How does GONG get disk and limb simultaneously?

[Heliotown]

I wonder how the online GONG h-alpha telescopes get the full disk and the limb view simultaneously?

Thomas

[Spectral Joe]

Partly bandpass, partly processing. The GONG telescope filters are 0.4 Angstrom wide and the images are processed heavily before you see them. With narrower bandpasses you can see edge and disk details in one image. Another problem with many mass produced H alpha telescopes is continuum leakage, which reduces disk contrast further still. The GONG telescopes are engineered for performance, not a compromise of performance versus price, and have correspondingly better performance.

[Montana]

It looks like they do 2 exposures to me, then clip and paste them together, I could be very wrong though. It must be automatic software.

Alexandra

[Heliotown]

Spetral Joe wrote: "
Partly bandpass, partly processing. The GONG telescope filters are 0.4 Angstrom wide and the images are processed heavily before you see them. With narrower bandpasses you can see edge and disk details in one image. Another problem with many mass produced H alpha telescopes is continuum leakage, which reduces disk contrast further still. The GONG telescopes are engineered for performance, not a compromise of performance versus price, and have correspondingly better performance.
"

Hi Joe,

I thought that narrower bandwidth, like .4 angstroms, enhances disk features and contrast but attenuates limb activity like proms.

To me it looks like the GONG disk image is .4 or .5 and their outer limb prominence ring area is .7 or so. I am just guessing though.

I wonder if they have posted information on their process anywhere? I can't find any.

Thomas

[solarchat]

I'm no expert but you can't see proms lower than about .5 A .

I get a slightly better than gong image with my 12 bit and 16 bit cameras all the time. Thd DMKs are 8 bit and therefore not capable of displaying the same images simultaneously

[Pedro]

http://halpha.nso.edu/index.html#top

best

[Cschur]

Dang Pedro, I was going to answer this and your too dang fast! I wonder what a 16 bit camera that will do 1600 x 1200 would cost - with fast 30fps frame rates? My SBIG ST10xme is 16 bit, but the shutter is horridly slow at 1/10 second and good for less than 10000 frames. At that rate, a few good sized solar shots would trash my shutter!

[Pedro]

Hi Chris:

I was wondering the same thing. Maybe Stephen can tell us which 16 bit camera he is using.

I am looking more and more at the FLEE CCD cameras

http://www.ptgrey.com/

best regards

[solarchat]

Thosr ate the specs for the Grasshopper series. That's what I use. I don't do 2 panel proms surface, there mostly all single Avis

[Pedro]

Thanks Stephen

Thsi is new to me. I didin't know that the PG cameras had 8, 12, 16 and 24-bit digital data output... Something to explore. We need a new computer with USB3...

best

[Spectral Joe]

0.35 Amgstrom bandpass, single exposure:

Attached files

[Spectral Joe]

Here's the same day, from Big Bear's patrol telescope (not the dedicated GONG one). 100mm aperture, 0.25 Angstrom Zeiss filter, 12 bit Pulnix camera, single exposure.

Attached files

[earl]

I think they are 2 shots. I looked closely at all 6 images cand I can see the sharp difference between the disk and limb. A single shot would have a smoother transition.

[Pedro]

surely 2 shots with different settings.

best

[mattwastell]

Sometimes you can see the images jumping and I believe you can see a surface disc and prominence disc!?

[Merlin66]

I, as mentioned previously, using the DS SM60 (0.4A) need a separate exposure for the surface and a re-tune (to chase the doppler) for the proms. I think we all agree, unless you're using >1.5A bandwidth you'll loose some of the prom detail with a single exposure....

[Spectral Joe]

I wasn't clear enough earlier, this image was taken on August 6, 2010 with a system having 0.35 Angstrom bandpass and very good spectral purity. It is a single exposure, the only processing done was a flatfield and some contrast enhancement. The proms are just as evident in the raw image. I know what the parameters were because I took the image. The very good spectral purity part is really important, because most etalon systems have off-band light leakage that reduces disk contrast. Double stacking sometimes improves this, but not always. With this extra light in the disk it's hard to get enough exposure on the proms without overexposing the disk. Images from GONG, Big Bear's patrol telescope and ISOON (http://nsosp.nso.edu/node/45) show disk and proms in single exposures. The difference is that they were engineered for the task and are a lot more expensive than the instruments we usually discuss here. If you want to know for sure that GONG takes a double exposure to enhance the proms, call or email them,the system is run by NSO in Tucson, but my bet is that they are single exposures.

Attached files

[solarchat]

.4 A SM60 DS?

[Merlin66]

Stephen,
yes, a double stacked SM60 external etalon pair on an ED80 refractor. It gets down to the 0.4A mark.

[marktownley]

Partly bandpass, partly processing. The GONG telescope filters are 0.4 Angstrom wide and the images are processed heavily before you see them. With narrower bandpasses you can see edge and disk details in one image. Another problem with many mass produced H alpha telescopes is continuum leakage, which reduces disk contrast further still. The GONG telescopes are engineered for performance, not a compromise of performance versus price, and have correspondingly better performance.

Question is then, what can we do as amateurs to reduce continuum leakage in our Ha telescopes???

[oldfrankland]

Here's an informative poster on the addition of the Daystar filters to the GONG system. Indeed, a 0.4A passband filter.

http://gong.nso.edu/gallery/disk2k10/da ... d11/jh.pdf

[marktownley]

Good find Jim! That pretty much says all the details one could want, some interesting stuff about the processing side too...

[Spectral Joe]

Mark
One possible solution is the 1.5 Angstrom interference filter often found on Ebay. The issue is the quality of the blocking filters in packaged systems, a recent set of bandpass plots you posted reveals that even the estimates I thought were pessimistic about the characteristics of blocking filters actually turned out to be optimistic. The 5% transmission immediately blueward of Ha in the Coronado blocker is very bad. The photospheric light at the wavelengths of the etalon modes where the blocker passes that much is 5 times as bright as the Ha core. It adds up fast. Anything that blocks this light helps. One of my theories has been that the reason for the performance of the Luntanado combination is that the design difference between the two filters results in the etalon modes of the two filters not coinciding, and the filters act as blockers for each other. I suggest that you don't want "matched" filters for a double stack, but just the opposite. I would really like to see someone do complete system transmission plots of available telescopes. Until then, Excel and I make the best guesses possible using what data is available.

Joe

[marktownley]

Mark
One possible solution is the 1.5 Angstrom interference filter often found on Ebay. The issue is the quality of the blocking filters in packaged systems, a recent set of bandpass plots you posted reveals that even the estimates I thought were pessimistic about the characteristics of blocking filters actually turned out to be optimistic. The 5% transmission immediately blueward of Ha in the Coronado blocker is very bad. The photospheric light at the wavelengths of the etalon modes where the blocker passes that much is 5 times as bright as the Ha core. It adds up fast. Anything that blocks this light helps. One of my theories has been that the reason for the performance of the Luntanado combination is that the design difference between the two filters results in the etalon modes of the two filters not coinciding, and the filters act as blockers for each other. I suggest that you don't want "matched" filters for a double stack, but just the opposite. I would really like to see someone do complete system transmission plots of available telescopes. Until then, Excel and I make the best guesses possible using what data is available.

Joe

Very interesting Joe! You have got my mind ticking over with various ideas... Be good to hear more of you thoughts on here

[Heliotown]

From the poster information:
"Polarizing beam splitter diverts otherwise unused 656 nm light through two afocalizing lenses"

So then, one light beam but the beam is split into two. So essentially this is two different views that they merge for one image.

I think I am reading this right.

Thomas

[marktownley]

From the poster information:
"Polarizing beam splitter diverts otherwise unused 656 nm light through two afocalizing lenses"

So then, one light beam but the beam is split into two. So essentially this is two different views that they merge for one image.

I think I am reading this right.

Thomas

I reckon you've got it there Thomas...

[Spectral Joe]

The beamsplitter is "stealing" light that isn't used for GONG's primary mission (GONG has no use for H alpha) and sending it through a 0.4 Angstrom Daystar filter to a DVC 2K X 2K camera. The edges of their images show all the signs of a wavelet shadow artifact. Still firm in my conviction, one camera, one image.

[marktownley]

The beamsplitter is "stealing" light that isn't used for GONG's primary mission (GONG has no use for H alpha) and sending it through a 0.4 Angstrom Daystar filter to a DVC 2K X 2K camera. The edges of their images show all the signs of a wavelet shadow artifact. Still firm in my conviction, one camera, one image.

You just think different processing is used for the limb and the disk from the same image - then superimposed?

[Spectral Joe]

Just got off the phone with Jack Harvey (the first author on the poster Jim posted the link to) and here's the deal: a single image is taken, two processing filters are applied, a wavelet filter to sharpen the details and a radial intensity filter to enhance off-limb details. The shadow and step at the limb are an artifact of the wavelet filter.