[mapserver-users] Re: Missing row of pixels in MapServer generated tiles

[Roger André]

Nope, that's not it.  Had the correct params in that entry after all:

<900913> +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0
+y_0=0 +k=1.0 +units=m +nadgrids=@null +no_defs

This matches documentation at
http://spatialreference.org/ref/user/google-projection/

Reprojected my data file with gdalwarp to epsg 900913, then updated the
mapfile to reflect that input layer is in that projection and tested.  Got
rid of the missing rixel rows at the bottom of tiles, but now the image only
extends between 80 to -80 latitude, and there is a weird, scalloped seam at
the international dateline.  So something is still not right, but it would
appear that it's projection related.
--


On 10/19/08, Roger André <randre at gmail.com> wrote:
>
> One thing I did notice is that my EPSG file definition for "Spherical
> Mercator" is missing a parameter, "+nadgrids=@null".  I'll test tomorrow to
> see if adding it to the definition removes the problem.
> --
>
>
> On Sat, Oct 18, 2008 at 7:25 PM, Roger André <randre at gmail.com> wrote:
>
>> This is probably more of a TileCache question, than a MapServer one, but I
>> figure one of you may have had the same issue in the past.  I'm using
>> MapServer to feed TileCache a fairly large GeoTiff with global extents,
>> which I'm then displaying in Google Maps as a tile layer.  This has worked
>> well in the past, but now I'm encountering a weird problem.  Some of my
>> tiles are being rendered by MapServer with a row of pixels missing on the
>> bottom of the image.  So for example, in Zoom Level 2, all of the 0.png
>> tiles will be 256 x 256 pixels, but row 256 will not contain data from my
>> global image, but instead will be set entirely to whatever background color
>> I have set in my Mapfile. The problem appears in multiple, but not all, Zoom
>> levels, and at different geographic locations, but always at the bottom row
>> of a tile.
>>
>> The source image doesn't have the lines in then, so I have to believe that
>> the "dead" row is an artifact of the rendering process.  My seed image is in
>> lat/lon WGS84, and the tiles are being requested in EPSG:900913.  The
>> reprojection seems to be working correctly, insofar as image lineup with
>> Google Maps seems to indicate.  If I take the seed image and create a very
>> reduced resolution version of it, then run it through TileCache, I don't get
>> any lines.  So it seems like the lines *might* be a result of downsampling,
>> but since they're not present at the bottom of every image, I'm not
>> convinced that's the problem.
>>
>> Oh, and I'm using the GD driver.
>>
>> Any ideas?
>>
>> Roger
>> --
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapserver-users/attachments/20081020/1ecaa03c/attachment.htm>