specific color in YUV color space.  Commonly used to mask out
greenscreens and bluescreens in live video.
2015-03-25 14:14:38 -07:00
jp9000 bee5b3a2d2 obs-filters: Add color correction filter
Adds a color correction filter if one wishes to just apply
gamma/constrast/brightness/opacity/color to a source.
2015-03-25 14:12:05 -07:00
jp9000 bd1ede883d obs-filters: Add crop filter
Allows any source to be cropped, though note that this renders to
texture first, so for more optimal results, cropping values should
probably be put in to capture sources that can be cropped as they're
actually rendered by the source.
2015-03-25 14:12:05 -07:00
jp9000 9138766620 obs-filters: Add mask/blend filter
This filter allows the ability to use a texture to modify the video
output of a source, the ability to:
- apply a color-based mask (dark = transparent, light = opaque)
- apply a mask based upon the alpha of an image
- blend an image via subtraction, addition, or multiplication
2015-03-25 14:12:04 -07:00
jp9000 55534d5d8d obs-filters: Add async delay filter
Adds a filter for delaying asynchronous video/audio data, for example
from webcams, video devices, or media playback sources.  Mainly intended
to allow users to sync up their webcams to other devices.
2015-03-25 14:12:03 -07:00
Gol-D-Ace f374c7198e rtmp-services: Update Vaughn Live / iNSTAGIB.tv ingests 2015-03-24 21:37:25 +01:00
kc5nra 946b9d01e3 obs-ffmpeg: Clearer meaning for ClearMediaOnEnd 2015-03-24 07:25:57 -05:00
kc5nra dd2d6494a1 obs-ffmpeg: Add clear on media end option
Clears source video when the media stream ends.
2015-03-24 00:21:24 -05:00
kc5nra c7ad555eae obs-ffmpeg: Handle null frames in frame callbacks
This is in preparation of an API change that will send null
frames to signal the end of the media being played.
2015-03-24 00:16:35 -05:00
kc5nra b091b03291 obs-ffmpeg: Split definition and assignment 2015-03-24 00:12:10 -05:00
kc5nra 54baa1a369 obs-ffmpeg: Change from sws_* cacheing to our own
Since sws_getCachedContext possibly logs a message every
time it is used (depending on the input), use our own caching
logic to minimize log spam.
2015-03-23 22:14:42 -05:00
jp9000 2e1eb4d7eb decklink: Fix log message 2015-03-22 20:54:18 -07:00
jp9000 716ef61cf9 decklink: Don't use error log if not supported
Using LOG_ERROR here isn't really necessary, it just means that there's
no blackmagic support.
2015-03-22 19:18:20 -07:00
jp9000 2d5c6dfae9 win-capture: Rename 'Monitor' to 'Display' capture
The name 'Monitor' is a bit inconsistent with what it's called by the
operating system.
2015-03-22 19:18:18 -07:00
jp9000 5247434d14 win-dshow: Change avermedia encoder names 2015-03-22 19:18:17 -07:00
jp9000 24053f4a9c win-dshow: Fix bad settings causing frozen video
If the settings are reset to defaults or if the settings are just bad,
the video would get stuck on the last frame that was displayed, which
feels a bit awkward.  Best to make it stop video output entirely rather
than get stuck on the last video frame.
2015-03-22 19:18:16 -07:00
jp9000 d5631dd1f8 win-dshow: Make devices active by default
Having devices not active by default may cause confusion for some users.
2015-03-22 19:18:16 -07:00
jp9000 f3104d92d5 mac-capture: If channels above 8, force to stereo
Certain devices (particularly certain mixers and soundflower 64ch) would
have an arbitrary number of channels, and wouldn't really be mappable to
a specific speaker layout supported by libobs.

So to fix this issue, if the channel count is above 8, force the data to
stereo to ensure playback can still occur, rather than cause it to just
fail.
2015-03-22 19:18:08 -07:00
jp9000 e20ec366b2 Make capture sources w/o alpha use opaque effect
This fixes an issue primarily with filter rendering: when capturing
windows and displays, their alpha channel is almost always 0, causing
the image to be completely invisible unintentionally.  The original fix
for this for many sources was just to turn off the blending, which would
be fine if you're not rendering any filters, but filters will render to
render targets first, and that lack of alpha will end up carrying over
in to the final image.

This doesn't apply to any mac captures because mac actually seems to set
the alpha channel to 1.
2015-03-22 19:18:05 -07:00
jp9000 233004631f linux-capture: Mark captures as custom drawn
The linux window and monitor captures draw in a custom way, therefore
they must be marked as types of sources that do custom rendering.
2015-03-22 18:38:49 -07:00
Jim 23bb999470 Merge pull request #376 from SkylerLipthay/feat/blackmagic-decklink-plugin
Blackmagic DeckLink plugin
2015-03-21 19:22:04 -07:00
Skyler Lipthay dd739d1e49 decklink: Add Windows DeckLink project
The code specific to Windows: helps convert `BSTR` instances to
`std::string`s; provides a Windows COM-specific implementation of
`CreateDeckLinkDiscoveryInstance`; aliases CFUUIDGetUUIDBytes,
CFUUIDBytes, and IUnknownUUID (the Linux SDK does this, but for some
reason the Windows SDK does not).
2015-03-21 16:43:16 -07:00
Skyler Lipthay b959f56291 decklink: Add Windows DeckLink SDK
Some changes were made to the stock DeckLink SDK: removed all references
to legacy API headers in DeckLinkAPI.idl; removed all instances of
`importlib("stdole2.tlb");`.
2015-03-21 16:43:00 -07:00
Skyler Lipthay 70f8369ae3 decklink: Add Linux DeckLink project
The code specific to Linux helps convert `const char *` instances to
`std::string`s.
2015-03-21 16:42:59 -07:00
Skyler Lipthay c9ff86dd01 decklink: Add Linux DeckLink SDK 2015-03-21 16:42:42 -07:00
Skyler Lipthay 496031e8c8 decklink: Add Mac DeckLink project
The code specific to Mac OS X helps convert `CFString` instances to
`std::string`s.
2015-03-21 16:42:40 -07:00
Skyler Lipthay fd83f5c85f decklink: Add Mac DeckLink SDK 2015-03-21 16:40:15 -07:00
Skyler Lipthay 4225fa8eee decklink: Add Blackmagic DeckLink capture plugin
This is a cross-platform Blackmagic device capture plugin that makes use
of the manufacturer provided DeckLink SDK.
2015-03-21 16:40:05 -07:00
Jim caa4e36704 Merge pull request #402 from Radzaquiel/patch-3
Update hitbox.tv ingest servers list
2015-03-21 09:51:37 -07:00
John Bradley d69f03c6a5 obs-ffmpeg: Use deferred update flag for properties 2015-03-19 16:02:24 -05:00
John Bradley c539c16ecc obs-ffmpeg: Instead of failing with unsupported CS, use default 2015-03-19 14:40:24 -05:00
John Bradley 52aae96f74 obs-ffmpeg: Add alias of CS601 support 2015-03-19 14:40:24 -05:00
John Bradley 17f8e3c58b obs-ffmpeg: Add frame drop controls to the ffmpeg source UI 2015-03-19 14:40:24 -05:00
Radzaquiel 3e1a521acb Update hitbox.tv ingest servers list
Added: US-West (San Francisco)
2015-03-19 11:18:52 +01:00
Radzaquiel 57c5c5ee52 Update hitbox.tv ingest servers list
Added: EU-South
2015-03-18 21:32:38 +01:00
Jim b4bca874b4 Merge pull request #400 from Gol-D-Ace/GoodGame.ru-ingests
rtmp-services: Update GoodGame.ru ingests
2015-03-16 15:26:48 -07:00
Gol-D-Ace 3421112ac9 rtmp-services: Update GoodGame.ru ingests 2015-03-16 23:14:43 +01:00
jp9000 9832a760b8 (API Change) Always use planar float audio output
Core API functions changed:
-----------------------------
EXPORT bool obs_reset_audio(struct audio_output_info *aoi);
EXPORT bool obs_get_audio_info(struct audio_output_info *aoi);

To:
-----------------------------
EXPORT bool obs_reset_audio(const struct obs_audio_info *oai);
EXPORT bool obs_get_audio_info(struct obs_audio_info *oai);

Core structure added:
-----------------------------
struct obs_audio_info {
	uint32_t            samples_per_sec;
	enum speaker_layout speakers;
	uint64_t            buffer_ms;
};

Non-interleaved (planar) floating point output is standard with audio
filtering, so to prevent audio filters from having to worry about
different audio format implementations and for the sake consistency
between user interfaces, make it so that audio is always set to
non-interleaved floating point output.
2015-03-12 22:22:02 -07:00
Gol-D-Ace a4b8114cab rtmp-services: Update Twitch Singapore ingest 2015-03-12 15:40:45 +01:00
John Bradley f23d958122 obs-ffmpeg: Cache picture buffer in the same way as sws context 2015-03-10 14:37:28 -05:00
John Bradley baea0f583a obs-ffmpeg: Remove extra curly braces 2015-03-10 14:34:37 -05:00
John Bradley a73c15ae16 obs-ffmpeg: Fix bug where update_sws_context result not checked 2015-03-10 14:33:21 -05:00
John Bradley e13bb8f479 obs-ffmpeg: Cache swscale context during frame callback
Some codecs don't report the correct dimensions until the first
frame is rendered.
2015-03-10 12:53:17 -05:00
John Bradley 0e944735e0 obs-ffmpeg: Fix error fetching properties from wrong pointer 2015-03-10 12:25:24 -05:00
Jim eb7d631694 Merge pull request #394 from kc5nra/ffmpeg-source-plugin
Ffmpeg source plugin
2015-03-09 19:41:59 -07:00
John Bradley 2fc6c4d725 obs-ffmpeg: Add FFMpeg input source 2015-03-09 21:22:32 -05:00
John Bradley 8a32345071 obs-ffmpeg: Add FFmpeg format -> obs format func 2015-03-09 21:22:32 -05:00
jp9000 b03eae57c6 (API Change) Fix "apply service settings" functions
API changed from:
------------------------
EXPORT void obs_service_apply_encoder_settings(obs_service_t *service,
		obs_encoder_t *video_encoder,
		obs_encoder_t *audio_encoder);

void obs_service_info::apply_encoder_settings(void *data
			obs_encoder_t *video_encoder,
			obs_encoder_t *audio_encoder);

To:
------------------------
EXPORT void obs_service_apply_encoder_settings(obs_service_t *service,
		obs_data_t *video_encoder_settings,
		obs_data_t *audio_encoder_settings);

void obs_service_info::apply_encoder_settings(void *data
			obs_data_t *video_encoder_settings,
			obs_data_t *audio_encoder_settings);

These changes make it so that instead of an encoder potentially being
updated more than once with different settings, that these functions
will be called for the specific settings being used, and the settings
will be updated according to what's required by the service.

This fixes that design flaw and ensures that there's no case where
obs_encoder_update is called where the settings might not have
service-specific settings applied.
2015-03-07 16:32:00 -08:00
jp9000 56f4dd5359 obs-outputs: Fix bug (headers not always sent)
If a disconnection occurs, the sent_headers would still be set to true,
and when a reconnect occurs, it would not send the headers again.
2015-03-07 10:11:18 -08:00
Christoph Hohmann c6008316e2 image-source: Add option (unload when not showing)
If this option is on, the image will unload when the image isn't
displayed anywhere, and then reload it when it's displayed again to
reduce the amount of memory images take up on VRAM at any given time.
If this option is off, then it's always loaded in VRAM regardless of
whether it's displayed or not.

Closes Pull Request #380

(edited by Jim)
2015-03-05 00:21:36 -08:00
Christoph Hohmann 52ff7b9734 image-source: Add a macro for logging info messages 2015-03-05 00:21:35 -08:00
jp9000 c5b8c744ed rtmp-services: Remove hitbox Vienna at request
Apparently it's not meant to be used quite yet.
2015-03-04 23:49:12 -08:00
Jim f0370f556a Merge pull request #395 from fryshorts/v4l2-buffering
linux-v4l2: Replace system timing option
2015-03-04 23:25:12 -08:00
fryshorts cdbf193e2e linux-v4l2: Add buffering option
Add a source property to enable buffering of frames, which is enabled by
default. This replaces the old "Use system timing" option by setting the
unbuffered source flag instead of using different timestamps.
While similar in intentions to the old option, this method should reduce
latency even more.
2015-03-04 21:21:22 +01:00
fryshorts c0f47c7efd linux-v4l2: Remove the system timing option
Remove the system timing property and code-path.
2015-03-04 21:01:18 +01:00
Radzaquiel d40e459e5c Update hitbox.tv ingest servers list
(Updated from Hitbox API)
Added: EU-West (Paris), US-East (New York)
2015-03-03 21:28:29 +01:00
Manuel Kroeber 8a55f3bda7 text-freetype2: Fix bug 0000151
Fix bug 0000151: File loading not properly handled.

Link to bug: https://obsproject.com/mantis/view.php?id=151

A newly selected font is not loaded properly if "read from file" is
active without a valid file.  Old error handling lead to random memory
being displayed.

Closes Pull Request #390

(message edited by Jim)
2015-03-01 23:39:13 -08:00
Skyler Lipthay 7d3eedbcb0 mac-avcapture: Default to unbuffered video capture
By default, video plays back based upon the timestamp for each frame,
and buffers the frames as needed to ensure that they play back at the
expected timing.

However, this can add some minor additional delay to the video, and may
not be ideal for certain devices such as webcams and generally any
device that has minimal latency.  However, because those are the only
type of devices that typically have drivers, there's no real need to
have it on by default.

This adds an option to use buffering, and leaves it off by default.

Closes pull request #384

(message added by jim)
2015-03-01 23:18:56 -08:00
fryshorts 6b95b2ccf7 linux-v4l2: Improve logging of pixel format
Use the macro from the mac capture plugin to convert the fourcc integer
value to a string. This makes the debug statement for the pixel format
slightly more readable for the casual developer.
2015-03-01 18:53:51 +01:00
fryshorts 7905197b7c linux-v4l2: Remove "Leave Unchanged" option
Remove the "Leave Unchanged" option for the input and video format
select.

This option was primarily added for cases in which the
resolution and framerate are set by another program or the capture
device itself and the values are not directly supported by the plugin.
One major usecase here would be capture devices for tv signals which
might be set to a spcific resolution and refresh rate, and might fail
to initialize in case any other combination of those settings is forced.

In case of the input this option did not make much sense, as the first
input is probably the best default option in most cases.

For the video format this default is even bad in some cases. If an
format emulated by libv4l2 is selected for example, this will usually
configure the device to use mjpeg with libv4l2 converting it to some
format obs can use. When obs or the source is then restarted and the
"Leave Unchanged" is enabled the plugin will fail, because it will only
notice that the device is set to mjpeg, without any knowledge about the
possibility of letting libv4l2 handle the conversion.

Using the first available and supported format is not nescessarily the
best choice, but still preferable to some random format that will
cause the plugin to not capture at all. Forcing a choice here will
hopefully also make the plugin behaviour more predicatable for the user.
2015-03-01 18:53:51 +01:00
Radzaquiel b86af28380 linux-v4l2: Fix string for translation
Changed "ImageFormat" for "VideoFormat"
(checked in v4l2-input.c)
2015-02-27 00:52:17 +01:00
fryshorts 1c0df043fa linux-v4l2: Use dv timing property in source
Use the dv timing selected in the source properties to configure the
device instead of resolution/framerate.
2015-02-24 23:32:46 +01:00
fryshorts 99b6a018cc linux-v4l2: Add helper function to set dv timing
Add a helper function to set the dv timing for the input.
2015-02-24 23:32:46 +01:00
fryshorts bcf8400137 linux-v4l2: Show dv timings when supported
Show the dv timing property when the selected input supports it.
2015-02-24 23:32:46 +01:00
fryshorts 5d317920f1 linux-v4l2: Add dv timing property
Add the dv timing property and hide it by default.
2015-02-24 23:32:46 +01:00
fryshorts bf27d4c1cb linux-v4l2: Add function to list dv timings
Add a helper function to enumerate dv timings supported by the
selected input and add them to a property.
2015-02-24 23:32:46 +01:00
fryshorts 0a6cf208f1 linux-v4l2: Add helper function to get dv timings
Add a helper function to get a dv timing by index for the input.
2015-02-24 23:32:46 +01:00
fryshorts f38347acc6 linux-v4l2: Use video standard property in source
Use the video standard selected in the source properties to configure
the device instead of resolution/framerate.
2015-02-23 21:27:08 +01:00
fryshorts 7dc9069c1b linux-v4l2: Add helper function for video standard
Add a helper function to get/set the video standard for the input.
2015-02-23 21:27:08 +01:00
fryshorts a9df1d921a linux-v4l2: Use input caps for properties
Use the capabilities for the selected input to decide which properties
are enumerated and displayed.
2015-02-23 21:27:08 +01:00
fryshorts bab2aa24d5 linux-v4l2: Add video standard property
Add the video standard property and hide it by default.
2015-02-23 20:34:44 +01:00
fryshorts cc57222aa9 linux-v4l2: Add function to enum video standards
Add a helper function to enumerate video standards supported by the
selected input and add them to a property.
2015-02-23 20:34:44 +01:00
fryshorts 96b994afb2 linux-v4l2: Add helper function to get input caps
Add a helper function to get the capabilities of a specific or the
currently selected input of the device.
2015-02-23 20:34:44 +01:00
fryshorts ff8ac210fb linux-v4l2: Remove camera constraint for inputs
Remove the constraint for device inputs to be of the type "CAMERA".
This was added under the false assumption that inputs of the type
"TUNER" are only used for control purposes.
2015-02-23 20:34:44 +01:00
jp9000 3c8a3d7866 mac-capture: Add update callbacks for coreaudio
Somehow managed to forget to add this function at an earlier point.
Allows the audio device to be changed via a simple update.
2015-02-20 21:53:00 -08:00
jp9000 8bc2288dfe mac-capture: Separate audio shudown code
This allows the code to be used elsewhere as needed.
2015-02-20 21:49:51 -08:00
jp9000 c69b70b847 obs-outputs: Fix bug (RTMP only working once)
This was caused to do the new RTMP code that added support for multiple
streams; the stream index needs to be reset on RTMP_Close otherwise it
will keep using the wrong stream information.
2015-02-17 20:38:23 -08:00
jp9000 d3c8f81cc2 obs-outputs: Fix potential RTMP auth issue 2015-02-17 17:23:32 -08:00
jp9000 8e9924d6a1 win-dshow: Do not update encoders if active
Do not update the avermedia encoders if they are already active, they
weren't designed to be updated in real time, so it will cause encoding
to break.
2015-02-14 15:04:11 -08:00
Martell Malone f079bc9f7f win-dshow: Reorder CapsMatch for gcc 2015-02-14 19:14:27 +00:00
jp9000 104415d5c3 win-capture: Fix hook bug that could cause crashes
I had this issue where IDXGISwapChain::ResizeBuffers would fail in the
hooks, causing games to crash when they resized their backbuffers
because ResizeBuffers would return an 'invalid call' HRESULT value.  In
the ResizeBuffers documentation it says that it will only happen if a
backbuffer currently has any outstanding references, but there's no way
this would happen unless ResizeBuffers internally calls Present or vise
versa.

After ResizeBuffers has been called, the very first call to Present will
somehow seemingly invalidate and/or destroy the current backbuffer.
It's very strange, but that seems to be what's going on, at least for
the game I was testing.  So if you are performing a post-overlay
capture, then you must ignore the capture on the very first call to
Present.

It's Microsoft's code so you can't really know what's going on, you just
have to work around these strange issues seemingly in the dark.
2015-02-14 09:12:24 -08:00
jp9000 8b59b606c5 win-capture: Use right call type for hook funcs
Although STDMETHODCALLTYPE is technically WINAPI (__stdcall), it's best
not to make any assumptions.
2015-02-14 07:48:36 -08:00
jp9000 5d9cac1f80 win-dshow: Rename the avermedia encoders again
Slight clarification (I guess)
2015-02-14 07:48:35 -08:00
jp9000 ae931b32cf win-capture: Fix debug function strings
These had an older name of what the function used to be named
2015-02-14 07:48:35 -08:00
Bl00drav3n 1f9f50c800 Fixed a bug in color conversion code for 16bit textures. 2015-02-14 01:08:38 +01:00
dodgepong 2b01030607 Add latest translations from crowdin 2015-02-11 19:15:08 -05:00
jp9000 04925993b0 win-dshow: Fix names of AVerMedia encoders 2015-02-11 15:29:31 -08:00
dodgepong 187f5181c7 Add latest translations from crowdin 2015-02-11 16:27:24 -05:00
jp9000 c90e55739f win-dshow: Allow overriding of color space/range
Certain devices may have a different color space or color range than
they may report, so allow the user to set it manually themselves if they
so wish.
2015-02-11 13:10:55 -08:00
jp9000 7e0a86e583 win-dshow: Fix stack overflow bug
Martell changed this function without realizing that this was calling a
function below it, not recursively calling itself.  The reason why he
got the warning was because there was no forward declaration of the
function that was being called; I think he's used to C where only one
function definition can exist with the same name.  In this case, it was
another function with the same name but with different parameters,
something that's permitted in C++.  I wish I had realized this sooner.

This fixes the crashes people have been having with devices.
2015-02-11 13:10:54 -08:00
Palana 8a3dd2ef8a librtmp: Try to parse IPv6 addresses
This should (hopefully) allow URLs of the form rtmp://[abcd::1]/app
2015-02-11 17:52:56 +01:00
Palana ce25cf86b6 librtmp: Remember addrlen from getaddrinfo
OS X doesn't like sizeof(struct sockaddr_storage) as last parameter
for connect
2015-02-11 17:52:56 +01:00
Palana b7375fd1be Revert "Revert "obs-outputs: Update to latest librtmp version""
This reverts commit ac72e0392d.
2015-02-11 17:40:14 +01:00
jp9000 9b850c7be1 mac-syphon: Name plugin "Game Capture (Syphon)"
People were getting a bit confused over what the "Syphon" source was,
and often wouldn't use it because they didn't know what it was for.
2015-02-10 22:28:56 -08:00
jp9000 2149817af7 obs-x264: Set lower buffer size limit to 0
0 is a common value used for recording.
2015-02-10 19:32:35 -08:00
jp9000 5a4a9befd2 rtmp-services: Add encoder settings callback
Swaps out the old initialize callback code that applied encoder settings
to the new apply_encoder_settings callback
2015-02-10 19:27:37 -08:00
jp9000 ac72e0392d Revert "obs-outputs: Update to latest librtmp version"
This reverts commit 36666c6712.
2015-02-09 19:30:11 -08:00