Juipter, it was worth a giggle!

[Rustang]

No award winning image here but I had fun!

I brought a 100mm f10 refractor earlier this year which I used for Solar white light imaging. Since I'm now doing Ha Solar with a different scope, Ive been contemplating selling the scope but before I do I thought I would have a bash at Jupiter. 100mm obviously is on the small side but its not a bad image I guess for my first go. Captured last night through high thin cloud and the bright moon not to far away. It was almost impossible to see if I was in focus, I'm used to at least seeing if what I'm capturing is at least pretty much there but planetary is a different ball game! so not sure if I'm 100% focused. Captured using a 224mc, and celestron x2 barlow, UV/IR cut filter through an f10 100mm refractor

Software: Firecapture, Autostakkert, Registax and Photoshop. 35 percent of 2000 frames

 

Any thoughts on how I could get a better result or am I flogging a dead horse with this set up? There's obviously not alot of detail, would it be possible to bring out anymore with 100mm? Am I under or over sampling to much?

Edited November 13, 2022 by Rustang

[Ags]

That's really very good, better than I ever got using a 102 mm Maksutov. My suggestions: put an alignment point on the moon (assuming you can, I use AutoStakkert for stacking), shoot a 3 minute sequence so you get more than 2000 frames, and use auto white balance in Registax.

[Rustang]

4 minutes ago, Ags said:

That's really very good, better than I ever got using a 102 mm Maksutov. My suggestions: put an alignment point on the moon (assuming you can, I use AutoStakkert for stacking), shoot a 3 minute sequence so you get more than 2000 frames, and use auto white balance in Registax.

Thanks What do you mean by alignment point on the moon? I forgot I could capture more frames, I believe as long as I can keep it under two minuets there shouldn't be any rotation?

[vlaiv]

With scope of this size - you can comfortably go to 4-5 minutes.

You might be slightly over sampled here - but you can change that if you have a way to change barlow to camera distance. Shortening that distance will reduce barlow magnification factor - to say x1.5-1.8 instead of x2.

Try to use higher gain and shorter exposures. Set exposures to 5ms

Maybe add yellow #8 filter next to UV/IR cut filter, or something like Baader contrast booster.

It will help with color, but it will also remove out of focus light that is lowering contrast and possibly causing issues with stacking.

As far as alignment point on the moon goes - that is good advice and to be clear - in AS!3 - click once on the Jupiter's moon to add alignment point to it. That will make it round and nice instead of smeared all over.

Also - try experimenting with different alignment point sizes, and with number of stacked subs.

Now that I look at list of our software - I don't see autostakkert in it? Did you stack with registax? Switch to autostakkert for stacking - you'll get better results. Do wavelet sharpening in registax but stacking itself in AS!3

[Ags]

In AutoStakkert you place alignment points on regions of detail to guide the stacking. AutoStakkert can compensate for some rotation with the alignment points (and you have a smaller scope with less resolution) so three minutes is fine in my limited experience.  

[Rustang]

12 minutes ago, vlaiv said:

With scope of this size - you can comfortably go to 4-5 minutes.

You might be slightly over sampled here - but you can change that if you have a way to change barlow to camera distance. Shortening that distance will reduce barlow magnification factor - to say x1.5-1.8 instead of x2.

Try to use higher gain and shorter exposures. Set exposures to 5ms

Maybe add yellow #8 filter next to UV/IR cut filter, or something like Baader contrast booster.

It will help with color, but it will also remove out of focus light that is lowering contrast and possibly causing issues with stacking.

As far as alignment point on the moon goes - that is good advice and to be clear - in AS!3 - click once on the Jupiter's moon to add alignment point to it. That will make it round and nice instead of smeared all over.

Also - try experimenting with different alignment point sizes, and with number of stacked subs.

Now that I look at list of our software - I don't see autostakkert in it? Did you stack with registax? Switch to autostakkert for stacking - you'll get better results. Do wavelet sharpening in registax but stacking itself in AS!3

Thanks Vlaiv, sorry yes I used Autostakkert to stack, I will add it to the list! I dont think i can change the barlow to sensor distance so would a x1.5 or x1.8 barlow be better for sampling then?

[Rustang]

21 minutes ago, Ags said:

In AutoStakkert you place alignment points on regions of detail to guide the stacking. AutoStakkert can compensate for some rotation with the alignment points (and you have a smaller scope with less resolution) so three minutes is fine in my limited experience.  

Thanks for the tip, I will give that ago, so a 3 min video would be ok then and not show any rotation with my setup? when you mentioned 'auto white balance' in Registax, did you mean auto RGB balance?

Edited November 13, 2022 by Rustang

[Ags]

Yes, I may have misremembered the name of the feature.

[Rustang]

4 minutes ago, Ags said:

Yes, I may have misremembered the name of the feature.

No problem, that's what Ive been using but the colour is still off, maybe its something in fire capture I need to look at

[vlaiv]

20 minutes ago, Rustang said:

I dont think i can change the barlow to sensor distance so would a x1.5 or x1.8 barlow be better for sampling then?

I would not invest in a new barlow in that case - just keep using the one you have as is.

[Rustang]

4 minutes ago, vlaiv said:

I would not invest in a new barlow in that case - just keep using the one you have as is.

Ok cool. Iin regards to gain and shutter, I had a look at the log at the gain was at 350, shutter 5.066ms so about right, it was a pretty noisy video, well looked noisy so would going any higher on the gain add to more noise losing the little detail that I managed to capture?

[vlaiv]

52 minutes ago, Rustang said:

o would going any higher on the gain add to more noise losing the little detail that I managed to capture?

As long as you don't saturate - raising gain won't do anything to the SNR - in fact, it can be beneficial.

Raising gain has benefit of lowering the read noise (almost always, but there are few fringe cases where this is not true). Other than that - it has no effect on SNR (unless you saturate of course).

What was our FPS?

That is also important metric. Use ROI and USB3.0 connection. Make sure you have usb turbo thingy turned on in sharp cap. With ~5ms - you should be able to get 200fps.

In 4-5 minute capture that will amount to 5 x 60 x 200 = ~60000 frames. Stacking only fraction of that (like 5-6% top frames) will give you stack of ~3600 - which will give you x60 SNR improvement.

You say you stacked 35% of 2000 subs, right?

What was your FPS and how long did you record for? With proper settings, in even one minute you should get ~10000 frames. 2000 is too low count.

[Rustang]

11 minutes ago, vlaiv said:

As long as you don't saturate - raising gain won't do anything to the SNR - in fact, it can be beneficial.

Raising gain has benefit of lowering the read noise (almost always, but there are few fringe cases where this is not true). Other than that - it has no effect on SNR (unless you saturate of course).

What was our FPS?

That is also important metric. Use ROI and USB3.0 connection. Make sure you have usb turbo thingy turned on in sharp cap. With ~5ms - you should be able to get 200fps.

In 4-5 minute capture that will amount to 5 x 60 x 200 = ~60000 frames. Stacking only fraction of that (like 5-6% top frames) will give you stack of ~3600 - which will give you x60 SNR improvement.

You say you stacked 35% of 2000 subs, right?

What was your FPS and how long did you record for? With proper settings, in even one minute you should get ~10000 frames. 2000 is too low count.

I think my limit was set to 3000 not 2000, the average FPS was 127. Ive got the camera plugged in at the moment, with gain at 375, shutter 5ms and high speed on its only getting up to around 138 FPS, ROI is 640 x 480. I think Ive been doing it wrong by setting a frame limit not a time limit!

Edited November 13, 2022 by Rustang

[vlaiv]

2 minutes ago, Rustang said:

I think my limit was set to 3000 not 2000, the average FPS was 127. Ive got the camera plugged in at the moment, with gain at 375, shutter 5ms and high speed on its only getting up to around 138 FPS, ROI is 640 x 480

Check to see if you are doping 12bit capture (16 bit) instead of 10bit (8bit format), or maybe switch to 320x240 roi.

Jupiter is small with that scope and you won't need large ROI anyway (unless trying to capture moons as well).

Above says you should be able to even do 800x600 at 240fps - so it might be down to USB and your computer (SSD?) if everything is setup ok and you still can't get high fps.

[Rustang]

6 minutes ago, vlaiv said:

Check to see if you are doping 12bit capture (16 bit) instead of 10bit (8bit format), or maybe switch to 320x240 roi.

Jupiter is small with that scope and you won't need large ROI anyway (unless trying to capture moons as well).

Above says you should be able to even do 800x600 at 240fps - so it might be down to USB and your computer (SSD?) if everything is setup ok and you still can't get high fps.

Its on 16 bit, so its on 640 x 480 so my FPS seem right if averaging at 127, Ive just done a test 3min video and not frame limited like before and got 22970 frames so I will do that next time plus going down to 300 x 300 ROI does boost the FPS to over 200

Edited November 13, 2022 by Rustang

[vlaiv]

12 minutes ago, Rustang said:

I will do that next time plus going down to 300 x 300 ROI does boost the FPS to over 200

If you are doing 5ms exposure - you will be limited to 200fps by exposure. You can do smaller ROI to save on file size - but you can't get higher FPS because 1000ms / 5ms = 200fps.

You can push FPS further only by reducing exposure time - and I'm not sure you'd want to do that.

Whole point of planetary imaging (lucky type) is to find that sweet spot exposure - one that freezes the seeing - and not to go lower than that as SNR suffers.

[Rustang]

8 minutes ago, vlaiv said:

If you are doing 5ms exposure - you will be limited to 200fps by exposure. You can do smaller ROI to save on file size - but you can't get higher FPS because 1000ms / 5ms = 200fps.

You can push FPS further only by reducing exposure time - and I'm not sure you'd want to do that.

Whole point of planetary imaging (lucky type) is to find that sweet spot exposure - one that freezes the seeing - and not to go lower than that as SNR suffers.

Just to confirm what reducing exposure time is, it's choosing a quicker 'shutter' so 4 or 3ms? So shall I stay with having a small ROI, 5ms and a 3min video? 

Also I was devayering while capturing in firecapture but I've read its best to turn debayer off and capture in mono then debayer in Autostakkert later. I have read though that some people seem to struggle with getting the colour back later. 

[Nigella Bryant]

I'm new to this planetary imaging too. I've use ROI and several two minute stacked image's and used WinJupos to de-rotate them. Try taking several and de-rotate. This is my best image so far albeit with a C11. I'm far, far from an expert, lol. 

[Rustang]

22 minutes ago, Nigella Bryant said:

I'm new to this planetary imaging too. I've use ROI and several two minute stacked image's and used WinJupos to de-rotate them. Try taking several and de-rotate. This is my best image so far albeit with a C11. I'm far, far from an expert, lol. 

Lovely, your doing well! Not sure il be able to resolve that much detail with my little 100mm but one day il get a set up to suit. For now I'm happy to improve with what I've got. 

[vlaiv]

33 minutes ago, Rustang said:

So shall I stay with having a small ROI, 5ms and a 3min video? 

You can comfortably go with longer videos. AS!3 will compensate for any rotation that happens in that time frame as any rotation that can be resolved with 4" scope will be smaller than seeing induced distortion - and AS!3 can deal with those with alignment points.

I'd say - shoot 5 minute video and don't worry about derotation. It is useful when you have longer videos (for example 5 minute on C11 or C14 might already be too much as they resolve much more than 4" scope - at least x2.5 more) - or videos that are spaced apart.

You won't need smaller ROI either if you use 8bit capture and 5ms as camera is capable 200+ FPS on 8bit and 800x600, but for sake of storage, unless you want to also capture moons - use 640x480 (4-5 minute video at those speeds can easily reach 20GB or more).

5ms is good exposure time for most nights and I don't think you need shorter than that.

[Rustang]

10 minutes ago, vlaiv said:

You can comfortably go with longer videos. AS!3 will compensate for any rotation that happens in that time frame as any rotation that can be resolved with 4" scope will be smaller than seeing induced distortion - and AS!3 can deal with those with alignment points.

I'd say - shoot 5 minute video and don't worry about derotation. It is useful when you have longer videos (for example 5 minute on C11 or C14 might already be too much as they resolve much more than 4" scope - at least x2.5 more) - or videos that are spaced apart.

You won't need smaller ROI either if you use 8bit capture and 5ms as camera is capable 200+ FPS on 8bit and 800x600, but for sake of storage, unless you want to also capture moons - use 640x480 (4-5 minute video at those speeds can easily reach 20GB or more).

5ms is good exposure time for most nights and I don't think you need shorter than that.

Thanks vlaiv, I appreciate your help again. I'm not guiding while planetary imaging like I do with DSO imaging because I can't be bothered to set it up, I'm just side rail tracking via my HEQ5 Pro. I'm guessing 5ms exposures over a 5min video I don't have to worry about things like star travel because I'm not tracking and as long as jupiter is still in frame I'm good to go without tracking!? I ask because I used to get travel over 2min exposures without tracking but as I hope its different because I'm taking ms exposures and not long exposures. 

Edited November 13, 2022 by Rustang

[vlaiv]

1 minute ago, Rustang said:

Thanks vlaiv, I appreciate your help again. I'm not guiding while planetary imaging like I do with DSO imaging because I can be bothered to set it up, I'm just side rail tracking via my HEQ5 Pro. I'm guessing 5ms exposures over 5min videos I don't have to worry about things like star travel because I'm not tracking and as long as jupiter is still in frame I'm good to go without tracking!? 

No guiding is needed for planetary and in fact - it is going to interfere with normal imaging.

Just simple tracking should be enough to keep the planet in the FOV for duration of 5 minutes.

 

[Rustang]

Just now, vlaiv said:

No guiding is needed for planetary and in fact - it is going to interfere with normal imaging.

Just simple tracking should be enough to keep the planet in the FOV for duration of 5 minutes.

 

Perfect, nice a simple to set up still which is what I want 🙂

[Nigella Bryant]

20 minutes ago, vlaiv said:

You can comfortably go with longer videos. AS!3 will compensate for any rotation that happens in that time frame as any rotation that can be resolved with 4" scope will be smaller than seeing induced distortion - and AS!3 can deal with those with alignment points.

I'd say - shoot 5 minute video and don't worry about derotation. It is useful when you have longer videos (for example 5 minute on C11 or C14 might already be too much as they resolve much more than 4" scope - at least x2.5 more) - or videos that are spaced apart.

You won't need smaller ROI either if you use 8bit capture and 5ms as camera is capable 200+ FPS on 8bit and 800x600, but for sake of storage, unless you want to also capture moons - use 640x480 (4-5 minute video at those speeds can easily reach 20GB or more).

5ms is good exposure time for most nights and I don't think you need shorter than that.

I stand corrected, I'll butt out cus my knowledge is limited. 

[vlaiv]

1 minute ago, Nigella Bryant said:

I stand corrected, I'll butt out cus my knowledge is limited. 

What about?

Derotation?

The way you use it - it is perfectly fine and necessary. If you take several videos and stack each of them and then stack those resulting images - then you need to derotate.

AS!3 will stack each of those videos to average feature position (or rather to feature position in reference frame it creates), and that is usually midway thru the video. Several videos will therefore produce images that are slightly rotated each and need to be derotated.

Joining short videos to single long video and stacking all at once can circumvent this if total recording is short enough for aperture (or rather sampling rate - but I assume one is using critical sampling anyway - all the telescope can resolve).

Feature shift is not hard to calculate. We need rotational period of planet and circumference on equator to calculate tangent speed. From that we can calculate shift in given time and from distance to earth - we can translate that shift distance into angular size.

AS!3 is capable of handling at least 1-2" of distortion, so if this angular size is less than that - you are fine in time frame you've chosen - then you are fine.

By the way - it is better to take one longer video (if you are within above time limit) - then several short ones - or at least to join short ones.

This is because of lucky imaging approach. We never know if spell of good seeing is going to be equally distributed over whole imaging time - or perhaps better seeing will be towards beginning or the end of recording. If you stack short videos - then one of those videos might contain poor seeing - but if you stack whole video, even if you choose same percentage - it will select best subs overall not just in segment.

(top n% of whole population will be better than top n% population of each partition joined)