Issue 333 in webp: bitstream error for incremental decoder for webp encoded with latest

55 views
Skip to first unread message

ryan.luc… via monorail

unread,
Feb 28, 2017, 12:37:09 PM2/28/17
to webp-d...@webmproject.org
Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 333 by ryan.luc...@gmail.com: bitstream error for incremental decoder for webp encoded with latest
https://bugs.chromium.org/p/webp/issues/detail?id=333

What steps will reproduce the problem?
1. Decode the attached webp file
2. Notice decode error VP8_STATUS_BITSTREAM_ERROR

What is the expected output? What do you see instead?
I suspect there was an error encoding this webp file, resulting in the bitstream error on decode

What version of the product are you using? On what operating system?
* Tested with 0.6.0 and revision 6f75a51b58fe2d846eb9f490154e0b537f29ddd1
* Encoded with libwebp encoder in skia on Linux
* Decoded with libwebp decoder in skia on Linux

Please provide any additional information below.
Encoding the same SkBitmap with libwebp-0.5.1 produces a webp file that can be decoded successfully by 0.6.0


Attachments:
9369.webp 126 KB

--
You received this message because:
1. The project was configured to send all issue notifications to this address

You may adjust your notification preferences at:
https://bugs.chromium.org/hosting/settings

pascal.m… via monorail

unread,
Feb 28, 2017, 5:12:11 PM2/28/17
to webp-d...@webmproject.org
Updates:
Mergedinto: 332
Status: Duplicate

Comment #1 on issue 333 by pascal.m...@gmail.com: bitstream error for incremental decoder for webp encoded with latest
https://bugs.chromium.org/p/webp/issues/detail?id=333#c1

This bug is similar to entry #332

pascal.m… via monorail

unread,
Feb 28, 2017, 5:15:18 PM2/28/17
to webp-d...@webmproject.org

Comment #2 on issue 333 by pascal.m...@gmail.com: bitstream error for incremental decoder for webp encoded with latest
https://bugs.chromium.org/p/webp/issues/detail?id=333#c2

a bit more details:
* the patch proposed in entry #332 seems to fix the issue with 9369.webp
* encoding algorithms changed between 0.5.1 and 0.6.0. So, the bitstream produced by 0.5.1 just doesn't exercise the same problematic code path than the one produced by 0.6.0.
Reply all
Reply to author
Forward
0 new messages