Re: Making BIG images

From: Don Morton <morton_at_nyahnyahspammersnyahnyah>
Date: Tue, 02 Jun 2009 14:42:25 -0800

Thanks for the input, Dave. Although it worked for the simple case I
presented, what I haven't told you all yet is that this is being applied
to a complicated panel plot, and in that case the PNG driver wouldn't
even give me output.

The good news, however, is that through the RTFM methodology, I've
learned that I can specify custom page sizes to "gs" with the
-dDEVICEWIDTHPOINTS and -d DEVICEHEIGHTPOINTS flags and actually see the
full image. Hence, more evidence that NCL is doing just fine, and that
my real issue is getting the resultant ps, which seems good, converted.
Hopefully a little more RTFM will get me there :)

Thanks,

Don

Dave Allured wrote:
> Don,
>
> I have had good luck making large hi-rez GIF images for web with
> NCL's "experimental PNG driver". It is described as problematic,
> but so far I have not had any trouble.
>
> http://www.ncl.ucar.edu/FAQ/#o_formats_003
> http://www.ncl.ucar.edu/Support/talk_archives/2008/1151.html
>
> Use the special attributes on workstation type to increase the pixel
> size of the PNG output frame. Then use NCL plot routines in the
> usual way.
>
> rez = 2000 ; makes 2000 x 2000 pixel PNG
> output = "png"
> output_at_wkWidth = rez ; must be square; deviations
> output_at_wkHeight = rez ; either way seem to be wasted
> wks = gsn_open_wks (output, outfile)
>
> Optionally, use ImageMagick to reduce or remove any unwanted white
> border. The following command works for me and makes a lossless
> conversion, no blurring of pixels with our current version of
> convert. HTH.
>
> convert -trim out.png out.gif
>
> Dave Allured
> CU/CIRES Climate Diagnostics Center (CDC)
> http://cires.colorado.edu/science/centers/cdc/
> NOAA/ESRL/PSD, Climate Analysis Branch (CAB)
> http://www.cdc.noaa.gov/psd1/
>
> Don Morton wrote:
>
>> Hi,
>>
>> I'm sure this must be trivial, and perhaps not an NCL issue, but the
>> solution has eluded me.
>>
>> I'm posting some forecast graphics to the web, and users would like them
>> to be larger, but without a sacrifice in resolution. To date, I've been
>> able to use gsnPaperWidth and gsnPaperHeight resources to accomplish
>> this, but I end up with only a portion of the image - it seems like my
>> viewing device wants to continue treating the image as if it were being
>> displayed on 8.5" x 11.0" media.
>>
>> I've reduced the problem to a very simple script (see below). I run the
>> script to produce "Big.ps" and when I view it using "gs" or "gv" it only
>> shows me the left bottom section. However, by playing with page sizes
>> in "gv" I can get more or less of the image. So, I suspect that NCL is
>> probably doing what it should, and that my problems stem from my
>> ignorance in dealing with different media types.
>>
>> Once I get "Big.ps" I try to use ImageMagick to convert to GIF
>>
>> convert Big.ps Big.jpg
>>
>> but Big.jpg ends up, again, being only a partial image.
>>
>> Is there something I should be setting in NCL, or are there a set of
>> flags I should use with "convert" to ultimately get the full, but
>> bigger, image in GIF or PNG format?
>>
>> Thanks for any help. My simplified script follows:
>>
>> ; Include files
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/wrf/WRFUserARW.ncl"
>>
>> begin
>>
>> wrfoutFile =
>> "/projects/ARSCWTHR/ForecastProducts/AKTrial2/wrfout/2009052612/wrfout_d03_2009-05-30_09:00:00.nc"
>> ; Full path to the wrfout file
>>
>> f = addfile(wrfoutFile,"r") ; Opens the netCDF file
>>
>> wks = gsn_open_wks("ps", "Big") ; Open a workstation for plotting
>>
>> gsn_define_colormap(wks,"WhViBlGrYeOrReWh") ; Define the colormap
>>
>> ;;;;;;;;;;;;;; Temperature plot ;;;;;;;;;;;;;;;
>>
>> tc = wrf_user_getvar(f, "tc", 0) ; T in deg C
>> tc2d = tc(0,:,:)
>>
>> res = True
>> res_at_gsnMaximize = True ; Maximize size of plot
>> res_at_gsnPaperOrientation = "auto"
>> res_at_gsnPaperWidth = 17.0
>> res_at_gsnPaperHeight = 22.0
>>
>> plot = gsn_contour(wks, tc2d, res) ; Plot the data
>>
>> end
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Jun 02 2009 - 16:42:25 MDT

This archive was generated by hypermail 2.2.0 : Wed Jun 03 2009 - 12:52:48 MDT