In-band tear detection with compression
A system detects a screen tear in a display device for frames of video by retrieving an identification code associated with each pixel row of the frame of video and analyzing identification codes across the pixel rows of the frame of the video to detect occurrence of a screen tear. To support compressed (e.g., lossy) video data, the identification code associated with each pixel row of the frame of video is inserted into the most significant bit (or bits) of a color channel of a first pixel of the pixel row. Identification codes across the pixel rows of the frame of the video are then compared for tear detection. Because modification of the most significant bit of a color channel of a pixel may result in color distortion, the pixels including inserted identification codes are hidden from view by the display device.
1. A method comprising:
for each pixel row of a frame of video stored in one or more buffers of a display: retrieving an identification code from only a most significant bit of a pixel of the pixel row from a buffer, wherein a color of the pixel of the pixel row is defined by a number of bits, the number of bits including the most significant bit, all pixel rows of even numbered frames of the video including a first identification code and all pixel rows of odd numbered frames of the video including a second identification code;
detecting a mismatch between identification codes of at least two of the pixel rows of the frame of video;
responsive to detecting the mismatch, outputting a signal indicating an occurrence of a screen tear in the frame of video presented by the display; and
occluding the pixel of each pixel row of the frame of video from view on the display device with an opaque element disposed over the pixel of each pixel row.
2. The method of claim 1 , wherein the frame of video includes decoded compressed video data.
3. The method of claim 1 further comprising generating the frame of video by decoding compressed video data.
4. The method of claim 3 further comprising inserting the identification code associated with each pixel row of the frame of video into the compressed video data before the compressed video is decoded.
5. The method of claim 1 , wherein the identification code associated with the pixel row of the frame of video is inserted into a first pixel of a color channel of the pixel row.
6. The method of claim 5 , wherein bits of the pixel of the pixel row represent a color channel value.
7. The method of claim 1 , further comprising:
for each pixel row of a second frame of video, retrieving an identification code from only a most significant bit of a pixel of the pixel row;
detecting a lack of a mismatch between the retrieved identification codes of the pixel rows of the second frame of video; and
responsive to detecting the lack of the mismatch, outputting a signal indicating no occurrence of a screen tear in the second frame of video.
8. The method of claim 1 , wherein the detecting the mismatch is performed during a vertical blanking period after the frame of video has been scanned by the display device.
9. A system, comprising:
a display panel configured to present a frame of video;
one or more buffers configured to store the frame of video presented by the display panel;
a timing controller configured to:
for each of multiple pixel rows of the frame of video stored in the one or more buffers, retrieve an identification code from only a most significant bit of a pixel of the pixel row from a buffer, wherein a color of the pixel of the pixel row is defined by a number of bits, the number of bits including the most significant bit, a first set of pixel rows of a first group of frames of the video including a first identification code and a second set of pixel rows of a second group of frames of the video including a second identification code;
detect a mismatch between identification codes of at least two of the pixel rows of the frame of video; and
responsive to detecting the mismatch, output a signal indicating an occurrence of a screen tear in the frame of video presented by the display panel; and
an opaque element disposed over the pixel of each pixel row.
10. The system of claim 9 , wherein:
the frame of video is generated by decoding compressed video data including the identification code associated with each pixel row of the frame of video; and
the identification code associated with each pixel row of the frame of video is inserted into the compressed video data before the compressed video is decoded.
11. The system of claim 9 , wherein the identification code associated with the pixel row of the frame of video is inserted into a first pixel of a color channel of the pixel row.
12. The system of claim 11 , wherein bits of the pixel of the pixel row represent a color channel value.
13. The system of claim 9 , wherein the timing controller is further configured to:
for each pixel row of a second frame of video, retrieve an identification code from only a most significant bit of a pixel of the pixel row;
detecting a lack of a mismatch between the retrieved identification codes of the pixel rows of the second frame of video; and
responsive to detecting the lack of the mismatch, output a signal indicating no occurrence of a screen tear in the second frame of video.
14. The system of claim 9 , wherein the timing controller is configured to detect the mismatch during a vertical blanking period after the frame of video has been scanned by the display panel.