You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@pdfbox.apache.org by "Andreas Lehmkühler (JIRA)" <ji...@apache.org> on 2015/09/26 22:47:04 UTC
[jira] [Comment Edited] (PDFBOX-2934) Support smask /S /Alpha
[ https://issues.apache.org/jira/browse/PDFBOX-2934?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14909475#comment-14909475 ]
Andreas Lehmkühler edited comment on PDFBOX-2934 at 9/26/15 8:46 PM:
---------------------------------------------------------------------
I've found several issues.
The parent paint context of the softmask paint takes an affine transformation into account but the softmask not. This explains the shifted/flipped soft masks. I didn't found a solution to solve that yet. I just did a diryt hack to add the missing translation to the soft mask. As result PDFJS-1710-p1.pdf and PDFJS-1710-p7.pdf are rendered correct.
There are obviously different rules on how to apply the soft mask. missing_imager.pdf attached to PDFBOX-1416 can be rendered if the alpha values are reverted. But I don't know how to determine when to revert those values. Maybe that reversion has to be done some where else.
The calculation of the alpha value was wrong. It must not be premultiplied but simply be copied to the alpha channel.
I've attached a patch demonstrating the issues and the dirty hack.
was (Author: lehmi):
I've found several issues.
The parent paint context of the softmask paint takes an affine transformation into account but the softmask not. This explains the shifted/flipped soft masks. I didn't found a solution to solve that yet. I just did a diryt hack to add the missing translation to the soft mask. As result PDFJS-1710-p1.pdf and PDFJS-1710-p7.pdf are rendered correct.
There are obviously different rules on how to apply the soft mask. missing_imager.pdf attached to PDFBOX-1416 can be rendered if the alpha values are reverted. But I don't know how to determine when to revert those values. Maybe that reversion has to be done some where else.
The calculation of the alpha value was wrong. It must not be premultiplied but simply be copied to the alpha channel.
> Support smask /S /Alpha
> -----------------------
>
> Key: PDFBOX-2934
> URL: https://issues.apache.org/jira/browse/PDFBOX-2934
> Project: PDFBox
> Issue Type: Task
> Affects Versions: 2.0.0
> Reporter: Tilman Hausherr
> Attachments: PDFBOX2934_softmask.patch, PDFJS-1710-p1.pdf, PDFJS-1710-p1.pdf-1-better.png, PDFJS-1710-p1.pdf-1.png, PDFJS-1710-p7.pdf, PDFJS-1710-p7.pdf-1-better.png, PDFJS-1710-p7.pdf-1.png, PDFJS-5526-p11-13.pdf, gs-bugzilla689439-alphamask.pdf
>
>
> The smask feature /S /Alpha is either not implemented or not working at all. I have two files that are wrongly rendered with the current software, and correctly rendered when the feature is disabled \(!\). As a first step, I am disabling it for now and attach test files. Johns proposed patch has also disabled the feature, so I think we can agree on that.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: dev-help@pdfbox.apache.org