[ncl-talk] Plot looks fuzzy when transparency is used

[Rick Brownrigg]

Hi Kyle,

I am not aware of any interactions between the HLU objects and the use of
the opacity channel.  The "fuzzy" images look like the vectors have been
drawn without any kind of antialiasing.  Can you provide your code -- you
can send it directly to me if you don't to post it publicly.

Rick


On Thu, Jan 15, 2015 at 3:02 PM, Kyle Griffin <ksgriffin2 at wisc.edu> wrote:

> Hi all,
>
> Curious case I've been looking at for the past few weeks, and I think I
> finally have enough to share with you. The two images below are made by the
> same set of resources applied (via setvalues) directly to relevant map and
> field objects in NCL - the only difference is from some changes in the fill
> opacities and the draw order. The image ending in 000 (the first link)
> appears, to me, to be fuzzy - especially with the vector edges, especially
> when compared to the 012 image (second link).
>
> With partially opaque fills:
> http://aos.wisc.edu/dynamaps/15011512/jet_pv_irrwind/jet_pv_irrwind_africa_15011512_000.jpg
> Without partially opaque fills:
> http://aos.wisc.edu/dynamaps/15011512/jet_pv_irrwind/jet_pv_irrwind_africa_15011512_012.jpg
>
> Your first thought, like mine, may jump to the image type (JPEG) which is
> not designed to handle transparencies. Since NCL doesn't produce JPG files,
> it's obvious these have been run through 'convert', but I don't believe
> that to be the source of the problem either. The curious thing, however, is
> that this occurs regardless of the original source file. I've done this by
> directly producing EPS and PNG files and the final result, with
> transparencies, is the same.
>
> Direct product of NCL (png):
> http://aos.wisc.edu/dynamaps/15011512/jet_pv_irrwind/ncl_transp_orig.png
>
> Is there something inherent to my use of the object-oriented applications
> that degrades image quality with the use of opacity settings? I've made
> similar/nearly identical images via gsn_xxxx calls with perfect success.
> Any thoughts here would be appreciated. I can provide some code, but I can
> assure you it's not something that will be easily abbreviated into a
> typical ncl-talk example (sorry).
>
>
> Kyle
>
> ----------------------------------------
> Kyle S. Griffin
> Department of Atmospheric and Oceanic Sciences
> University of Wisconsin - Madison
> Room 1421
> 1225 W Dayton St, Madison, WI 53706
> Email: ksgriffin2 at wisc.edu
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150115/b75b17dc/attachment.html