James Brown <james@...>
A test of HRV using wavelet lossless compression will take place from 13:00 29/10/2003 until 13:00 30/10/2003.
These will be disseminated in place of the HRIT JPEG compressed HRV products.
Image data will be switched from the IQGSE to the IMPF on the 28/10/2003 at 13:00. There may be a short interruption to the
SEVIRI dissemination. Some differences are expected in the headers, image quality and compression efficiecy.
Looking at the news for the week ahead - some more teasers for David - I wonder if your thoughts about HRV have been listened to David?
Incidentally, does anyone understand the gain changes with reference to the channel numbers - I didn't even know there was a channel 13.4 !!
David J Taylor
Looking at the news for the week ahead - some more teasers forDavid - I wonder if your thoughts about HRV have been listened to
So do I, James, but I have no inside information about this. I had
suggested either reducing the compression level or bit shifting the
whole data to be 10-bit data left justified in the 12-bit field
offered by the compressed JPEG channel. At present, the data is
right-justified which leads to an accuracy loss, according to my
tests. If the lossless Wavelet data occupies a similar bandwidth to
the lossy JPEG data, then by changing there is no ill effect. As
you know, the HRIT channel is already about 1.7Mb/s over DVB versus
1.0Mb/s over a direct link, so there is more flexibility than before.
Incidentally, does anyone understand the gain changes withreference to the channel numbers - I didn't even know there was a
channel 13.4 !!
The 13.4um data is transmitted as channel number 11 - if you hold
your mouse over the thumbnail image the pop-up hint shows the
spectral range for the channel.