Back in May @Samrau started a thread on imaging the sun and @Echound mentioned spectroheliographs (SHG for short). It just so happens I had spent months researching building one (the Sol’Ex Project). Today I am starting a new thread on spectroheliographs. Their abilities and operation are so much different than regular solar filter or etalon based solar scopes it requires a separate discussion. One important fact to point out is spectroheliographs are only for imaging the sun, they can not be used visually.

I got started months ago but instead of building one I decided to order a commercially made SHG. This has only become possible recently. Building from scratch isn’t a bad option but it wouldn’t be much faster since you have to order all the optical components at the very least. I received my order this week, ahead of the original estimated delivery date. After spending a few days imaging with the MLAstro SHG 700 model I had to share the first results. Its that good. Maybe I’ll inspire a few to try this type of solar imaging out.

Now anyone that has any experience in astronomy or astrophotography knows the moment they get something new (camera, filter, t-shirt, anything) the weather gets cloudy. Yes that has happened to me of course. But there has been some usable sky to get started experimenting and learning. So what I'm sharing are my first efforts with less than ideal conditions. Even with that the results are exciting.

Two pictures below, HAlpha wavelength, both from the same scan taken today. These are resized and compressed for posting here. The captions link to Telescopius where you can examine larger versions. Frankly its amazing how well they turned out as I was shooting through thin clouds. These are essentially unprocessed, just what the scan processing software did. Much more could be done with stacking multiple images and additional processing.

Colorized HAlpha (automatically done by processing software):

A negative version of same image is also generated by the processing software. I’m particularly fond of negative versions as they seem to have a more 3D effect to them:

I’ve started with the Halpha wavelength to learn and work out my settings . Yet that is just one of many different wavelengths you can image with a spectroheliograph. You just dial in a different wavelength, refocus and scan. While spectroheliographs are not new, they were invented over a century ago, the recent diy projects and software have spawned a new type of amateur solar imaging. Affordable solar imaging compared to traditional dedicated solar telescopes. Now that a commercial SHG model is available this is going to get popular.

In the next post I will go over the hardware and related requirements for spectroheliograph imaging. Then I will post on the software required to scan the sun and create an image.

How long does each scan take?

This is very interesting. How can I get my hands on all of this? If I build it myself, I'll still have a hard time finding the optical components. I'll have to research this further.

This is so cool; to be able to see the topology of the surface of the Sun!

Edited July 3Jul 3 by Archmonoth

17 hours ago, astroHoward said:

Yet that is just one of many different wavelengths you can image with a spectroheliograph. You just dial in a different wavelength, refocus and scan.

This is all really awesome! What are the limitations to the wavelengths you can use? Could you e.g. get some images using the He D3 line at 5875.61 Å as well? And/or the Ca II K and H lines at 3933.663 Å and 3968.469 Å respectively? Or maybe H-beta at 4861.34 Å as well? Definitely would like to learn more about the specifics of such instruments myself, maybe even get my hands on one at some point, heh.

Also, how is the colorization done? Does it do it based on the wavelength to match the true color of it as well as possible? Would definitely love to see the Ca II lines in the true purple color and the light blue of H-beta!

How exactly does a spectroheliograph work? This is Minh’s description off MLAstro:

A slit is used to capture a small slice of sunlight. This light passes through a collimator, which turns it into parallel rays. These rays then hit a diffraction grating, which spreads the light into a spectrum. The spread beam is refocused by a lens onto a camera sensor. By adjusting the tilt of the grating, you can choose which part of the spectrum is focused on the sensor, allowing you to observe the Sun in different wavelengths (such as H-alpha, Sodium D, Helium D3, or Ca K).

The Sun’s details and contrast are captured in slices, recorded as a video while the mount slowly slews, causing the Sun’s image to "scan" through the slit. Specialized software then reconstructs the full image of the Sun from this video data

You can build your own spectroheliograph and the Sol’Ex project published last year is what got me started. Led by Christian Buil this is a great project combining 3D printed hardware and open source software. There is a company that supplies the optical kit. The Sol’Ex design has been around a few years and has been revised and improved.

A second design similar to Sol’Ex uses 3D printing and old Pentax camera lenses with M42 base as the optical elements. There are several variations of this and the one I was most interested in is on Printables. These older Pentax lens are very good optical quality for their age. This would be a good budget build for those able to source used lens but is significantly heavier (about 1800g) than original Sol’Ex.

Anyone with strong diy skills can build a relatively inexpensive spectroheliograph with either of these designs. And that is the only way to have one until now.

Somewhere around the start of this year MLAstro began shipping their design based on the Sol’Ex but slightly modified. Their current commercial model, SHG 700. has a aluminum body, custom optical elements and micrometers for focus and tuning. This is a solid design. However it weighs about twice as much as the original Solex 3D printed versions.

A spectroheliograph needs to be attached to a telescope to capture the sun. This is almost exclusively going to be a refractor type telescope. The original Sol’Ex design was optimized for smaller telescopes, around 400mm focal length or less. The SHG 700 model can use up to about 730mm focal length. These numbers are based on telescope focal ratio, slit size, camera sensor size and being able to image the full solar disk in one pass. There is some flexibility in this and you would be best to consult other sources to get a thorough understanding of the relationships.

I am using a 80mm F7 telescope which is 560mm focal length. This is a fairly common telescope size and it does not have to be high quality either. While mine is ED type glass even a plain doublet refractor is fine if the focal length is right. So inexpensive telescopes can be used. That said one thing to keep in mind is the focuser on the telescope needs to be sturdy; especially for the heavier versions.

The picture below is what the SHG 700 looks like attached to my small refractor.

The MLAstro SHG 700 as pictured with camera weighs 1300g. That is a fair bit of weight, certainly more than the standard Sol’Ex 3D build (about 700g). This is far more than any diagonal / eyepiece combination would weight. Also note the back focus on my telescope required me to add T2/M42 spacers to reach proper focus. So that weight is leveraged by the added distance. This is why a poor focuser mechanism will have problems. Fortunately this SVBony has a pretty good focuser.

Next you need a camera to capture sun scans and this should be a monochrome planetary (or guiding) type astro camera. They are optimized for reliable high frame rates. Planetary color sensors can be a waste of pixels/resolution in this case on a spectroheliograph but could be adapted if that is what you have. A DSLR or other common digital camera is not going to work at all for scanning (it would however show a neat color spectrum with absorption lines).

The most important thing for good images is matching the pixel size of the camera to the spectroheliograph setup you are using. There is some flexibility here but in general 2 – 4 micron pixels are needed. The Sol’Ex project website goes into detail with examples about this.

Lastly a solid telescope mount is helpful if not required. While its technically possible to let the sun drift across the spectroheliograph the results will not be good. And that would be pretty slow. To take quality images you need a mount that can slew the telescope at 4 – 16x the normal rate. In operation one starts capturing camera data then slews the telescope across the sun. Almost exactly like scanning a document on a flatbed scanner.

To achieve best results you need a equatorial type mount that is polar aligned and slews smoothly. In this day and age there are many options for a reasonable cost. Its possible anyone that does astrophotography already has something that would work. Alternatively, inexpensive or diy mounts could be modified to use the OnStep telescope controller project.

A spectroheliograph works much different than using a solar filter or standard solar telescope. It scans the sun slice by slice. One of the most important parts then is the software that combines these slices into a complete image. That is for the next part.

As a side note my sky was clear and seemed pretty steady (good seeing) this morning. I spent an hour acquiring scans and will have some more examples later. The resolution from this with a little 3 inch telescope is astounding.

3 hours ago, Siyu Zhang said:

How long does each scan take?

Interesting question, right now I am slewing at 8x speed so lets look at a SER log file from today:

StartCapture=2025-07-03T15:54:05.5587927Z
MidCapture=2025-07-03T15:54:16.4845712Z
EndCapture=2025-07-03T15:54:27.4103498Z

So about 22 seconds for full disk using my current settings (camera is running about 130fps with a capture area of 3836x156 pixels).

I'd like to get a higher slew speed but next step is 16x. I don't think the camera can capture the fps to support that as there is a relationship (speed / fps / exposure time). But there maybe adjustments I can do. Also it maybe possible to use scripting to send particular commands to the mount (unlike using hand controller) as it will accept ASCOM type command to set an integer slew rate (like 10x).

Absolutely stunning image! The contrast you achieved is great, it really shows just how narrow of a bandpass spectroheliographs can reach.

I've been hesitant about buying a SHG or dedicated solar telescope myself, as my daytime seeing always seems horrendous.
I wish it were possible to capture solar flares and filament lifts real-time with a SHG too, but I suppose you can make decent animations when making scans in quick succession.

Do you use any (energy rejection) filter at all by the way?

Minh mentioned some people are looking into developing lucky imaging software for SHG scans. Could be really interesting!

Edited July 4Jul 4 by Echound

2 hours ago, Echound said:

Absolutely stunning image! The contrast you achieved is great, it really shows just how narrow of a bandpass spectroheliographs can reach.

I've been hesitant about buying a SHG or dedicated solar telescope myself, as my daytime seeing always seems horrendous.
I wish it were possible to capture solar flares and filament lifts real-time with a SHG too, but I suppose you can make decent animations when making scans in quick succession.

Do you use any (energy rejection) filter at all by the way?

Minh mentioned some people are looking into developing lucky imaging software for SHG scans. Could be really interesting!

Thanks! You are absolutely correct that seeing is the 'weak' spot with a spectroheliograph. So far I think I have had average to poor seeing in the times I've used the SHG 700. Some scans do not look great on the edge and some have been pretty bad. I have only imaged the time frame about an hour before local noon. Theoretically the morning hours should be best but I have a tree in the way that prevents anything earlier (although it is dead and I have incentive to remove it now!). I'd like to experience good seeing at my location so I know what that is like with the SHG.

Be interesting to see improvements in software and that may happen. I certainly think there is room for improvement in the scan -> image conversion software that would help reduce seeing effects. Particularly the disk edge. Right now I see too many edge defects that look to me like they could be better handled. Then again I'm not the one writing the software, although I have some ideas to throw at Cedric on this topic. Unfortunately I'm a pixel peeper and can be rather critical of image defects 😨 But also, maybe I could better tweak my settings and scanning technique. Right now all this is pretty fun to learn, whole new world vs my usual imaging.

I am not currently using any ERF but seriously considering adding a ND filter like the Sol'Ex project uses. I can't reach the 'optimum' exposure time for my pixel/slew rate, even with camera gain at the minimum (100). I know Minh thinks they affect the image but if I can find a good 48mm ND I can add it further up from the SHG (since I need so much focus extension). I'm thinking the reduced light cone angle might be better (less distortion/wavefront error at filter) but that's just speculation and I could be wrong. There is a heck of a lot of energy concentrated on that slit without a ERF, on my 80mm you can not hold you hand at focus point for more than a few seconds. Clearly a high quality front element ERF (expensive) would work good but my budget is shot for the year.

Timelapse of flares and lifts should be possible by scripting the capture. I have seen a SharpCap script a Chinese user submitted on SolarChat that does that (would need Pro SharpCap license) and I plan on trying it out. Because SharpCap scripting can send ASCOM commands to the mount, ie move it, I could also adjust the mount slew speed as I mentioned before, programmatically setting a non standard rate which may help on my exposure time issue. Automation would certainly speed up the capture cadence. I think 1 scan per minute is easily possible with my settings right now and maybe even faster than that if optimized.

Last Part: Spectroheliograph Software

Unlike hardware the software for using a spectroheliograph is no cost, freely available. There are two parts to obtaining a picture like today's Sun below. The first part is getting a spectroheliograph scan of the sun. The second part is taking that scan file and turning it into a normal image.

Sun in H-Alpha 2025.07.04, click here for better version/size. This has some level/curve tweaks and light unsharp mask in GIMP.

Scanning: the software of choice to scan is SharpCap. It has a free version and everyone uses this. Doing a scan is simple in concept: you move the telescope to be just to the east side of the sun (no sun is visible). Then you start a video capture and slew the telescope in RA to the west going across the sun until it is all the way out of the field of view. Stop slew, stop scan and save it.

Now in reality there is a lot that goes into this. Framing, focusing correctly, choosing the right exposure time and frame rate, selecting a region of interest, etc. Fortunately there are some tutorials you can learn from and its not difficult. The results are more then worth the effort to master this part. Its just sort of weird compared to doing the usual astrophotography or auroral photography. I mean this is what the full live camera feed from a SHG with the Sun in it looks like on computer screen:

You are looking at absorption lines (H Alpha most prominent in this case), not a solar disk. Did not think I’d be learning and recognize spectral lines. In operation you are only scanning a small selected region (ROI) around that H Alpha line so the frame rate is much higher (and file size is much smaller).

Once you have a scan done and saved (SharpCap will name it a SER file) the next step is to convert that into one or more images. The original software to do this from the Sol’Ex project is INTI. This is a Python application and it works really well.

Another way to convert SER files is to use JSolex which is a Java based program. This is what I am using right now. It does essentially the same thing as INTI but is really convenient to use and has a few more options. To use it just open the SER file and choose Quick or Full processing and it does the rest. Full processing generates a bunch of images, like this in a folder:

If JSolex has a connection to the internet it will pull down annotation information and make some interesting images with that data. This can all be customized. In fact JSolex has an incredible toolbox including image math functions built in. It even has a bunch of spectral line tools. I’ve only started to scratch the surface using it myself. Whether you use JSolex or INTI at this point you have image files (png/jpg/tiff/fits whatever you want to save) that can be processed in other software (Affinity/GIMP/ Photoshop/Pixinsight).

There are a number of advantages and disadvantages to spectroheliograph imaging. Big advantages include the resolution of images and ability to image many different wavelengths. One disadvantage is how sensitive image reconstruction is to seeing conditions or scan irregularities. Its worth comparing conventional solar imaging with a solar telescope to spectroheliographs and we can talk about that later if there is interest.

This sums up my introduction to spectroheliograph imaging. Like your SpaceWeatherLive T-shirt I probably can not stop talking about it.

On 7/3/2025 at 4:09 PM, Philalethes said:

This is all really awesome! What are the limitations to the wavelengths you can use? Could you e.g. get some images using the He D3 line at 5875.61 Å as well? And/or the Ca II K and H lines at 3933.663 Å and 3968.469 Å respectively? Or maybe H-beta at 4861.34 Å as well? Definitely would like to learn more about the specifics of such instruments myself, maybe even get my hands on one at some point, heh.

Also, how is the colorization done? Does it do it based on the wavelength to match the true color of it as well as possible? Would definitely love to see the Ca II lines in the true purple color and the light blue of H-beta!

I think you probably already read the answer is yes. In fact nearly any line can be imaged, various Fe for sure as well, its amazing. But I suspect some lines are much harder to capture, that is the more obscure ones will be much lower intensity level (on a histogram) and have a poor S/N ratio in a final image. But that's just a guess at this point. There is some advanced tuning you can do in some of the video file to image conversion software like JSolex. For instance you can shift the region of interest in the scan +/- x number of pixels which can allow you to shift to obscure absorption lines. Its pretty incredible what these devices plus software can do, with a small telescope!

I'm not sure how many preset wavelength / colors are in JSolex. I know the major ones area there.

I don't know what spectrohelioscope generated this but Wah! show some of you colors in this video below. Certainly put a lot of aquisition effort into these images:

20 hours ago, astroHoward said:

Thanks! You are absolutely correct that seeing is the 'weak' spot with a spectroheliograph. So far I think I have had average to poor seeing in the times I've used the SHG 700. Some scans do not look great on the edge and some have been pretty bad. I have only imaged the time frame about an hour before local noon. Theoretically the morning hours should be best but I have a tree in the way that prevents anything earlier (although it is dead and I have incentive to remove it now!). I'd like to experience good seeing at my location so I know what that is like with the SHG.

Be interesting to see improvements in software and that may happen. I certainly think there is room for improvement in the scan -> image conversion software that would help reduce seeing effects. Particularly the disk edge. Right now I see too many edge defects that look to me like they could be better handled. Then again I'm not the one writing the software, although I have some ideas to throw at Cedric on this topic. Unfortunately I'm a pixel peeper and can be rather critical of image defects 😨 But also, maybe I could better tweak my settings and scanning technique. Right now all this is pretty fun to learn, whole new world vs my usual imaging.

I am not currently using any ERF but seriously considering adding a ND filter like the Sol'Ex project uses. I can't reach the 'optimum' exposure time for my pixel/slew rate, even with camera gain at the minimum (100). I know Minh thinks they affect the image but if I can find a good 48mm ND I can add it further up from the SHG (since I need so much focus extension). I'm thinking the reduced light cone angle might be better (less distortion/wavefront error at filter) but that's just speculation and I could be wrong. There is a heck of a lot of energy concentrated on that slit without a ERF, on my 80mm you can not hold you hand at focus point for more than a few seconds. Clearly a high quality front element ERF (expensive) would work good but my budget is shot for the year.

Timelapse of flares and lifts should be possible by scripting the capture. I have seen a SharpCap script a Chinese user submitted on SolarChat that does that (would need Pro SharpCap license) and I plan on trying it out. Because SharpCap scripting can send ASCOM commands to the mount, ie move it, I could also adjust the mount slew speed as I mentioned before, programmatically setting a non standard rate which may help on my exposure time issue. Automation would certainly speed up the capture cadence. I think 1 scan per minute is easily possible with my settings right now and maybe even faster than that if optimized.

It makes sense that the jagged lines will get more prominent with bad seeing.
I suppose stacking the results of multiple scans would even them out a bit, though I’m not sure whether the Sun might change too much between scans and cause stacking artifacts.

I recently learned that you can create magnetograms with spectroheliographs, with results similar to HMI Magnetograms.
https://solex.astrosurf.com/sol-ex-presentation-en.html
That blew my mind! I had kind of assumed you'd need a fancy space telescope for that.

Edited July 5Jul 5 by Echound

Absolutely amazing to see surface activity at this level. Absorption lines too. Perhaps even @Rudolph might find it useful if not interesting as far as magnetic interactions go.

1 hour ago, hamateur 1953 said:

Absolutely amazing to see surface activity at this level. Absorption lines too. Perhaps even @Rudolph might find it useful if not interesting as far as magnetic interactions go.

It'd be awesome to have near-live HMI-M but I recently bought a house, thousands of dollars for a telescope to detect sigmoids is currently out of reach 😂 The Doppler image captured by Christian Buil is breathtaking though.

On 7/2/2025 at 10:12 PM, astroHoward said:

Back in May @Samrau started a thread on imaging the sun and @Echound mentioned spectroheliographs (SHG for short). It just so happens I had spent months researching building one (the Sol’Ex Project). Today I am starting a new thread on spectroheliographs. Their abilities and operation are so much different than regular solar filter or etalon based solar scopes it requires a separate discussion. One important fact to point out is spectroheliographs are only for imaging the sun, they can not be used visually.

I got started months ago but instead of building one I decided to order a commercially made SHG. This has only become possible recently. Building from scratch isn’t a bad option but it wouldn’t be much faster since you have to order all the optical components at the very least. I received my order this week, ahead of the original estimated delivery date. After spending a few days imaging with the MLAstro SHG 700 model I had to share the first results. Its that good. Maybe I’ll inspire a few to try this type of solar imaging out.

Now anyone that has any experience in astronomy or astrophotography knows the moment they get something new (camera, filter, t-shirt, anything) the weather gets cloudy. Yes that has happened to me of course. But there has been some usable sky to get started experimenting and learning. So what I'm sharing are my first efforts with less than ideal conditions. Even with that the results are exciting.

Two pictures below, HAlpha wavelength, both from the same scan taken today. These are resized and compressed for posting here. The captions link to Telescopius where you can examine larger versions. Frankly its amazing how well they turned out as I was shooting through thin clouds. These are essentially unprocessed, just what the scan processing software did. Much more could be done with stacking multiple images and additional processing.

Colorized HAlpha (automatically done by processing software):

A negative version of same image is also generated by the processing software. I’m particularly fond of negative versions as they seem to have a more 3D effect to them:

I’ve started with the Halpha wavelength to learn and work out my settings . Yet that is just one of many different wavelengths you can image with a spectroheliograph. You just dial in a different wavelength, refocus and scan. While spectroheliographs are not new, they were invented over a century ago, the recent diy projects and software have spawned a new type of amateur solar imaging. Affordable solar imaging compared to traditional dedicated solar telescopes. Now that a commercial SHG model is available this is going to get popular.

In the next post I will go over the hardware and related requirements for spectroheliograph imaging. Then I will post on the software required to scan the sun and create an image.

The website linked for MLAstro has an AI-generated logo, makes me skeptical.

MLAstro hosted a Webex meeting yesterday where Cedric Champeau covered the basic features of JSol'ex. This was a good presentation and I learned a lot. Fortunately it was recorded and is available on Youtube (below). For anyone interested in spectroheliographs and how some features work its worth a watch. The first 10 minutes is MinH discussing the SHG700 product but after that its all information on using JSol'Ex and some interesting information on how the sun scans are interpreted and processed.

In particular its worth watching around the 30 minute point on to see how various images (continuum, doppler, colorized) are generated by pixel shifting. This also includes a discussion of spectral coloring and detecting Ellerman bomb.

As a side note for those that do not know: you can save videos to you computer using YT-DLP. I do this all the time because it bypasses any ads and gives me a local copy. One never knows when a video may disappear (poster deletes it, DCMA takedown, turns private, whatever).

As a tip for any Debian type Linux users: if you want to run JSol'Ex just download the Linux zip file. The deb package did not work for me (dependency issues) and Cedric confirmed that and point me in the right direction (just use the zip version). It turns out Cedric uses the same OS/desktop I do, Linux Mint, and you see that in his presentation.