Topics

XRIT2PIC.


Douglas Deans
 

On 01/09/2020 21:34, Ernst Lobsiger via groups.io wrote:
Rob and Douglas
I can confirm that xrit2pic latest version works here on Windows 10.
I can make images of GOES, Himawari, Meteosat 0°, RSS, IODC.
I can also make images of Metop-A/B/C and NOAA-19 as disseminated
in EUMETSAT-Data-Channel-1. File format says "Raw1 10", "Raw2 10".
When I want to make the same Metop-B/C from channel EPS-10 I have
file format "Raw3 14". When pressing "Preview" xrit2pic crashes. It
yet opens the preview window and then just silently disappears ...
Regards
Ernst
xrit2pic version 2020.3b
Build at Aug 27 2020, 15:36:46
Windows 10 PRO 2004
=====================================================================

Thanks Ernst. That only adds to the puzzle.
I am using windows 10 Home 2004.

Regards,
Douglas.


Ernst Lobsiger
 

Rob and Douglas

One last hint today:

When I want to directly export *.jpg a single Metop-C VIS006 segment from EUMETCast channel EPS-10 (Format "Raw3 14")
an error message window pops up saying: "JPEG Error! Maximum supported image dimension is 65500 pixels".
When I close this Error window by clicking its "close" button, xrit2pic silently disappears ...

Good night
Ernst


R. Alblas
 

Ernst and Douglas,
About METOP EPS-10: that works fine here, also on Windows-7.
How many segments does this contain? Open the tree, one level, so you can see the 5 channels; in column 'Segm / orbit' you ca see '8 of 8' or something like that.
Note that EPS is 1080 lines per segment; with 61 segments that would be an image of 65880x2048!
I have to check the plate carree / mercator projections. (To be honest, managing polar satellites are not the strongest capabilities of xrit2pic).
 

About MSG problem: For Wavelet a temp. file is generated, to do the decompression. This file is generated in '\', which is the top-directory of the used disk (e.g. C:\). (temp. name and place are determined by OS, not by the program). Maybe here something goes wrong. You may try to start xrit2pic at a location which is writeable for the user (start-icon, properties, 'Start in').

Regards,
Rob.


On 01-09-2020 23:01, Ernst Lobsiger via groups.io wrote:
Rob and Douglas

One last hint today:

When I want to directly export *.jpg a single Metop-C VIS006 segment from EUMETCast channel EPS-10 (Format "Raw3 14")
an error message window pops up saying: "JPEG Error! Maximum supported image dimension is 65500 pixels".
When I close this Error window by clicking its "close" button, xrit2pic silently disappears ...

Good night
Ernst


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

Ernst and Douglas,
[]
About MSG problem: For Wavelet a temp. file is generated, to do the decompression. This file is generated in '\', which is the top-directory of the used disk (e.g. C:\). (temp. name and place are determined by OS, not by the program). Maybe here something goes wrong. You may try to start xrit2pic at a location which is writeable for the user (start-icon, properties, 'Start in').

Regards,
Rob.
=============================

Rob,

Using the top level e.g. C:\ is often not allowed in Windows for better security. It would be better to use the path pointed to by the TMP or TEMP environment variable, or to use a path which you create and know is writeable. Perhaps that what you are doing and Douglas's path is wrong?

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


Ernst Lobsiger
 

On Wed, Sep 2, 2020 at 12:13 AM, R. Alblas wrote:
How many segments does this contain? Open the tree, one level, so you can see the 5 channels; in column 'Segm / orbit' you ca see '8 of 8' or something like that.
Note that EPS is 1080 lines per segment; with 61 segments that would be an image of 65880x2048!
Rob

I think I just selected 1 single Metop segment. But with unexpected results. See attached PDF.

Regards
Ernst


R. Alblas
 

On 02-09-2020 09:49, David J Taylor GM8ARV 🏴 🇪🇺 via groups.io wrote:
Ernst and Douglas,
[]
About MSG problem: For Wavelet a temp. file is generated, to do the decompression. This file is generated in '\', which is the top-directory of the used disk (e.g. C:\). (temp. name and place are determined by OS, not by the program). Maybe here something goes wrong. You may try to start xrit2pic at a location which is writeable for the user (start-icon, properties, 'Start in').

Regards,
Rob.
=============================

Rob,

Using the top level e.g.  C:\  is often not allowed in Windows for better security.  It would be better to use the path pointed to by the TMP or TEMP environment variable, or to use a path which you create and know is writeable.  Perhaps that what you are doing and Douglas's path is wrong?

Cheers,
David
David,
Yes, I think that's going wrong. But, as I said, location and name are not determined by "me" (xrit2pic) but by the OS. (Of course xrit2pic should give an error message if that fails, I'll see why that doesn't happen.) Anyway, I'll add a debug facility to check these kind of things.

Regards,
Rob.


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

David,
Yes, I think that's going wrong. But, as I said, location and name are not determined by "me" (xrit2pic) but by the OS. (Of course xrit2pic should give an error message if that fails, I'll see why that doesn't happen.) Anyway, I'll add a debug facility to check these kind of things.

Regards,
Rob.
========================================

OK, Rob.

So I guess you use both GetTempPath and GetTempFileName. That should be fine.

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


R. Alblas
 

Ernst,
Yes, that's one segment.
But I see that it generates in your case a pic. size with all segments but only the first one filled in.
So a jpeg of 480*1080 = 518400 rows is generated (with just 1080 rows containing image data).

I'll fix that.

Regards,
Rob.


On 02-09-2020 10:41, Ernst Lobsiger via groups.io wrote:
On Wed, Sep 2, 2020 at 12:13 AM, R. Alblas wrote:
How many segments does this contain? Open the tree, one level, so you can see the 5 channels; in column 'Segm / orbit' you ca see '8 of 8' or something like that.
Note that EPS is 1080 lines per segment; with 61 segments that would be an image of 65880x2048!
Rob

I think I just selected 1 single Metop segment. But with unexpected results. See attached PDF.

Regards
Ernst


R. Alblas
 

On 02-09-2020 11:59, David J Taylor GM8ARV 🏴 🇪🇺 via groups.io wrote:
David,
Yes, I think that's going wrong. But, as I said, location and name are not determined by "me" (xrit2pic) but by the OS. (Of course xrit2pic should give an error message if that fails, I'll see why that doesn't happen.) Anyway, I'll add a debug facility to check these kind of things.

Regards,
Rob.
========================================

OK, Rob.

So I guess you use both GetTempPath and GetTempFileName.  That should be fine.

Cheers,
David
Well, I actually use tmpnam() (C), but maybe here is the problem. It seems to use \\ as location, instead of something under the Users area.  On WIndows-7 it creates a file like \s2c4.
That may work most of the times (apparently even under Windows-10), but not always.
Hm, I'll have a look at this.

Thanks,
Rob.


Douglas Deans
 

Rob and Ernst,
I can confirm that I have been able to get a GOES 16 image to work fine today on XRIT2PC so it does look as if the issue with MSG is the decompression dll not functioning for whatever reason.

Noted your last mails on the subject and await any further help.
Thanks again.

Regards,
Douglas.


R. Alblas
 

Douglas, Ernst,

I have now adapted xrit2pic so you get a decent error message in case of temp. file errors.
Also, selecting one or limited amount of segments will now generate a picture just for that part, so no 500000 -row pictures with just 0.2% filling.
This version is not available yet, first I want to fix generation of temp files the right way for Windows.

Regards,
Rob.

On 02-09-2020 15:28, Douglas Deans via groups.io wrote:
Rob and Ernst,
I can confirm that I have been able to get a GOES 16 image to work fine today on XRIT2PC so it does look as if the issue with MSG is the decompression dll not functioning for whatever reason.

Noted your last mails on the subject and await any further help.
Thanks again.

Regards,
Douglas.



David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

Well, I actually use tmpnam() (C), but maybe here is the problem. It
seems to use \\ as location, instead of something under the Users area.
On WIndows-7 it creates a file like \s2c4.
That may work most of the times (apparently even under Windows-10), but
not always.
Hm, I'll have a look at this.

Thanks,
Rob.
============================

Rob,

Unless your code has changed recently, there's no reason why it shouldn't continue to work. I've not seen any examples of programs failing recently because of temporary name issues.

https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/tempnam-wtempnam-tmpnam-wtmpnam?view=vs-2019

One thing I've done it to designate a directory on a disk other than the main disk, in my case so that any intensive file activity on TMP/TEMP takes place off the SSD main drive, to minimise the wear on the SSD, which why - on one PC:

TEMP=D:\DAVIDS~1
TMP=D:\DAVIDS~1

but on another PC:

TEMP=C:\Users\David\AppData\Local\Temp
TMP=C:\Users\David\AppData\Local\Temp

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


R. Alblas
 

I have altered xrit2pic: 2020.3c. For temp. files it uses now the location pointed to by windows-variable TEMP (thanks, David).

With the non-gui version you can do a test, by typing in a shell:

xrit2pic_cmd -test_temp

and that will show if a temp. file can be created in write mode.
(xrit2pic, so gui mode, does do the same, but in Windows that doesn't output to a shell so that's pretty useless...)

Also, if one or some segments are selected then that will generate an image with just that size.

I hope this works now.

Regards,
Rob.

On 02-09-2020 15:28, Douglas Deans via groups.io wrote:
Rob and Ernst,
I can confirm that I have been able to get a GOES 16 image to work fine today on XRIT2PC so it does look as if the issue with MSG is the decompression dll not functioning for whatever reason.

Noted your last mails on the subject and await any further help.
Thanks again.

Regards,
Douglas.



Douglas Deans
 

On 03/09/2020 14:45, R. Alblas wrote:
I have altered xrit2pic: 2020.3c. For temp. files it uses now the location pointed to by windows-variable TEMP (thanks, David).
With the non-gui version you can do a test, by typing in a shell:
xrit2pic_cmd -test_temp
and that will show if a temp. file can be created in write mode.
(xrit2pic, so gui mode, does do the same, but in Windows that doesn't output to a shell so that's pretty useless...)
Also, if one or some segments are selected then that will generate an image with just that size.
I hope this works now.
Regards,
Rob.
===========================================================================

Perfect Rob. MSG images now showing. Many thanks for the fix.

One other question if I may. I am now starting to look at other data such as Himawari. Will I have to put a bz2 unzip program into the folder with XRIT2PIC exe or is there some way to tell the program to look elsewhere.

Thanks again,
Douglas.


R. Alblas
 

Douglas,

xrit2pic has bunzip2 built-in. It reads both zipped and unzipped Himawari-files. You can also define to keep files which are unzipped as such, so they don't need to be unzipped again; see in:
Preferences, Files, zipping.

Good to hear that it works now.

(Btw, there used to be a xrit2pic mailing list which would be better for these kind of questions, but it seems to be dead now. Long time no activity there...)

Regards,
Rob.

On 03-09-2020 17:18, Douglas Deans via groups.io wrote:
On 03/09/2020 14:45, R. Alblas wrote:
I have altered xrit2pic: 2020.3c. For temp. files it uses now the location pointed to by windows-variable TEMP (thanks, David).

With the non-gui version you can do a test, by typing in a shell:

xrit2pic_cmd -test_temp

and that will show if a temp. file can be created in write mode.
(xrit2pic, so gui mode, does do the same, but in Windows that doesn't output to a shell so that's pretty useless...)

Also, if one or some segments are selected then that will generate an image with just that size.

I hope this works now.

Regards,
Rob.
===========================================================================

Perfect Rob. MSG images now showing.  Many thanks for the fix.

One other question if I may. I am now starting to look at other data such as Himawari. Will I have to put a bz2 unzip program into the folder with XRIT2PIC exe or is there some way to tell the program to look elsewhere.

Thanks again,
Douglas.


Douglas Deans
 

On 03/09/2020 16:33, R. Alblas wrote:
Douglas,
xrit2pic has bunzip2 built-in. It reads both zipped and unzipped Himawari-files. You can also define to keep files which are unzipped as such, so they don't need to be unzipped again; see in:
Preferences, Files, zipping.
Good to hear that it works now.
(Btw, there used to be a xrit2pic mailing list which would be better for these kind of questions, but it seems to be dead now. Long time no activity there...)
Regards,
Rob.
===============================================================================

Ah that is interesting because yesterday evening I tried some Metop and Himawari data and kept getting a 'preview not available' warning.
Now trying those data with the new revised program all is working fine.
Perhaps they also use a temp folder as with MSG data so the changes have solved that issue as well. Excellent.

Thanks for your help and I will now try and master the advanced colour settings etc.

Regards,
Douglas.


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

Douglas,
[]
(Btw, there used to be a xrit2pic mailing list which would be better for
these kind of questions, but it seems to be dead now. Long time no
activity there...)

Regards,
Rob.
=========================================

Rob,

As it makes it easier for folk to follow one list rather than multiple, it's fine for XRIT postings here. We cover Windows and Linux, and there's been quite a lot of Python/PyTroll postings recently. Your choice, of course.

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


R. Alblas
 

Yes, this part uses the same temp-file construction. So 2 fixes in one...

Rob.

On 03-09-2020 18:06, Douglas Deans via groups.io wrote:
On 03/09/2020 16:33, R. Alblas wrote:
Douglas,

xrit2pic has bunzip2 built-in. It reads both zipped and unzipped Himawari-files. You can also define to keep files which are unzipped as such, so they don't need to be unzipped again; see in:
Preferences, Files, zipping.

Good to hear that it works now.

(Btw, there used to be a xrit2pic mailing list which would be better for these kind of questions, but it seems to be dead now. Long time no activity there...)

Regards,
Rob.
===============================================================================

Ah that is interesting because yesterday evening I tried some Metop and Himawari data and kept getting a 'preview not available' warning.
Now trying those data with the new revised program all is working fine.
Perhaps they also use a temp folder as with MSG data so the changes have solved that issue as well.  Excellent.

Thanks for your help and I will now try and master the advanced colour settings etc.

Regards,
Douglas.