NGC 1365 in colour

[ericwbenson (Eric)]

Originally Posted by Placidus

Thanks, Eric. You've solved a long-standing mystery.

Definitely RBI, and not just from the focus star but from every star in the image. Had a look at the raw subs. It's not just in the green. We dither a lot between subs. There's always a perfect but shifted copy of the previous image. We absolutely don't want to go down the infrared pre-flash path. We've looked very closely both theoretically and experimentally at the effect of that on noise, and it's not acceptable (NASA recommends only using IR pre-flash at chip temperatures below about minus 70). Since we know the between-sub dithering exactly, and know the guy who wrote GoodLook 64 (our image-processing software), we could predict the amount of ghosting and either partially subtract or de-weight those pixels in that image, but for the moment I think I'll just file it under "irritating cosmetic issue to be addressed one day".

Thanks again for that.

Best,
Mike

I agree with the pre-flash decision, my camera has it but I don't use it either. What cooler temperature are you using? I do 30min subs for almost everything at -25C and don't see such ghosting on field stars. It could be the 'CCD lottery' unfortunately. Processing them out would be a chore, remembering the order of the subs to now the previous exposure position, etc, certainly only a nice to have.
EB

[Placidus (Mike and Trish)]

Originally Posted by ericwbenson

I agree with the pre-flash decision, my camera has it but I don't use it either. What cooler temperature are you using? I do 30min subs for almost everything at -25C and don't see such ghosting on field stars. It could be the 'CCD lottery' unfortunately. Processing them out would be a chore, remembering the order of the subs to now the previous exposure position, etc, certainly only a nice to have.
EB

Chip normally at -30C. The colder it is, the less strong the residual image, but the longer it takes to go away.

[gregbradley]

Originally Posted by Placidus

Chip normally at -30C. The colder it is, the less strong the residual image, but the longer it takes to go away.

I don't use it either. Although I did see some affected subs the other night so I turned the camera off and let it warm up and then turn it back on and cool back down and the residual image was gone.

Greg.

[Placidus (Mike and Trish)]

Originally Posted by gregbradley

I don't use it either. Although I did see some affected subs the other night so I turned the camera off and let it warm up and then turn it back on and cool back down and the residual image was gone.

Greg.

Thanks for that, Greg. I'm starting to design a program that will sit between calibration and registration. It will go through each image that we took in a session in chronological order, and have a go at guestimating the residual image at the start of each exposure. It will have just two parameters: the strength of the residual image and the time constant for the exponential rate of decay. One can then use the guestimate to automatically either correct the actual image taken, or provide a set of weights for ignoring particular parts of the frame when stacking. That second part will sit on the back burner as we're busy fixing cattle yards and buying steers at the moment.

[gregbradley]

Originally Posted by Placidus

Thanks for that, Greg. I'm starting to design a program that will sit between calibration and registration. It will go through each image that we took in a session in chronological order, and have a go at guestimating the residual image at the start of each exposure. It will have just two parameters: the strength of the residual image and the time constant for the exponential rate of decay. One can then use the guestimate to automatically either correct the actual image taken, or provide a set of weights for ignoring particular parts of the frame when stacking. That second part will sit on the back burner as we're busy fixing cattle yards and buying steers at the moment.

Well that sounds pretty advanced.
It may just be my lack of testing but I have only really noticed RBI on bright objects.

I have only really seen it a few times. Once when bright jet lights went through an image, the other I was referring to was NGC253 and the image was repositioned in the frame and I could see a ghost image of the earlier 10 minute sub below the actual and copies of bright stars mirrored imaged after the meridian flip.

Greg.

[strongmanmike (Michael)]

Originally Posted by Placidus

I'm starting to design a program that will sit between calibration and registration. It will go through each image that we took in a session in chronological order, and have a go at guestimating the residual image at the start of each exposure. It will have just two parameters: the strength of the residual image and the time constant for the exponential rate of decay. One can then use the guestimate to automatically either correct the actual image taken, or provide a set of weights for ignoring particular parts of the frame when stacking.

https://www.youtube.com/watch?v=37OWL7AzvHo

That second part will sit on the back burner as we're busy fixing cattle yards and buying steers at the moment.

Now that I understand...you area a steering committee

[Placidus (Mike and Trish)]

Now that I understand...you area a steering committee [/QUOTE]



Sixteen black Angus steers arrived today. They're all looking happy, healthy, and quiet. Probably just waiting till we're not paying attention and then they'll eat us. Must photograph something in Taurus soon, to celebrate.

[Atmos (Colin)]

Reminds me of the plot of Black Sheep