How to Improve Planets

[JohnSadlerAstro]

Hi,

I'm not sure if help pls questions are meant to go in here, but I thought this would be a good place to catch the imagers with amazing photos that post here regularly. If this should be here, mods please move as needed.  

So I've been having loads of fun this summer messing about with my new refractor, and OpticStar 90/800mm achro. I've used it primarily for planetary imaging and observing with my colour ASI-120. I'm really pleased with the results I've been getting so far in both visual and imaging, but as I'm really new to planetary I'm a bit of a noob.  

I don't want to throw lots of money at this, as I have already spent too much (trademark) on a DSO imaging setup. Is there anything I can do or anything cheap I can buy that will help?

My standard method is:

Use 90S Gold frac with ASI 120 and 2x Barlow (on EQ5)

Connect to PC, run SharpCap.

Crop to smallest res. 

Capture at about 50-100fps, for 90sec.

Open in Registax, stack best 40%, use linked wavelets to get it nice and sharp. 

Try to remove disgusting colour.

My main gripe is the colour--it looks so horrible. However hard I try, Jupiter looks like it was in a bar fight earlier in the evening, with purple-brown banding. These are the photos I'm getting at the minute.

Mars, low elevation over bad seeing area. I was amazed I got the polar cap and a tiny hint of shading, but I want to redo this tomorrow morning.

Jupiter, without and then with some Star Tools colour messing.

Saturn, with no Barlow. I want to redo this one tomorrow.

I have been using 90 sec videos on Jupiter because I've heard that it moves too much for anything longer. Is this true? How long can I go on other planets? Will it make much difference?

Thanks for looking!

John

[souls33k3r]

I would seriously recommend one more gadget and that's an ADC. After this lots and lots of frames and lastly processing.

Hope it works out for you

[michael.h.f.wilkinson]

First of all, those are nice results, especially given the aperture you are using. I would have a bash with the 130mm scopes and appropriate Barlows. Aperture is king in planetary imaging, and my trusty old Celestron C8 will hit a fancy 4" APO for six when imaging. a 130mm scope with central obstruction will be better than a 90mm with residual chromatic aberration. Much of course depends on seeing. The very best way of improving planetary images is moving south. I will be in Italy and southern France this summer, and planets will be 9 degrees or so higher up, plus I will be 900m or so above sea level. Both facts will help.

 

[vlaiv]

I agree that results are really nice. That means that your workflow works good so not much to be done there except for a bit of experimenting (like trying to stack less than 40% or more than 40% - learn to judge how much will give best results depending on seeing on particular night).

I would recommend few things that you can try to push your results to be even better.

First off, at this image scale you can safely go for 4-5 minute captures. If using AS!3 (or !2) it will handle a bit of rotation and there will be no blurring. If you decide to go for even longer captures - look into WinJupos for derotation. It can help with longer sequences (more than 5 minutes).

Now for color - since you are using achromat there are two ways that you can try to deal with residual color. First is fairly simple, and it will not cost you much - Use Wratten #8 filter (yellow planetary filter - these should not cost more than 10-15 monetary units). It will disturb color balance a bit, and since these are not interference filters, but absorption ones - it will have small impact on SNR, but it will remove purple haze around bright targets and help with sharpness a bit due to this.

There is another option that will cost you nothing to try, but it might not give as good results, and it is something to be tried and tested (just a theoretical idea, and not sure if it will work good in practice). Shoot 2-3 movies (here you will probably need WinJupos derotation), but changing focus slightly between shots. Now, idea is since you are working with achromat, to focus on particular color and use only that color (channel) from each recording. Depending on achromat, you can get away with either 2 movies (and extract red and green from first and blue from second, with single "refocusing'), or you need 3 (one for each channel). Toughest thing will be focusing on particular color. Here software that can display a single channel instead of color image would be helpful (not sure if SharpCap or Firecapture can do that).

Apart from things said, here are things that will not cost you anything and can improve your results:

1. Take darks / flats for your camera and apply them in Pipp

2. Shoot in 16 bit mode rather than 8 bit (or keep 8 bit format but optimize gain for it). Optimize for coherence time rather than for histogram (this means exposure in 5-10msec range). Use raw of course (ser file format).

3. Keep planet as close to optical axis as possible - so when you crop don't put planet in top left corner but in center. This mostly applies to newtonians, but I would say it is a good practice for any scope. Optical axis tends to be the sharpest part of the field for any design.

4. Don't debayer your images - use bayer drizzle algorithm in AS!3. When working with Pipp tell it not to debayer and to preserve bayer pattern when cropping / centering. For bayer drizzle to work you need to let planet drift across the sensor during capture - so don't do the best polar align that you can, and turn off PEC if you've done it on your mount so that it does not track perfectly.

5. Make sure you optimize your shooting conditions for planetary - don't shoot over heat absorbing surfaces (avoid shooting over walls, pavements, houses - look for grass). Try different times in day for your shots - atmosphere is sometimes more stable in early evening while there is still some daylight left (just after sunset) - additional light will hurt your SNR a bit, but seeing might be better then. Same goes for early before sunrise.

Other than above, I would say - get aperture, but that will cost some ...

[JohnSadlerAstro]

Thank you for the advice, folks, much appreciated!  

I have a set of planetary colour filters, I tried using it on Jupiter a few nights ago and it seemed to improve the contrast--almost like the blue one. Hopefully it is the right kind.

On point 2; if I reduce exposure down to 10msec, it gets horribly underexposed looking. Is this an issue, or can I adjust this later in processing? If gain on the asi120 is raised above 30%, the image becomes really noisy. ? 

I will be trying my 130P-DS with the camera tonight, the only issue is that it has f650mm length, so I am trying to work out weird adapter configurations to act as an extender for my Barlow. Does anyone know how to calculate the "x" number for a Barlow with n mm of extension past the Barlow optic itself?

John

[vlaiv]

2 minutes ago, JohnSadlerAstro said:

Thank you for the advice, folks, much appreciated!  

I have a set of planetary colour filters, I tried using it on Jupiter a few nights ago and it seemed to improve the contrast--almost like the blue one. Hopefully it is the right kind.

On point 2; if I reduce exposure down to 10msec, it gets horribly underexposed looking. Is this an issue, or can I adjust this later in processing? If gain on the asi120 is raised above 30%, the image becomes really noisy. ? 

I will be trying my 130P-DS with the camera tonight, the only issue is that it has f650mm length, so I am trying to work out weird adapter configurations to act as an extender for my Barlow. Does anyone know how to calculate the "x" number for a Barlow with n mm of extension past the Barlow optic itself?

John

On the matter of exposure - go for gain of 64 for asi120 - it appears that it is of the lowest read noise, so best for planetary. Depending on your seeing on a particular night, you might go for 20 or 30ms exposure, but you need really good seeing for this. In average seeing, your best exposure is in range of 5 to 10ms. It will "under expose" single frames, but you should not look at "proper exposure" for planetary.

For planetary imaging two things are important - like in normal imaging signal strength alone is not proper measure - it is SNR. So even faint signal can be signal with good SNR (provided that it indeed is good SNR) - brightness is just the matter of interpretation (one can look at 0-100 as brightness scale, so signal in range 0-1 will be exceptionally dim, but multiply that signal by 100 and it will be really bright - by multiplying signal you multiply noise as well so SNR remains the same - and if SNR is good then you are all set ). But by using short exposures you are hurting SNR indeed (as fewer photons are captured by pixels in shorter time) - this is where second important thing for planetary imaging comes into play. Seeing is not constant distortion - it changes over short period of time. If you set your exposure length such that for duration of that exposure seeing is distorting image but does not change how image is distorted this is good thing. If your exposure length is higher than this speed of change of distortion - and this is called coherence time, then besides distortion you get another difficulty to deal with and that is blurring - or to be more precise - motion blur. It is a bit like when you try to capture someone running (or passing car) with regular camera - if you set your exposure to short - image will be sharp, if you set it too long, you will get motion blur.

Same thing happens with seeing - there is distortion in wavefront that "scrambles" image - a bit like those funny / distorting mirrors. As long as you capture single distortion and don't let change of distortion make blur - it helps stacking program in two different ways - first is, it can tell more easily good frame from a bad one, and second - it can do a bit of returning of frame in proper shape (by comparing it to other frames and undoing large level distortions - this is why you place alignment points - stacking program tries to bend back each frame so that detail on particular alignment point aligns across the frames). As for each frame having low SNR because short exposure - this is the key for planetary imaging. It is balancing act. You want to collect enough frames that are not "motion blurred", of sufficient SNR that upon stacking give you good resulting image in both SNR and blur level (so that you can properly deblur it with wavelets or whatever).

By shooting longer sequences you increase your chances that enough frames will be good enough to end up in stack. By going short exposure, you are hurting your SNR but because of shorter exposure, you will still get more frames (and again increase odds for total number of frames in stack). Short exposures also increase probability of frame not being blurred too much and ending up in stack. This is what distinguishes different cameras as being good for planetary - ability to do high FPS (you want it to be able to capture 100-200 FPS that you can achieve with 5-10ms exposure) - thus USB 3.0 cameras are better, and also you want your planetary camera to be as low read noise as possible. In short exposures, because signal will be low - read noise becomes more and more important term - and you want that to be minimal.

As for barlow - there is simple formula for magnification, but you need to know focal length of your barlow lens.

Formula is as follows 1 - distance / barlow focal length.

Mind you, barlow focal length should have negative sign (negative lens), so above formula will in fact be 1 + distance / focal length (if you take positive focal length for barlow).

If you don't know exact focal length of your barlow, you can sort of measure it (it will not be precise measurement). Take a ruler and try to find "middle" of the lens (barlow element is usually about 5 to 7mm thick, so subtract 3mm if you measure from field lens surface. Measure distance to where eyepiece shoulder rests - as this is usually taken to be focal point.

Look at diagram for help (roughly measure distance between two arrows):

Now let's suppose that you measured 60mm and barlow is declared as being x2.

This would mean that 2 = 1 + 60 / focal length or focal length = 60mm

So if you place your sensor at 80 mm from barlow lens (20 mm further than barlow shoulder), you will get m = 1 + 80/60 = x2.333 magnification factor.

But before you measure barlow this way - look up on internet for your particular barlow, there might be focal length of it published somewhere, like here:

https://www.cloudynights.com/topic/538194-how-do-you-measure-the-focal-length-of-a-barlow-lens/

(few barlows and their respective FLs listed).

There is another simple method to measure barlow magnification for different distance (and obtain its focal length). Just examine any of your images and measure planet diameter from them (like Jupiter). Then calculate what is your arcsecond per pixel ratio from this (just look up Jupiter apparent diameter on that particular date, and measure number of pixels across). From that you can get effective focal length if you know your camera pixel size (3.75um for ASI120).

Formula for that is:  arcsec per pixel = 206.3 x pixel size (in um) / focal length (in mm)

And then divide effective focal length with focal length of your scope.

 

[JohnSadlerAstro]

Hi,

Thank you so much for taking the time to give all this information, plenty to read and digest! I will be trying out Saturn and Mars again tonight, so I can try out all the tips. Hopefully the pictures will be really improved!

EDIT: I've measured out and constructed a 3.333333x Barlow Which will provide very close to the optimal F-ratio with my camera on the PDS. I'm just hoping that the collimation I did is good enough. ?

John

[des anderson]

Thanks a bundle, you`ve explained a lot about planetary exposures and how to get the best from a barlow. I have been under the impression that if a barlow had x2 on it, it magnify by 2 the same as an e.p. Des

[vlaiv]

3 minutes ago, des anderson said:

Thanks a bundle, you`ve explained a lot about planetary exposures and how to get the best from a barlow. I have been under the impression that if a barlow had x2 on it, it magnify by 2 the same as an e.p. Des

Just make sure that you use a Barlow lens, rather than telecentric amplifier like TV powermate.

Telecentric amplifiers are different design (four lenses) and they tend to keep (almost, depends on particular model) their magnification over range of distances.

[JamesF]

4 hours ago, vlaiv said:

There is another simple method to measure barlow magnification for different distance (and obtain its focal length). Just examine any of your images and measure planet diameter from them (like Jupiter).

I believe Firecapture estimates focal length for you if you write the data file as well as the recording.  Obviously you have to tell it the target, but I assume it must have some ephemeris data that allows it to calculate the apparent angular diameter of the target and knowing the pixel size of the camera can calculate the image scale and from that the effective focal length.

James

[JamesF]

Another "simple" thing you can do that might help...

Focus can often be a pain with planets because once you have a fair focal length it's quite hard to touch the focuser and not have the image bounce around all over the place.  The ultimate solution for this is a motor-driven focuser, but that costs money

On OTAs with small focuser knobs an alternative is to use a clothes peg clipped onto the knob.  Then you can move the focuser by gently pushing the peg one way or another with one finger.  With your OTA I don't think that's possible, but you may still be able to fix some sort of rod to the focuser knob, perhaps with a cable tie or something similar.  I believe some people used a rod glued to a bottle cap or jar lid that would fit snugly over the knob.

Before I had a motor-driven focuser I found this enabled me to watch the laptop screen and "fine tune" the focus far more easily.

James

[JohnSadlerAstro]

No planets last night--the seeing was very bad, even Mars was twinkling a tiny bit. I found, though, that the planets are sharper through the refractor than through the 130pds. This is probably because I can never get collimation quite right with it.

I did a reprocess of the Mars image, though. Imo it has more natural colour this way.

John

[Stub Mandrel]

I get my best results with a long series of short runs (5000 frames at 80fps - demands use of a crop to a small frame) then derotating in Winjupos.

Caveat - for good results from Winjupos process every run exactly the same way including the sharpening. Try and keep the runs with minimal gaps between them.

The end result usually allows a further round of sharpening, but sometimes winjupos just 'destroys' your data.

[MarsG76]

To add my 2 cents worth.. I wouldn't be stacking more than 20% of the frames in registax. This ensures that the best quality frames are used and the softer ones are not getting through.

Capturing around 3000 frames per capture is a good number, but thst will depend on the planet your imaging, 40 seconds max for Jupiter, around a minute for Saturn and about 2 minutes for Mars is about the max times you want to capture before rotation stars to blur the details.

Also look into WinJupos de rotation.

 

[Tommohawk]

I agree with everything above! I also agree your images are looking good for a modest aperture. BTW I usually get best results with just 15%.

Couple of other things - I also get consistently better results with a TV powermate than a barlow, even my 4 element "apo" barlow. More £££ though.

Also - I don't think anyone mentioned this - with a colour camera and targets at low elevation, the RG and B channels will suffer lateral displacement. If you don't have an ADC, Registax has a feature to help align the channels. It can make quite a difference. 

One last thing re focussing - Firecapture has an image stabiliser so when you twiddle the focusser the image remains dead still - it brilliant. I don't think SharpCap has that - might be wrong though!

Good luck!