Re: [Gmod-help] Re: Hi-Res images of conservation tracks have the background turn blue.
Angie Hinrichs <
angie@...>
2009-06-01 16:57:41 GMT
I took a look at the latest C source code for gdImageCopy()
(http://cvs.php.net/viewvc.cgi/gd/libgd/src/gd.c?revision=1.83&view=markup)
and it looks like since gd version 2.0.10, if a pixel is transparent
in the source, it is not copied to the destination. I can see how
that might interact badly when copying into something other than
another pixellated image.
Is there a hook in GBrowse by which a plugin or glyph can tell that it
is in high-res mode, so it can avoid using transparency?
Thanks,
Angie
On Fri, 29 May 2009, Lincoln Stein wrote:
> Ok. I'll poke around in GD::SVG to see what might be the problem with this
> strategy. It ought to be ok.
>
> Lincoln
>
> On Fri, May 29, 2009 at 5:18 PM, Angie Hinrichs <angie@...> wrote:
>
> > > Angie, is there any reason to call transparent()? The other tracks render
> > > their background in a solid color.
> >
> > In a nutshell, to avoid hiding the vertical light blue ruler lines in
> > the background by drawing white on top of them.
> >
> > Currently, ucsc_glyph::draw gets a colormap+pixelmap representation of
> > an image with a white background (from UCSC C code by way of
> > UcscConservation plugin), creates a GD::Image $im as described below
> > and uses $gd->copy to plop that on top of the GBrowse image:
> >
> > $gd->copy($im, $x-$leftLabelWidth, $y, 0, 0, $width, $height);
> >
> > -- so it bypasses all of the usual glyph draw() infrastructure, and
> > renders the entire allocated space in the main image. Using
> > transparency preserves the light blue lines and whatever else might be
> > in the background... is it not worth the trouble? Is there a better
> > way?
> >
> > Thanks,
> > Angie
> >
> > On Fri, 29 May 2009, Lincoln Stein wrote:
> >
> > > Angie, is there any reason to call transparent()? The other tracks render
> > > their background in a solid color.
> > >
> > > Lincoln
> > >
> > > On Fri, May 29, 2009 at 3:56 PM, Angie Hinrichs <angie@...>
> > wrote:
> > >
> > > > Mark, Do you see any error_log warnings like "ucsc_glyph: colors will
> > > > be out of sync ($i --> $index)!" ?
> > > >
> > > > Wherever the bad interaction is taking place, it is definitely an
> > > > alpha-channel problem. To construct the GD::Image using the colormap
> > > > and pixelmap returned by the UCSC C code,
> > > > Bio/Graphics/Glyph/ucsc_glyph.pm uses these calls:
> > > >
> > > > my $im = GD::Image->new($width, $height);
> > > > $im->transparent( <UCSC colormap's white index> );
> > > >
> > > > ... loop on UCSC colormap calls $im->colorAllocate(),
> > > > ** hoping that calling $im->colorAllocate() in the same order
> > > > results in same colormap indices
> > > >
> > > > ... loop on UCSC pixelmap calls $im->setPixel()
> > > >
> > > > ** Here is where something could go wrong. It looks like I should
> > > > call $im->transparent *after* the call to $im->colorAllocate() for
> > > > white, making sure that I pass $im's actual white index to
> > > > $im->transparent. However, when I call transparent after
> > > > colorAllocate, the white pixels are shaded light gray instead of being
> > > > transparent, even in the default display. Calling $im->transparent
> > > > first was just a hack that worked, and then I forgot about it. I will
> > > > try updating GD too, to see if that makes any difference.
> > > >
> > > > FWIW I am constructing the GD::Image one color and pixel at a time
> > > > because when I used GD::Image->newFromGdData(), the image was created
> > > > OK but when I returned the new $im from a sub{}, perl crashed with a
> > > > 'glibc detected double free or corruption'. Does that ring a bell for
> > > > anybody?
> > > >
> > > > Angie
> > > >
> > > > On Fri, 29 May 2009, Todd W. Harris wrote:
> > > >
> > > > > Mark -
> > > > >
> > > > > Try upgrading your GD::SVG to the most current version on CPAN.
> > > > >
> > > > > Todd
> > > > >
> > > > >
> > > > > On May 29, 2009, at 9:42 AM, mark meyer wrote:
> > > > >
> > > > > > We use the GD::SVG and was never able to get the inkscape (pdf
> > > > generation)
> > > > > > operational, if that helps any.
> > > > > >
> > > > > > On Fri, May 29, 2009 at 9:52 AM, Scott Cain <scott@...>
> > > > wrote:
> > > > > > Hi all,
> > > > > >
> > > > > > Could this be a bad interaction between the plug in and GD::SVG or
> > > > > > inkscape (depending on what sort of hi res image you are getting)?
> > > > > > For example, I think that GD::SVG has a hard time with alpha
> > channels.
> > > > > >
> > > > > > Scott
> > > > > >
> > > > > >
> > > > > > On Fri, May 29, 2009 at 10:31 AM, mark meyer <
> > mmeyerposting@...>
> > > > > > wrote:
> > > > > > > Hi Angie,
> > > > > > >
> > > > > > > Bill is out of the office today, but I can try to answer your
> > > > questions.
> > > > > > > Attached are the two screen shots, one is what the plugin looks
> > like
> > > > in
> > > > > > > the
> > > > > > > gbrowse, just fine as it looks in UCSC. The second is what
> > happens
> > > > when
> > > > > > > you
> > > > > > > click on high-res image.
> > > > > > >
> > > > > > > We are running gbrowse v1.69. The plugin code was implemented
> > and
> > > > > > > downloaded about 1 month ago, just recently ran into this issue
> > since
> > > > we
> > > > > > > are
> > > > > > > trying to make some publication figures from our Gbrowse data.
> > > > > > >
> > > > > > > I can access conf files, etc. if you need to see them and try to
> > do
> > > > my
> > > > > > > best
> > > > > > > to answer questions, but I may not be able to answer very
> > detailed
> > > > > > > questions
> > > > > > > in Bill's absence since he has been the one configuring this.
> > > > > > >
> > > > > > > Thanks for checking into this for us,
> > > > > > >
> > > > > > > Mark
> > > > > > >
> > > > > > > On Thu, May 28, 2009 at 5:29 PM, Angie Hinrichs <
> > angie@...>
> > > > > > > wrote:
> > > > > > > >
> > > > > > > > Dave - yes, that sounds likely.
> > > > > > > >
> > > > > > > > Bill, what version of GBrowse are you running? How recently
> > did
> > > > you
> > > > > > > > grab the plugin code? I didn't get the screenshot attachment
> > so if
> > > > > > > > you wouldn't mind sending that to me, that would be helpful.
> > If
> > > > your
> > > > > > > > GBrowse is on a public server, can you send the URL?
> > > > > > > >
> > > > > > > > Angie
> > > > > > > >
> > > > > > > > On Thu, 28 May 2009, Dave Clements, GMOD Help Desk wrote:
> > > > > > > >
> > > > > > > > > Hi Bill,
> > > > > > > > >
> > > > > > > > > I have no particular insight on why this is happening. My
> > guess
> > > > is
> > > > > > > > > that it is due some obscure interaction between the UCSC
> > plugin
> > > > and
> > > > > > > > > core GBrowse, but I don't actually know.
> > > > > > > > >
> > > > > > > > > I've CC'd Angie at UCSC to see if she may have any insight.
> > > > > > > > >
> > > > > > > > > Dave C.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > On Tue, May 26, 2009 at 2:25 PM, Bill Pulec <
> > wpulec@...>
> > > > wrote:
> > > > > > > > > > Hi All,
> > > > > > > > > > I am attempting to view high-res images of conservation
> > tracks
> > > > but
> > > > > > > > the
> > > > > > > > > > background of these tracks in high-res is blue. This isn't
> > the
> > > > case
> > > > > > > > > > when
> > > > > > > > > > these tracks are displayed in gbrowse normally, it only
> > happens
> > > > when
> > > > > > > > > > viewed
> > > > > > > > > > as high-res. Also all the non-conservation tracks are fine
> > in
> > > > high
> > > > > > > > > > res. Is
> > > > > > > > > > there any reason why outputting these images in high-res
> > would
> > > > do
> > > > > > > > > > this? Has
> > > > > > > > > > anyone encountered other tracks doing this in high-res? Any
> > > > insight
> > > > > > > > > > would be
> > > > > > > > > > appreciated. I have attached an example image of the track
> > > > below.
> > > > > > > > > >
> > > > > > > > > > Thanks,
> > > > > > > > > > Bill
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > > Was this helpful? Let us know at
> > > > > > > > > http://gmod.org/wiki/Help_Desk_Feedback
> > > > > > > > >
> > > > > > > > > Learn more about GMOD at:
> > > > > > > > > SMBE:
> > > > > > > > >
> > > > http://ccg.biology.uiowa.edu/smbe/symposia.php?action=view&sym_ID=27
> > > > > > > > > Arthropod Genomics:
> > > > http://www.k-state.edu/agc/symp2009/seminar.html
> > > > > > > > > AGA Next Gen Seq in Non-Models:
> > > > > > > > > http://www.regonline.com/Nextgeneration
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > >
> > ------------------------------------------------------------------------------
> > > > > > > > Register Now for Creativity and Technology (CaT), June 3rd,
> > NYC.
> > > > CaT
> > > > > > > > is a gathering of tech-side developers & brand creativity
> > > > professionals.
> > > > > > > > Meet
> > > > > > > > the minds behind Google Creative Lab, Visual Complexity,
> > > > Processing, &
> > > > > > > > iPhoneDevCamp as they present alongside digital heavyweights
> > like
> > > > > > > > Barbarian
> > > > > > > > Group, R/GA, & Big Spaceship.
> > > > http://p.sf.net/sfu/creativitycat-com
> > > > > > > > _______________________________________________
> > > > > > > > Gmod-gbrowse mailing list
> > > > > > > > Gmod-gbrowse@...
> > > > > > > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > >
> > > >
> > ------------------------------------------------------------------------
> > > > > > Scott Cain, Ph. D. scott at
> > scottcain
> > > > dot
> > > > > > net
> > > > > > GMOD Coordinator (http://gmod.org/)
> > 216-392-3087
> > > > > > Ontario Institute for Cancer Research
> > > > > >
> > > > > >
> > > >
> > ------------------------------------------------------------------------------
> > > > > > Register Now for Creativity and Technology (CaT), June 3rd, NYC.
> > CaT
> > > > > > is a gathering of tech-side developers & brand creativity
> > > > professionals.
> > > > > > Meet
> > > > > > the minds behind Google Creative Lab, Visual Complexity,
> > Processing, &
> > > > > > iPhoneDevCamp as they present alongside digital heavyweights like
> > > > Barbarian
> > > > > > Group, R/GA, & Big Spaceship.
> > http://p.sf.net/sfu/creativitycat-com
> > > > > > _______________________________________________
> > > > > > Gmod-gbrowse mailing list
> > > > > > Gmod-gbrowse@...
> > > > > > https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
> > > > >
> > > >
> > >
> > >
> > >
> > >
> >
>
>
>
>
------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, &
iPhoneDevCamp as they present alongside digital heavyweights like Barbarian
Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
RSS Feed