Full disk imaging with a Quark chromo

[StuartT]

I'm about to purchase a Quark chromo for use with a Starfield 102 f/7 scope (FL=700mm).

I've read that full disk imaging is only possible with FL < 450mm, so presumably if I use a 0.5x reducer I'll achieve this. Can anyone confirm? Also, which eyepiece would I need for this? Presumably something longer like a 40mm 

(sorry, I'm rubbish with optical calculations!)

[vlaiv]

No it won't be possible.

Problem is with integrated x4.2 telecentric amplifier and blocking filter diameter.

All of this happens before there is chance for reduction - so you'll be reducing already clipped image.

Sun is about half a degree in diameter - with 714 x 4.2 = 3000mm of focal length.

That makes radius (quarter of a degree) of sun image (little trigonometry) about 13mm or diameter about 26mm

Regular quark has 12 mm blocking filter. That is about half of what we calculated. For this reason you need less than 450mm for full solar disk observing - and even less for full solar disk imaging (observing tolerates small vignetting as eye can't distinguish it easily).

If you want to image rather than observe - then look at quark combo. That one does not have integrated telecentric lens - you need to add your own to get to F/25 - F/30 (either add telecentric lens or use aperture mask or combination). It has 25mm blocking filter with 21mm clear aperture - so almost double of regular quark.

Only issue is that you need your telecentric lens - something like this:

https://www.firstlightoptics.com/barlows/explore-scientific-2x-3x-5x-barlow-focal-extender-125.html

For example - x2 model will give you F/14 scope with 1400mm of focal length. That will be suitable for full disk imaging and you can get to say F/28 by using 50mm aperture mask.

With 1400mm of focal length, solar disk will be about 12mm in diameter - so you need sensor that has height of 13mm or more. ASI1600 is good example. Or in this case - you can use 0.5 reducer to shrink Sun image. But be careful - you'll loose some of sharpness with reducer.

Another thing to consider - at F/28, ideal pixel size for Ha wavelength is 9.2um. This means that you'll need to bin your pixels, so it is best to fiddle around with parameters (aperture mask size, telecentric amplification factor, pixel size and binning and sensor size) in order to get best match for full solar disk.

 

[StuartT]

Thanks @vlaiv this is really helpful!

Edited February 21, 2022 by StuartT

[StuartT]

So a couple of additional questions, if I choose to go with the Quark Combo

1) I currently have the Televue 2x and 3x Barlows, but it sounds like these would not be suitable?

2) I was planning on using the Apollo M mini camera, but this has a sensor height of 6.6mm. Would that only allow me a partial disk view? (I do have a ASI2600MC Pro which is a APS-C size, but it only has a low frame rate, so I am guessing this is not ideal for solar)

What I would like to be able to do (ideally) is to have the choice of imaging both the full disk and close up views

Edited February 21, 2022 by StuartT

[vlaiv]

28 minutes ago, StuartT said:

1) I currently have the Televue 2x and 3x Barlows, but it sounds like these would not be suitable?

Barlow and telecentric lens do almost the same thing - they amplify the image at focal plane (in effect extend the focal length of telescope) and they both increase F/ratio of system.

There is subtle difference in how they achieve this. We say that that solar filter best operates at F/25-F/30, but what we really mean is - solar filter best operates if light rays are almost parallel to each other and perpendicular to filter.

Look at this diagram:

Barlow diverges rays of light. More away from center they are - more angled they end up being. Telecentric lens "separates" them more - but keeps them more "parallel" in the end.

Both will operate the same in the center of the field - but as you start moving away from center - barlow will start causing issues for filter - it will stop performing to the specs and contrast will be impacted - like when using filter in say F/10 beam instead of F/25-F/30.

You can try with barlow to see what sort of result you get, but ultimately, you want telecentric to get the best results.

35 minutes ago, StuartT said:

2) I was planning on using the Apollo M mini camera, but this has a sensor height of 6.6mm. Would that only allow me a partial disk view? (I do have a ASI2600MC Pro which is a APS-C size, but it only has a low frame rate, so I am guessing this is not ideal for solar)

Good thing about quark combo is that you can choose how you achieve F/ratio of your system. You can increase it in two ways:

- use of tele extender

- use of aperture mask.

Difference between the two is that first one increases focal length of the system, while second one does not. Larger focal length requires larger sensor for same FOV (same as solar disk in this case - or just a bit larger than half a degree).

You can image full solar disk with your camera if you don't use tele extender, but instead use aperture mask.

In order to achieve say F/25 with 714mm of focal length - will need 714/25 = 28.56mm aperture mask - let's say 28mm one.

What is the drawback of this method? Well, you won't get very detailed full disk image. Your camera is 4.5um pixel size and resolution 1944*1472, but we have seen that you need pixel size of around 9um so you'll bin your camera 2x2 to avoid oversampling and get sharp image.

This means that your solar image will have only about 700px across. You can't expect sharper more detailed image with 28mm of aperture. In order to get more detailed solar image - you'll need larger sensor.

If you are happy with this - then there is your solution - you can have several different zoom modes for your scope:

- no telecentric lens + 28mm aperture mask for full disk solar imaging

- x2 telecentric lens + ~50mm aperture mask for medium zoom imaging

- x3 telecentic lens and ~80mm aperture mask for close up imaging

[StuartT]

thanks again. I am not sure I understand how stopping down the aperture would change the size of the image. I though this just limited the amount of light coming through? For example, when I change the aperture on a camera lens, the image size doesn't change, just the brightness

Edited February 21, 2022 by StuartT

[vlaiv]

Just now, StuartT said:

thanks again. I am not sure I understand how stopping down the aperture would change the size of the image. I though this just limited the amount of light coming through?

It won't change the size of image. It will make rays of light more parallel - that is requirement by solar filter.

We add telecentric lens in the first place in order to satisfy F/25-F/30 requirement by solar filter (it says that quark combo can work from F/15 - but I would advise to experiment and see how much contrast one looses with F/15 in comparison to F/25-F/30, after all - regular quark has x4.2 and works best with F/6-F/7 scopes - precisely F/25-F/30).

That as a consequence makes solar image larger in the focal plane (increases focal length of the system and gives "more magnification").

If we don't use telecentric lens - that keeps image small as originally is - but creates problem of having system at native F/7 - which will cause issues for filter. We use aperture mask to increase F/ratio and make light beams work properly with filter.

Make sense?

[StuartT]

err... not really. There are a few things I don't understand there!

So I think basically you are saying I can't have my cake and eat it. So I should probably just image partial disk (with a regular Quark) in high quality, rather than try and do everything 

[vlaiv]

5 minutes ago, StuartT said:

err... not really. There are a few things I don't understand there!

So I think basically you are saying I can't have my cake and eat it. So I should probably just image partial disk (with a regular Quark) in high quality, rather than try and do everything 

Quite the opposite.

With regular quark - you get two constraints that you don't have with quark combo:

1. smaller blocking filter

2. integrated x4.2 telecentric lens that you can't change.

Quark combo does not have those restrictions so it's better option in general as far as imaging is concerned.

Now onto the rest, so let's do it one by one:

Quark needs to operate at certain F/ratio. It really does not care how that F/ratio is achieved (to a point - it prefers telecentric over barlow as discussed).

F/ratio is ratio of aperture and focal length - you change it by altering one of the two.

You alter focal length by use of telecentric extender. You alter aperture by use of aperture mask. Any combination of the two that produces required F/ratio is good.

For given F/ratio there is optimum pixel size - around 8.2um in the case of F/25 and 9.84um in case of F/30. If you want exact formula - it goes like this pixel_size = F/ratio * 0.656 / 2 (that is just inverse of f/ratio = pixel_size * 2 / wavelength - which is formula for critical sampling with wavelength set to 656nm or Ha wavelength).

This shows that regardless of the setup you are using - detail on the sun will be limited by size of sensor. Your sensor has 6.6mm of height. There is only so much pixels you can squeeze into 6.6mm if each pixel is ~9um. You need to leave some room for proms - but that is it.

Want to fit whole solar disk onto sensor with higher number of pixels - get larger sensor.

This holds for any aperture size and tele extender combination used - because all result in same F/ratio and same ideal pixel size and if you have one ideal pixel size - well, there is so many pixels of that size (either native size or binned) you can put on sensor.

Back to last post - I just showed you that you can have both full disk image and zoomed in image with combination of equipment that you already have if you get quark combo. Only limit is that it will have certain resolution. This is not the limit of neither quark nor scope nor telecentric lens nor aperture mask - it is limit of sensor and ideal pixel size.

Want to do the same but have more resolution? Get larger sensor.

I think it is sensible route to start of with sensor you already have, get some good images - both close ups and full disk (neither is easy and both require practice in capture and processing) and when you feel confident and ready (and funds allow) - look into getting bigger sensor.

[StuartT]

Thanks so much for taking the time to explain all this @vlaiv

I do actually have a larger sensor (the 2600MC Pro has APS-C sensor), but the frame rate is not so good as the Player One camera. Also I read that binning is best done in post processing, not at acquisition.

[vlaiv]

1 hour ago, StuartT said:

Thanks so much for taking the time to explain all this @vlaiv

I do actually have a larger sensor (the 2600MC Pro has APS-C sensor), but the frame rate is not so good as the Player One camera. Also I read that binning is best done in post processing, not at acquisition.

There are some differences with camera binning that are be beneficial for planetary type imaging - like faster download rates and hence high FPS.

I don't have hands on experience with Ha solar so I have no idea of how fast the features are changing and what is imaging time limit, but from what I've gathered sun is rather dynamic in Ha so there is time limit on individual recordings and higher FPS is definitively going to help (for lunar for example - one can use slow fps and get enough frames over say half an hour, but for solar I think feature motion blur would occur on these time scales).

Play around with binning settings on your Apollo to test things.

ASI2600MC can be used but you have to be careful of how you use it. It is color sensor and only red pixels will record Ha light properly.

This means that every other pixel in X and Y will be effective. Problem is - you'll still need to download full image, so you'll be wasting 3/4 of your USB bandwidth / data transfer speed on data you'll throw away.

Then there is problem of pixel size matching. Single pixel is 3.76um but again - you are not using every pixel but every other pixel, so your actual pixel size is 7.52um. That is neither here nor there with respect to optimum sampling. We have seen that optimum pixel size is around 9um. 7.52um is smaller and you'll be over sampling with that - and that is bad because of SNR. I have no idea how strong Ha signal is with quark combo in a planetary type exposure and will SNR of single sub be issue (for planetary it is, but for lunar it most of the time isn't as moon is very bright - unfiltered at least). Binning that in software later will fix SNR thing - but again, it will be under sampling at 15um.

In any case - ASI2600MC is not the best camera for this - no color camera is. Mono is the best solution for Solar Ha. That does not mean that you can't try with this camera - just keep in mind limitations of it.

 

[StuartT]

1 hour ago, vlaiv said:

There are some differences with camera binning that are be beneficial for planetary type imaging - like faster download rates and hence high FPS.

I don't have hands on experience with Ha solar so I have no idea of how fast the features are changing and what is imaging time limit, but from what I've gathered sun is rather dynamic in Ha so there is time limit on individual recordings and higher FPS is definitively going to help (for lunar for example - one can use slow fps and get enough frames over say half an hour, but for solar I think feature motion blur would occur on these time scales).

Play around with binning settings on your Apollo to test things.

ASI2600MC can be used but you have to be careful of how you use it. It is color sensor and only red pixels will record Ha light properly.

This means that every other pixel in X and Y will be effective. Problem is - you'll still need to download full image, so you'll be wasting 3/4 of your USB bandwidth / data transfer speed on data you'll throw away.

Then there is problem of pixel size matching. Single pixel is 3.76um but again - you are not using every pixel but every other pixel, so your actual pixel size is 7.52um. That is neither here nor there with respect to optimum sampling. We have seen that optimum pixel size is around 9um. 7.52um is smaller and you'll be over sampling with that - and that is bad because of SNR. I have no idea how strong Ha signal is with quark combo in a planetary type exposure and will SNR of single sub be issue (for planetary it is, but for lunar it most of the time isn't as moon is very bright - unfiltered at least). Binning that in software later will fix SNR thing - but again, it will be under sampling at 15um.

In any case - ASI2600MC is not the best camera for this - no color camera is. Mono is the best solution for Solar Ha. That does not mean that you can't try with this camera - just keep in mind limitations of it.

 

Right, I take the points about the need for higher frame rates. I can get away with imaging the moon with my OSC, but that is much less dynamic a target, as you say. Also, I understand that mono is better for the sun in any case. Thanks. I shall look into mono cameras with larger pixels and high frame rates (though I suspect options are very limited with these three requirements!)

The one thing I still don't really understand (sorry to be dumb) is why there is a constraint on f-ratio for the Quark. How does the Quark 'know' what f-ratio rays it's receiving? Why does this matter? 

I assume then that the telecentric barlow (powermate or whatever) needs to be 'before' the Quark combo in the optical train?

Edited February 22, 2022 by StuartT

[vlaiv]

27 minutes ago, StuartT said:

I assume then that the telecentric barlow (powermate or whatever) needs to be 'before' the Quark combo in the optical train?

yes

28 minutes ago, StuartT said:

The one thing I still don't really understand (sorry to be dumb) is why there is a constraint on f-ratio for the Quark. How does the Quark 'know' what f-ratio rays it's receiving? Why does this matter? 

This is needed because of the way Ha filter in quark (and other solar filters) work. It is interference type filter often called Fabry-Perot etalon (although quark uses some sort of mica crystal that operates pretty much the same). In any case - that is best described / explained as two parallel plates that must be at exact distance (filter tuning is getting these plates at exact distance).

It is needed to be so with respect to wavelength of light - it must "bounce" back and forth (quantum phenomena) at exact rate to cancel all other frequencies out.

This requires light to hit that filter at right angle to it. If angle is slightly different - then path of light thru the filter gets longer:

Distance at an angle is longer than direct distance - red arrow is longer than black.

If light is coming at an angle - it throws the filter off band (or rather presents itself to filter as different wavelength). This happens also with regular filters - that is why you have special filters optimized for fast optics. You can't use regular Ha (for night time imaging) filter on say F/2 or F/3 system - it won't work as good.

Regular Ha filter for night time imaging has something like 3-7nm pass band. Solar filter has 0.5-0.7 angstrom pass band or 0.05-0.07nm - that is x100 narrower pass band. For this reason solar ha filter needs much more parallel rays as it is much easier to throw it off band if light is at an angle.

High F/ratio creates almost parallel rays. In ideal setup - light going thru solar Ha filter would be perfectly collimated. This is how front aperture Ha filters work (light coming from far away is effectively parallel rays) or sub aperture filter in telescope - like with Lunt 50 scope:

(screen shot describing how etalon works - there are two lenses - one that makes beams parallel, other that refocuses light). In reality - rays are not at 90 degrees with this setup either - only principal ray is - because sun is not point source - it has some width - so if you point the scope at the center light from edge will come at 0.25 degree angle to optical axis - but those angles are small.

[StuartT]

thanks @vlaiv for this thorough explanation. I really appreciate your help.

So I think I will probably go for the Quark combo then, for greater flexibility.