This repository was archived by the owner on Jul 22, 2025. It is now read-only.
[360SCVP] Add scaling for low-res RWPK info generation #62
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Compiling of testMainEncoder cannot pass because of changes brings from local mode in WorkSession. This change is commited to fix the issue so it won't block CI porcess.
1. KeyCount of shared memory was accumulated from 1, for now there is an array to collect random values. 2. Add Local mode for WorkSession to imporve performance on local machine. Replace ip address such as "127.0.0.1" with "local" to start local mode.
Set targetSocket for SVT-HEVC to imporve encode performance. See details in following link: https://github.com/OpenVisualCloud/SVT-HEVC/blob/master/Source/API/EbApi.h#L585
…ection Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
This issue corresponds to bug VCD-1855 [immersive][client] Client's tile splicing sequence is not correct. Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
1. Enbale 360SVCP library which modified face index for selected tiles when source is from cubemap projection. 2. Add stdc++ dynamic linking file incase of `GLIBCXX_3.4.22' not found issue.
Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
…jection format setting from player to MPD file parsing result. 2. modify ineffective projection format setting for each adaptation set. 3. modify 360SCVP API calling according to actual projection format 4. add region wise packing information calculation for tiles merged picture when in Cubemap projection 5. add tile position conversion according to cubemap faces relationship between standard Cube-3x2 faces layout defined in OMAF spec and faces layout used in 360SCVP. Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
…ct additional tiles num in initial viewport in ERP source. 2. Add an additional tile selected when selected tile num in 360SCVP is prime number in cubemap source. Signed-off-by: Luo, Ying <ying2.luo@intel.com>
…dd install_safestringlib.sh script and add lib copy in fpm.sh. 2. Replace memcpy with memcpy_s in safestring library in player. Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Since memory copy is reduced in local mode, call to the SVT-HEVC interface is also changed to synchronous to avoid memory release problems when distributed encoder is used as FFmpeg plugin.
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
…cked in multi-threads Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Last commit did not change the sending of frame to synchronous completely, modified sending process in LocalWorkSession to fix the issue.
Call SwapBuffers when render status is not RENDER_NO_FRAME or RENDER_ERROR. Signed-off-by: Luo, Ying <ying2.luo@intel.com>
…ownloading 2. add downloading task priority 3. add downloading params: timeout limit/retry/proxy, etc. 4. rewrite the client readmanager by using state queues with timeline point information 5. add curl url resource checker 6. add the segment range sync logic 7. use the timeline information as Packet PTS. the timeline values is inited from the segment number.
Signed-off-by: dahanhan <dahan.han@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Signed-off-by: dahanhan <dahan.han@intel.com>
Signed-off-by: dahanhan <dahan.han@intel.com>
A shared memory KeyCount list was saved to allocate shared memory, but there were preempting happend when using multiple threads or processes. To avoid this issue, MainEncoder creates a circular queue of FrameSM at its initialization step, each FramesSM holds necessary information for the shared memory of one single Frame. Keep adding new FrameSM to the queue when shared memory in the head of queue is not able to free, this dynamically growing queue controls all of shared memory.
After adding Cubemap support in 360SCVP, new structure "Param_VideoFPStruct" is added into "Param_ViewPortInfo". And this new structure should also be initialized for ERP in VROmafPacking for extractor track generation Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Signed-off-by: dahanhan <dahan.han@intel.com>
…ck in OmafTilesStitch class. Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Thread in StreamReceiver cannot exit normally during destruction, fixed by add flag to exit endless loop, also modified thread quit function in SubEncoderManager_SVT and SubEncoderSVT.
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Signed-off-by: dahanhan <dahan.han@intel.com>
…lass elements missed initialization in the copy constructor 2. Dynamic memory needs to be released and reallocated in the operator assignment Signed-off-by: dahanhan <dahan.han@intel.com>
… for the viewport endpoints will use ArcSin function which required the input value be limited to [-1, 1]. In the implementation this input value is calculated as a floating point division, and when the number and denumber is very close, the division results may be slightly bigger than 1. It's necessary to add a limitation on the input value to avoid unpredictable results from the ArcSin function. Signed-off-by: Dahan Han <dahan.han@intel.com>
Fix bug VCD-2592, and the root cause is that EOS status is not processed in audio track segments writing Signed-off-by: Chen Xiaomin <xiaomin.chen@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Signed-off-by: Sun, Yanying <yanying.sun@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Replace the address check with an integer return, reduce the calculation and fix the random hang issue at exit.
Signed-off-by: Sun, Yanying <yanying.sun@intel.com>
Signed-off-by: Sun, Yanying <yanying.sun@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
Signed-off-by: Luo, Ying <ying2.luo@intel.com>
…_fix Update dockerfile for travis CI
Signed-off-by: dahanhan <dahan.han@intel.com>
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The low-resolution region RWPK information should be modified based on the image size percentage of high-res and low-res
Signed-off-by: dahanhan dahan.han@intel.com