[Gdal-dev] Flashy colors in a pyramid

Wed Aug 19 09:20:35 EDT 2009

I see the same strange areas when I'm viewing the tiles with Irfanview, so geoserver is excluded. 

I re-created the first tile that gave me the problem (and only this tile) from the source files using gdal_retile and gdal_translate -projwin. The problem isn't there any more. So I guess it was only a random error ? I still need to recreate the higher tiles, if that works. 

Thanks for the tip on the option "TILED=YES". It comes from gdal_translate, I didn't think much when I started using gdal_retile. So I have no specific reason to use it and the output tiles of gdal_retile aren't tiled inside. 

Alain 

----- Mail Original ----- 
De: "Christian Müller (via Nabble)" <ml-user+68004-9764870 at n2.nabble.com> 
À: "acangi" <aca at ngi.be> 
Envoyé: Mercredi 19 Août 2009 14h34:41 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne 
Objet: Re: [Gdal-dev] Flashy colors in a pyramid 

Hi 

The thing I want to know is if you have the blue or green areas in the 
output files of gdal_retile.py or if you see them when looking at the images 
with geoserver. So, is at problem of gdal_retile.py or of geoserver loading 
the output from gdal_retile.py. 

Btw, you use the option -co "TILED=YES". Why ?. Each output tile of 
gdal_retile.py has again tiles in it ? 

acangi writes: 

> 
> Hi Christian, 
> 
> Thanks for your reply and thanks for writing gdal_retile. I use it a lot and 
> it works great. 
> 
> I'm not sure I understand you well. On the highest level of the pyramid (the 
> top) there's only one file. Gdal_retile already created the files needed for 
> the mosaic (shp, shx, dbf and prj). Would you like me to re-create them ? 
> 
> Or do you mean at the base of the pyramid ? 
> 
> Alain 
> 
> 
> Christian Müller wrote: 
>> 
>> Hi, Christian speaking, I wrote the gdal_retile.py utility. 
>> 
>> Could you try to run a gdal_merge.py on the highest pyramid level 
>> containing 
>> this strange area. What is the result.? I want to know if the problem is 
>> gdal_retile.py or geoserver. 
>> 
>> 
>> 
>> acangi writes: 
>> 
>>> 
>>> Hello list, 
>>> 
>>> I created a pyramid of aerial photos using gdal_retile.py and now see 
>>> there 
>>> are strange zones in the pyramid. I'm trying to understand why I got this 
>>> strange results and how I can fix it (Bombing the area to make sure it 
>>> looks 
>>> like what I have ?). Description : 
>>> 1. The command I used was gdal_retile.py -targetDir pyramid/ -levels 12 
>>> -pyramidOnly -v -s_srs /path/to/prjs/3812.prj -ps 10000 10000 -r bilinear 
>>> -co "TILED=YES" -co "COMPRESS=LZW" -co "INTERLEAVE=PIXEL" -tileIndex 
>>> orthocol -tileIndexField location /path/to/source/files/*.tif & 
>>> 2. In the source files, I see nothing special. The tiles in this area 
>>> look 
>>> like the other ones. 
>>> 3. On the first (generated) level of the pyramid, I see a small area 
>>> (alias 
>>> SMALL_SQUARE : 256m x 256m) in blue : 
>>> http://n2.nabble.com/file/n3470903/BlueSquare.jpg 
>>> 4. On the next levels of the pyramid, a black rectangle 512m x 256m is 
>>> now 
>>> visible where SMALL_SQUARE was : 
>>> http://n2.nabble.com/file/n3470903/BlackRect.jpg 
>>> and the area corresponding to the first level tile containing 
>>> SMALL_SQUARE 
>>> (alias BIG_SQUARE : 10000m x 10000m) is in green : 
>>> http://n2.nabble.com/file/n3470903/GreenSquare.jpg 
>>> 5. I view the pyramid in geoserver, the tiles look like they are on disk, 
>>> except at the first level, geoserver doesn't show me the full 
>>> SMALL_SQUARE, 
>>> only parts of it. In geoserver logs, I find nothing very interesting, it 
>>> all 
>>> looks the same whether or not I'm looking at BIG_SQUARE. 
>>> 6. If I set OutputTransparentColor to #000000 (the background color), 
>>> BIG_SQUARE is now invisible (in geoserver) 
>>> 
>>> So, my questions are : 
>>> Why did it happen ? I suppose it's just a random error... 
>>> How can I fix it ? Is there any way other than rebuilding the whole 
>>> pyramid 
>>> ? 
>>> 
>>> Any idea is welcome, 
>>> 
>>> Alain 
>>> -- 
>>> View this message in context: 
>>> http://n2.nabble.com/Flashy-colors-in-a-pyramid-tp3470903p3470903.html 
>>> Sent from the GDAL - Dev mailing list archive at Nabble.com. 
>>> _______________________________________________ 
>>> gdal-dev mailing list 
>>> [hidden email] 
>>> http://lists.osgeo.org/mailman/listinfo/gdal-dev 
>> 
>> 
>> _______________________________________________ 
>> gdal-dev mailing list 
>> [hidden email] 
>> http://lists.osgeo.org/mailman/listinfo/gdal-dev 
>> 
>> 
> 
> -- 
> View this message in context: http://n2.nabble.com/Flashy-colors-in-a-pyramid-tp3470903p3471888.html 
> Sent from the GDAL - Dev mailing list archive at Nabble.com. 
> _______________________________________________ 
> gdal-dev mailing list 
> [hidden email] 
> http://lists.osgeo.org/mailman/listinfo/gdal-dev 

_______________________________________________ 
gdal-dev mailing list 
[hidden email] 
http://lists.osgeo.org/mailman/listinfo/gdal-dev 

View message @ http://n2.nabble.com/Flashy-colors-in-a-pyramid-tp3470903p3471969.html 
To unsubscribe from Re: Flashy colors in a pyramid, click here . 

-- 
View this message in context: http://n2.nabble.com/Flashy-colors-in-a-pyramid-tp3470903p3472194.html
Sent from the GDAL - Dev mailing list archive at Nabble.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/gdal-dev/attachments/20090819/f940e952/attachment.html