Issue 330 in webp: webp_info

25 views
Skip to first unread message

pdk… via monorail

unread,
Feb 8, 2017, 1:11:02 PM2/8/17
to webp-d...@webmproject.org
Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330

I'd like to request a tool possibly called webp_info, akin to webm_info, which prints the structure of a file.

https://chromium.googlesource.com/webm/libwebm/+/master/webm_info.cc

PNG has pngcheck.

http://www.libpng.org/pub/png/apps/pngcheck.html


--
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

pdk… via monorail

unread,
Feb 8, 2017, 5:39:56 PM2/8/17
to webp-d...@webmproject.org

Comment #1 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c1

It appears an integrity check is done, but I'm not sure how robust it is.

$ dwebp file.webp

pascal.m… via monorail

unread,
Feb 10, 2017, 7:38:57 AM2/10/17
to webp-d...@webmproject.org
Updates:
Labels: -Type-Defect Type-Enhancement
Owner: hu...@google.com
Status: Accepted

Comment #2 on issue 330 by pascal.m...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c2

Having a webp_info utility is indeed a good idea!

And yes, dwebp should be robust to pretty much any bitstream it is fed with.
And also the direct calls to WebPGetFeatures() in the library.

bugdro… via monorail

unread,
Mar 14, 2017, 2:32:24 PM3/14/17
to webp-d...@webmproject.org

Comment #3 on issue 330 by bugd...@chromium.org: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c3

The following revision refers to this bug:
https://chromium.googlesource.com/webm/libwebp/+/274daf541551690fefa8d27cd2b99caded5c3750

commit 274daf541551690fefa8d27cd2b99caded5c3750
Author: Pascal Massimino <pascal.m...@gmail.com>
Date: Tue Mar 14 10:41:17 2017

Add webpinfo tool.

A command line tool to print out the chunk level structure of
WebP files along with basic integrity checks.

BUG=webp:330

Change-Id: Ic69f646f649abb655b1854621d99afedeed158d7

[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/examples/Android.mk
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/makefile.unix
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/configure.ac
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/examples/Makefile.am
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/Makefile.vc
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/CMakeLists.txt
[modify] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/build.gradle
[add] https://crrev.com/274daf541551690fefa8d27cd2b99caded5c3750/examples/webpinfo.c

hu… via monorail

unread,
Mar 20, 2017, 1:32:11 PM3/20/17
to webp-d...@webmproject.org

Comment #4 on issue 330 by hu...@google.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c4

Hi pdk...@gmail.com, hope you got a chance to play with the new webpinfo tool. Please let us know if you have any comments or suggestions.

pdk… via monorail

unread,
Mar 20, 2017, 3:14:28 PM3/20/17
to webp-d...@webmproject.org

Comment #5 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c5

​Yep, I did. Works well. I have a few suggestions for lossless images, but
I'm not sure how useful they'd be to most users.​

pngcheck prints the filters (predictors) used for the image when -vv is
used. As PNG filters on scanlines, this is easy to parse (for the user)
when printed out, but probably less so for the block-based predictors of
WebP. I mean that when the predictors are just printed out sequentially,
it's difficult to correlate them with the image region.

Another suggestion is to print some of the information that is printed
during encoding like transforms used, although I don't know how much of
that can be retrieved from the file. And number of palette entries for
palette images.

pdk… via monorail

unread,
Mar 20, 2017, 3:40:30 PM3/20/17
to webp-d...@webmproject.org

Comment #6 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c6

Currently no warning is printed for files hat have alpha, but don't have the bit set in VP8X. I'm aware that it's merely informative, so not sure what is correct.

pdk… via monorail

unread,
Mar 25, 2017, 4:09:37 AM3/25/17
to webp-d...@webmproject.org

Comment #7 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c7

​Another minor observation: pngcheck skips chunk identifier, chunk size,
(and chunk CRC) when reporting chunk length, whereas webpinfo does.

pdk… via monorail

unread,
Mar 25, 2017, 4:09:57 AM3/25/17
to webp-d...@webmproject.org

Comment #8 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c8

I mean doesn't.

pdk… via monorail

unread,
Mar 26, 2017, 3:41:46 AM3/26/17
to webp-d...@webmproject.org

Comment #9 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c9

​Another minor observation: pngcheck skips chunk identifier, chunk size, (and chunk CRC) when reporting chunk length, whereas webpinfo doesn't.

pdk… via monorail

unread,
Mar 26, 2017, 3:42:47 AM3/26/17
to webp-d...@webmproject.org

Comment #10 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c10

(For some reason my comments are auto-deleted.)

​Another minor observation: pngcheck skips chunk identifier, chunk size, (and chunk CRC) when reporting chunk length, whereas webpinfo does.

pdk… via monorail

unread,
Mar 26, 2017, 5:44:54 AM3/26/17
to webp-d...@webmproject.org

Comment #11 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c11

I mean doesn't.

pdk… via monorail

unread,
Mar 26, 2017, 5:45:02 AM3/26/17
to webp-d...@webmproject.org

Comment #12 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c12

(I mean doesn't.)

pdk… via monorail

unread,
Mar 26, 2017, 5:45:28 AM3/26/17
to webp-d...@webmproject.org

Comment #13 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c13

I mean webpinfo doesn't.

pdk… via monorail

unread,
Mar 26, 2017, 5:45:51 AM3/26/17
to webp-d...@webmproject.org

Comment #14 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c14

I mean doesn't.

(Apparently short comments are auto-deleted.)

bugdro… via monorail

unread,
May 4, 2017, 2:30:29 PM5/4/17
to webp-d...@webmproject.org


The following revision refers to this bug:
https://chromium.googlesource.com/webm/libwebp/+/cbc1b92117d8302d77f56b82e8fe3f02cb635ac7

commit cbc1b92117d8302d77f56b82e8fe3f02cb635ac7
Author: Hui Su <hu...@google.com>
Date: Thu May 04 16:18:11 2017

webpinfo: add features to parse bitstream header

BUG=webp:330

Change-Id: If2a6ff010219fa25a2b5843725c616ee4dce65a7

[modify] https://crrev.com/cbc1b92117d8302d77f56b82e8fe3f02cb635ac7/examples/webpinfo.c

bugdro… via monorail

unread,
May 10, 2017, 1:23:53 AM5/10/17
to webp-d...@webmproject.org


The following revision refers to this bug:

bugdro… via monorail

unread,
May 17, 2017, 1:51:12 PM5/17/17
to webp-d...@webmproject.org


The following revision refers to this bug:
https://chromium.googlesource.com/webm/libwebp/+/dabda70733cf1366b3817e8453a1acfc1bf1a398

commit dabda70733cf1366b3817e8453a1acfc1bf1a398
Author: hui su <hu...@google.com>
Date: Wed May 17 16:35:58 2017

webpinfo: add support to parse Alpha bitstream

BUG=webp:330

Change-Id: I9ed0923d19558052d63bdbe93a542c51595fdce7

[modify] https://crrev.com/dabda70733cf1366b3817e8453a1acfc1bf1a398/examples/webpinfo.c

bugdro… via monorail

unread,
May 23, 2017, 2:51:51 PM5/23/17
to webp-d...@webmproject.org


The following revision refers to this bug:
https://chromium.googlesource.com/webm/libwebp/+/af6deaffa05ae578d8174b557a5ef4116136f121

commit af6deaffa05ae578d8174b557a5ef4116136f121
Author: hui su <hu...@google.com>
Date: Tue May 23 17:28:25 2017

webpinfo: handle alpha flag mismatch

Throw a warning for false positive vp8x alpha flag;
throw an error for false negative vp8x alpha flag.

BUG=webp:330

Change-Id: I3933588c9d6d76ecaf515517f7506dea299af638

[modify] https://crrev.com/af6deaffa05ae578d8174b557a5ef4116136f121/examples/webpinfo.c

hu… via monorail

unread,
Jun 6, 2017, 2:25:55 PM6/6/17
to webp-d...@webmproject.org
Updates:
Status: Fixed

Comment #19 on issue 330 by hu...@google.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c19

Closing this thread for now. Thanks for making the request. Please feel free to let us know any further suggestions and comments.

pdk… via monorail

unread,
Jun 6, 2017, 3:33:24 PM6/6/17
to webp-d...@webmproject.org

Comment #20 on issue 330 by pdk...@gmail.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c20

​Is there a reason why only the first transform is printed?​

hu… via monorail

unread,
Jun 6, 2017, 7:20:45 PM6/6/17
to webp-d...@webmproject.org

Comment #21 on issue 330 by hu...@google.com: webp_info
https://bugs.chromium.org/p/webp/issues/detail?id=330#c21

Before we can parse the next transforms, we need to entropy-decode the first transform's data completely. I think that is beyond the scope of webpinfo.
Reply all
Reply to author
Forward
0 new messages