i got a reply
-----------------------There are different types of processing that occurs. First, there is some
processing done on the spacecraft itself, which is explained in section
10.2 of the SECCHI documentation from the Principal Investigator:
As the data is downlinked, it comes down in packets that must be
re-assembled into images; as you look through the archive, you might
come across images that have large black regions in them; these are
from lost packets.
The images are then saved as a 'level 0' product, with the appropriate
scientific metadata attached using the FITS standard.
The images are next calibrated to create a 'level 1' product, which is an
attempt to remove known sensor effects from the data, such as correcting
for sensor degradation over time or other problems that might cause an
uneven response from different pixels.
To generate the JPEG files, the data is fit to a log scale to bring out features
the scientists are looking for and a color table is applied to mark the image
as being from a specific filter, the images are rotated so the top of the image
is solar north and the sun is in the center of the image, labels are inserted to
mark what the time and origin of the file are, and the image is saved out in a
number of resolutions, each one with JPEG compression applied.
No "sharpening" is done; the improved resolution is because the images
were transferred at the original resolution without the heavy levels of
compression that's required for the space weather beacon's lower
There *are* some science products from other missions that undergo
correction for point-spread, which is a type of sharpening:
Some of the data from SDO have undergone this processing, they refer to
the images as 'level 1.5' as they've undergone a level of processing between
correcting for the sensor (level 1) and processing to physical units (level 2):
I have no idea if the JPEG images on the SDO website are based on the
level 1 or level 1.5 processing.
> could i get hold of the image 20110125_094530_n7euB_195 before processing ?
> i have the n4 and the processed n7 but i would really like to see the
> inbetween @ 2048.
You can download the level 0 version of the file from:
You will need a FITS file reader, such as DS9 to view the file:
However ... the image is only 512 x 512. It's been upscaled for the JPEGs.
(I have no idea why they generate images of a higher resolution than the
source image; I;d have to ask the scientists in charge of the processing
If you view the headers on the image, you'll see:
NAXIS = 2 /
NAXIS1 = 512 /
NAXIS2 = 512 /
Which means the data is a two dimensional array, with each dimension
If you want the full size image that has come down before it's been calibrated
and converted to JPEG, see:
Note on this one, the headers have:
NAXIS = 2 /
NAXIS1 = 2048 /
NAXIS2 = 2048 /
... and the file size is 8MB, as opposed to 512 kB for the 'n7' version.