[Mapserver-users] Raster support - JPG problem

[Lowther, David W]

Have you guys considered ECW instead of JPEG?

-----Original Message-----
From: Lowell Filak [mailto:lfilak at medinaco.org]
Sent: Wednesday, January 15, 2003 3:57 PM
To: Hankley, Chip; 'Zatorsky.MichaelP at police.qld.gov.au';
mapserver-users at lists.gis.umn.edu
Subject: RE: [Mapserver-users] Raster support - JPG problem


I would have to agree with Chip's performance findings.
We originally received all our orthophotography in TIFF format and planned
on converting the images to JPEG to enable us to load the dataset on smaller
machines.
However the performance hit was major in every application (Mapserver,
Arc/INFO, & Arcview) because the entire JPEG tile had to be decompressed
before the software could determine which pixels of the image needed
rendered.
You can save space by flattening the source images to 8-bit WITH space
reserved for 256 colors (some applications flatten to 8-bit but reserve
space for 1000+ colors and that doesn't save any space).
If truly needed you can always save space by resampling the data to a larger
pixel size, ie. 1' -> 2'.
Lowell F.

The following message was sent by "Hankley, Chip" <Chip.Hankley at GASAI.Com>
on Wed, 15 Jan 2003 12:46:00 -0600.

> Michael,
> 
> As Norman and Daniel pointed out, there are known issues with JPEG and
> MapServer.
> 
> That said, I had a similar problem some months back trying to serve up
JPEG
> images to MapServer. I did some experimenting and was finally able to get
a
> JPEG image to come up in MS (sorry, I can't remember exactly what I did),
> but the performance was terrible.
> 
> My assumption is that MapServer has to somehow uncompress the JPEG images
> before rendering them as part of the final image, and that this was the
> cause of the performance hit (if I simply swapped the data source to a
TIF,
> there was no performance problem). I have no idea if my assessment of how
> the mechanics of JPEG rendering is correct... but in the end, I concluded
> that the trade-off for space savings with JPEG didn't make up for the
> performance hit I was taking.
> 
> ...just my experience.
> 
> Chip
> _______________________________________________
> Mapserver-users mailing list
> Mapserver-users at lists.gis.umn.edu
> http://lists.gis.umn.edu/mailman/listinfo/mapserver-users

_______________________________________________
Mapserver-users mailing list
Mapserver-users at lists.gis.umn.edu
http://lists.gis.umn.edu/mailman/listinfo/mapserver-users