Please Help - Jupiter Ghosting

[Christopher Davenport]

Hello,

Gave Jupiter a bash last night at 3000mm FL. The seeing was not ideal, which made focusing near impossible.

Anyway I managed a 300s run and captured 48K frames in colour, used 45% of frames.

Processed in DS!S and wavelets Registax and Curves in Image analyser.

Try as I may, I cannot get rid of the ghosting on the image of Jupiter.

Notice the ghosting in the image on the Northern Western edge.

Looks almost like atmosphere, but I know better

All my images of Jupiter do this? Saturn, the moon and galaxies are fine.

What am I missing?

[leelee970]

I can't quite see the ghosting - poor PC monitor, but have you tried RGB align in Registax. The only other thing is maybe overdoing it on wavelets.

[todd8137]

That's a lot of frames I never do that many 2000 max for me and that produces large clear images I do not see any ghosting

Pat

[Barv]

Hi Christopher.

Am I understanding you correctly - 48000 frames? That is an awful lot. As I am sure you know, the more frames you get the lower the noise levels will be. What camera, and  framerate are you using? If you are using a coloured cam, 300 secs is way to long on Jupiter unless you're de-rotating. I usually use no more than 90 seconds and that is pushing things a bit! You should aim for a final stack of between 5 and 10000 frames. It looks to me that this is a stacking issue but I cannot be sure!

Regards

Harvey

[Mr TamiyaCowboy]

Hello,

Gave Jupiter a bash last night at 3000mm FL. The seeing was not ideal, which made focusing near impossible.

Anyway I managed a 300s run and captured 48K frames in colour, used 45% of frames.

Processed in DS!S and wavelets Registax and Curves in Image analyser.

Try as I may, I cannot get rid of the ghosting on the image of Jupiter.

Jupiter 21-49-15-466_g4_b3_ap14 AS p45 1.25Lan 120Con Wave Align Bal Denoise.png

Notice the ghosting in the image on the Northern Western edge.

Looks almost like atmosphere, but I know better

All my images of Jupiter do this? Saturn, the moon and galaxies are fine.

What am I missing?

Hi Chris 

its the stacking software, some of your frames have the objects slightly moved.

when these frames are added you get a ghost effect ( suffered the same fate , but not now ) 

so the simple fix is download PIPP, use this to make sure all your frames with an object in are centered.

then throw into stacking software, you will find the output image has lost the ghosting and the wavelets 

become more crisp. 

https://sites.google.com/site/astropipp/

at 48k thats a big stack, so have the software choose only the best quality.

[Robp]

I'd second giving PIPP a try.

Ive managed to stack a 300s Jupiter AVI and didn't get the distortion you see.

You will also want to vary the stacksize and see which give the best result. Lots of lower quality frames or not so many higher quality frames. with 48K to play with you should be able to limit it to quite high quality and still get a decent amount of frames   

[Christopher Davenport]

Hmmm, thanks all, much to consider.

Sorry cam is QHY5L colour running at 160FPS for 5 minutes, my research suggests that 5 minutes is safe at 3000mm FL, but I know this is on the limit.

Seeing was ridiculously bad last night, Jupiter was dancing around like a maniac last night. (reason why I pushed up to 160FPS, it can go up to 200FPS )  

Will try following approaches to see if they help.

1.) PIPP - Time reduction,  Quality reduction and alignment.

2.) Reduce the stack size to down to 1% (480), 5% (2400), 10% (4800), 20% (9600)

If this does not work then I will trawl my archive for older data that was captured at slower frame rates.

Thanks to all for suggestions. Once I solve (if I solve), I will drop a note back in here.

[Mr TamiyaCowboy]

i am fairly confident just using PIPP to make the images and object centred will have a huge effect.

have suffered the same problem myself couple few frames shifted ever so slightly 

you do not notice, but the software see's everything and adds in those frames.

thats when you get the blobs from hell or the ghostly shadowing look. 

i would say Between 1-3% stacking and select the best quality frames.

the rest of the data could be centered and made smaller with PIPPS,

then made into a movie of jupiter. 

[Space Cowboy]

At that frame rate I would not stack less than 10,000 frames else you will have serious noise issues. Using a fast frame rate to beat the seeing is ok up to a point but you can end up with a very poor signal to noise ratio.  There appear to be horizontal lines on the image which suggests to me the planet was leaving the sensor during capture? PIPP is very good for weeding out those frames which could effect stacking. AS!2 does have its own cropping/centering built in unlike registax so its not always necessary to use pipp first.

Personally I keep the gain below 50% with the QHY and use 60-80fps.

[Christopher Davenport]

Thanks for the feedback.

Using Pipp I have reduced the frames to the first 10K, then I have got it to centre them and quality estimate down to 2000.

Once I have stacked the problem persists, and the quality of the image has degraded significantly.

This rules out the the timing as 10K of 48K, means I am only using the first 60 seconds

Using Pipp I have done a quality reduction to 2K and centering.

After stacking the ghosting is very slight, but the e quality of the image has degraded significantly.

This rules out pure stack size.

Using Pipp I have done a pure quality reduction 10K and centering. 

This image is actually quiet a good image, but the ghosting is similar to AS!2.

This rules out centering of the image using Pipp.

On creating a video using Pipp, the image of Jupiter, elongates and flattens in a wobble.

This must be due to extremely poor seeing conditions.

Thus, I think in very bad seeing, speeding up the FPS, will introduce clarity, but also ghosting as the planet elongates.

So I will try again tonight as the seeing is supposed to be better. I will reduce the FPS around 80 and see the results.

Thanks again to all.

[Christopher Davenport]

@Stuart, those horizontal lines are frames where the usb bus has saturated (due to speed) and introduced horizontal lines into the some frames.

Because they are a black horizon lines, both Pipp and AS!2 pick up the change in contrast include them in the stack thinking they are features.

Another reason to slow the camera down.

[Mr TamiyaCowboy]

glad a pipp video has shown where the error is coming from. 

shame it has not helped bring that ghosting to a stop fully.

but its a start. 

maybe just a case of working out the best frame rate to image quality/stability

atleast you nailed down the problem with a video build