pconvert changed from 4.6.1 to 4.9.0.1?

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-03-15T12:59:46Z

Dear Olga,

Thank you for detailing the difference between 4.6 and 4.9. and your suggestion of using an RGB text file. Using those clues, I made a file:
R=lat
G=lon
B=feq(lat,lon,sea_surface_temperature)

and this give me a jpeg exactly like the old one!

Many thanks,
Pauline

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-03-12T12:40:39Z

Dear Pauline,

can you send your e-mail-address to olga.faber@brockmann-consult.de? I will then send you login data for our FTP server, where you can upload the file.
Otherwise, could you provide us access to the MERCI you've downloaded the products from?

Best regards,
Olga

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-03-09T13:48:42Z

Hi Olga,

I've been investigating further and trying to replicate what you did, which I think was:

Use BEAM to open a ATS NR .N1 product, export it to netcdf, then use pconvert to make a jpeg. When I did this, I also generated a jpeg like the ones that you did, and like I did before.

So the issue is with the L2P netcdf product. These have filenames like:
20120229-ATS_NR_2P-UPA-L2P-ATS_NR__2PRUPA20120229_073323_000065273112_00135_52304_3711-v01.nc.bz2

and are available from the MERCI archive. If you want me to provide one, then what do you suggest? These files are ~14 Mb.

I tried again with a different file and am still seeing the problem I reported. Another thing to consider is that I am using pconvert on a 32-bit Linux system and not from Windows like the test you did.

I hope you can try again to replicate this problem.

Regards,

Pauline

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-02-28T09:06:40Z

Hi Norman, Olga,

Thanks for helping!

The debug output was from 4.9.0.1.

It looks like you have used an ATS NR product, rather than an L2P product (which is based on an NR product...).

To grab one from the MERCI archive, enter through the L2P directory, rather than /aatsr. I wouldn't want to send you one, since they are quite large. It would be a .nc.bz2 file that you will need to test this.

Danke schön,

Pauline

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-02-27T16:30:04Z

Hallo Pauline

I don't find any difference between pconvert in BEAM 4.6.1 and 4.9.0.1 by using an AATSR L2P netcdf file (see attached doc).
(commando: pconvert.bat -d -f jpg -b 8,9,1 ATS_NR__2PNPDK20101103_144346_000060673096_00097_45370_7126.nc
where the bands are 8 = latitude, 9 = longitude, 3 = sst_nadir)

Could you please send us your product?

Debug output for beam-4.9.0.1:
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
creating histogram for band 'latitude'...
Error: Could not load mediaLib accelerator wrapper classes. Continuing in pure Java mode.
Occurs in: com.sun.media.jai.mlib.MediaLibAccessor
com.sun.media.jai.mlib.MediaLibLoadException
beam-debug: Histogram: lower bin index = 6 (less than 1.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 498 (less than 4.0% of pixels skipped)
beam-debug: Histogram: histo sample min = -81.55762401223183; sample max = 79.54607979953289
creating histogram for band 'longitude'...
beam-debug: Histogram: lower bin index = 33 (less than 1.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 432 (less than 4.0% of pixels skipped)
beam-debug: Histogram: histo sample min = -156.68702286481857; sample max = 124.4021161198616
creating histogram for band 'sst_nadir'...
beam-debug: Histogram: lower bin index = 335 (less than 1.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 464 (less than 4.0% of pixels skipped)
beam-debug: Histogram: histo sample min = 21303.83203125; sample max = 29572.54296875
creating RGB image...
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'latitude'
beam-debug: RasterDataNode.processRasterData: numReadsMax=10, readBufferLineCount=4096
beam-debug: RasterDataNode.processRasterData: done
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'longitude'
beam-debug: RasterDataNode.processRasterData: numReadsMax=10, readBufferLineCount=4096
beam-debug: RasterDataNode.processRasterData: done
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'sst_nadir'
beam-debug: RasterDataNode.processRasterData: numReadsMax=5, readBufferLineCount=8192
beam-debug: RasterDataNode.processRasterData: done
writing RGB image to 'ATS_NR__2PNPDK20101103_144346_000060673096_00097_45370_7126.jpg'...

RE: pconvert changed from 4.6.1 to 4.9.0.1?

2012-02-27T15:50:03Z

Pauline Cocevar:

Hello Beam Team,
Debug output:

reading file 20120204-ATS_NR_2P-UPA-L2P-51941.nc
creating histogram for band 'lat'...

Dear Pauline, the debug output is from BEAM 4.9, right?
-- Norman

pconvert changed from 4.6.1 to 4.9.0.1?

2012-02-27T12:35:30Z

Hello Beam Team,

We are seeing a difference between pconvert in BEAM 4.6.1 and 4.9.0.1 when applied to an AATSR L2P netcdf file.

The following command was used in both releases:

> pconvert.sh -f jpg -b 1,2,3 20120204-ATS_NR_2P-UPA-L2P-51941.nc

where the bands are 1 = latitude, 2 = longitude, 3 = sea surface temperature.

4.6.1 gave output like the left pic on the attached doc.

4.9.0.1 gives output like the right pic on the attached doc.

We prefer 4.6.1's output since this gives us an indication of problems in lat/lon which can occur in L2P files.

Can you tell us what has changed, and if we can retrieve the old style of jpeg, please?

Many thanks,
Pauline Cocevar

(Other info:
Using the debug option, the only thing odd that I find in 4.9.0.1 is that the SST range is not realistic.

When we VISAT the file, the min/max are correct for lat and lon, but the
SST statistics chart shows min 277.01 K, max 305.180

Debug output:

reading file 20120204-ATS_NR_2P-UPA-L2P-51941.nc
creating histogram for band 'lat'...
beam-debug: Histogram: lower bin index = 0 (less than 0.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 511 (less than 0.0% of pixels skipped)
beam-debug: Histogram: histo sample min = -43.755001068115234; sample max = 83.85299682617188
creating histogram for band 'lon'...
beam-debug: Histogram: lower bin index = 0 (less than 0.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 511 (less than 0.0% of pixels skipped)
beam-debug: Histogram: histo sample min = -179.9969940185547; sample max = 179.9980010986328
creating histogram for band 'sea_surface_temperature'...
beam-debug: Histogram: lower bin index = 10 (less than 0.0% of pixels skipped)
beam-debug: Histogram: upper bin index = 511 (less than 0.0% of pixels skipped)
beam-debug: Histogram: histo sample min = 1701.00390625; sample max = 3157.0
creating RGB image...
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'lat'
beam-debug: RasterDataNode.processRasterData: numReadsMax=7, readBufferLineCount=4096
beam-debug: RasterDataNode.processRasterData: done
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'lon'
beam-debug: RasterDataNode.processRasterData: numReadsMax=7, readBufferLineCount=4096
beam-debug: RasterDataNode.processRasterData: done
beam-debug: RasterDataNode.processRasterData: Quantizing raster 'sea_surface_temperature'
beam-debug: RasterDataNode.processRasterData: numReadsMax=4, readBufferLineCount=8192
beam-debug: RasterDataNode.processRasterData: done
writing RGB image to '20120204-ATS_NR_2P-UPA-L2P-51941.jpg'...
beam-debug: AbstractProductReader.close(): org.esa.beam.dataio.netcdf.DefaultNetCdfReader[input=20120204-ATS_NR_2P-UPA-L2P-51941.nc])