Change HMI data source - dynamic range too small

[dpshelio]

Originally reported by: pohuigin (Bitbucket: pohuigin, GitHub: pohuigin)

---

currently we are using jpg images and giving them a dummy header. originally this was to save time downloading huge files. but it turns out that the dynamic range they use to make these images is so small that they are useless for max millennium...

So, I think we should switch to the FITS images: http://sdowww.lmsal.com/sdomedia/SunInTime/2013/07/16/fblos.fits or http://sdowww.lmsal.com/sdomedia/SunInTime/mostrecent/fblos.fits

and scale them to ~[-700,700], or whatever makes them match MDI.

Current: http://solarmonitor.org/full_disk.php?date=20120203&type=shmi_maglc&indexnum=1

Should look more like this: http://solarmonitor.org/full_disk.php?date=20070203&type=smdi_maglc&region=&indexnum=1

---

• Bitbucket: https://bitbucket.org/solarmonitor/solarmonitoridl/issue/13

[dpshelio]

Original comment by David PS (Bitbucket: dvdgc, GitHub: Unknown):

---

This issue is solved by pull request #9

[dpshelio]

Original comment by David PS (Bitbucket: dvdgc, GitHub: Unknown):

---

No, that was it :) I'm happy you find it looking good!! Then we can close this issue.

Cheers! David

[dpshelio]

Original comment by pohuigin (Bitbucket: pohuigin, GitHub: pohuigin):

---

OK, if you are specifically asking about the source of data for HMI as pointed to in get_hmi_latest.pro, and what scaling should work, then +-150 should work, as it will be getting (essentially calibrated) HMI magnetogram FITS files.

Scaling looks to be OK at the moment so is there a current problem that you wanted me to take a look at?

As far as I can see, everything looks good!

[dpshelio]

Original comment by David PS (Bitbucket: dvdgc, GitHub: Unknown):

---

pull request #9 (not the issue), has also a scaling addition (otherwise the image would be all white) between -150 and 150 - I've just copy/paste from mdi.

[dpshelio]

Original comment by pohuigin (Bitbucket: pohuigin, GitHub: pohuigin):

---

Hi David, I don't see the connection between issue #9 (the roll correction, which is done in arm_fd.pro) and get_hmi_latest.pro (which just downloads the data).

It appears we are getting HMI data now, so what problem with regards to issue #9 still remains?

Cheers,

[dpshelio]

Original comment by David PS (Bitbucket: dvdgc, GitHub: Unknown):

---

Please, @pohuigin could you check what's the situation now with the source used in get_hmi_latest.pro and update pull request #9 accordingly.

Thanks!