hello members
this is a first attempt at stacking two completely separate sets of exposures on different nights (of the same object of course)
however has anybody seen these striations before on the finished pic?
any suggestions on how to overcome them?
60da
pat
I've had them before on images with my 60Da where the sensor has been at ridiculously high temperatures - 35 degrees or more. Particularly with dark frames - I can actually see the striations on the darks themselves.
Do you know what your sensor temps were?
I'm guessing, but maybe also if the temps from the subs from the two nights were very different you could ave issues. Did you take a separate set of darks each night?
hmmm thats interesting jon, never would have thought about that
i dont know how to check a sensor temp though
havnt mastered darks yet
the subs would have been very different as they were along time apart
one set of subs was 180s@800 and tuther 240s@800
thanks jon i will remember that
pat
Hi Pat, usually a couple of things cause this but foremost is not matching darks to lights at same temperature.
Also over-stretching and I can see bad (aka hot) pixel trailing but over all it's not a bad result particularly good response to spiral arms Ha.
Try better matching of darks to lights temperature wise and dithering image subs (Dark library avail here, post #4710249). Temps of sensor can vary wildly from start of imaging to finish and darks can also vary the same. Dithering will help the bad pixel trailing and improve S/N.
If you manage your capture through something like Backyard Eos, you can set it up so it records the temperature of the sensor after each frame in the file name. Do the same with the darks then its straight forward when it comes to matching.There will be other software like APT which can also do this I think then of course there is Dark library which does the matching for you though I havnt tried that.
Generally I find with the 60Da that after the first couple of frames, if you leave say 30 to 60 seconds between subs to allow the data to download, the ambient sensor temperature will plateau out very quickly and you get relatively consistent temperatures over an imaging session. BYE though makes it really easy to manage this.
Just make sure you dont have live view running when you start imaging. its a killer. I saw mine leap to about 45 first time I used the camera. Almost wept. BYE defaults it to on when framing and focus which is fine but then you have to wait ages for it to cool down again
Hi Pat, usually a couple of things cause this but foremost is not matching darks to lights at same temperature.
Also over-stretching and I can see bad (aka hot) pixel trailing but over all it's not a bad result particularly good response to spiral arms Ha.
Try better matching of darks to lights temperature wise and dithering image subs (Dark library avail here, post #4710249). Temps of sensor can vary wildly from start of imaging to finish and darks can also vary the same. Dithering will help the bad pixel trailing and improve S/N.
thanks brendan i actually had no darks to add to the images so they are all lights......... i can probably see how important they are now
i think because we have such limited number of nights actually imaging i get too excited and snap away like a mad scientist until i drop ha ha!
it has actually taken some serious discipline to concentrate on a single object for a couple of hours
i suppose it may be better for me to take an internal dark with every light...... but it takes soooooooo long!
pat
If you manage your capture through something like Backyard Eos, you can set it up so it records the temperature of the sensor after each frame in the file name. Do the same with the darks then its straight forward when it comes to matching.There will be other software like APT which can also do this I think then of course there is Dark library which does the matching for you though I havnt tried that.
Generally I find with the 60Da that after the first couple of frames, if you leave say 30 to 60 seconds between subs to allow the data to download, the ambient sensor temperature will plateau out very quickly and you get relatively consistent temperatures over an imaging session. BYE though makes it really easy to manage this.
Just make sure you dont have live view running when you start imaging. its a killer. I saw mine leap to about 45 first time I used the camera. Almost wept. BYE defaults it to on when framing and focus which is fine but then you have to wait ages for it to cool down again
cheers
david both yourself and brendan may have hit the nail on the head here
i had no idea that the sensor could "boil" like that
i will try the delay technique and the "live view" is also very interesting
thanks guys
pat
Try a debanding routine. Startools and Pixinsight have this function.
Also, if using DSS, try a Median Kappa-Sigma, with 50 iterations and at least a 2 x drizzle, or perhaps consider super pixelation. It CAN have a big effect.
I always found APT better for DSLR capture than BackyardEOS. And APT shows the sensor temp after each frame as well - I have seen alarmingly high figures (once topped over 40°!!!)
Dithering will eliminate this effect. The frames are too closely matched.
My basic understanding of the problem (among other problems) is that the addition (average, median as the case may be) of pixel values during stacking increases the value of noise, where pixel x, y position is closely matched or coincident and creates this striated appearance. Calibrated frames will do this too. Light to dark frame temperature mismatch is common with DSLRs.
It is easier in the long run and more effective to space your light frames while imaging by 10 - 15 pixels. There will always be a little dark mismatch and dithering will hide this as well.
Berry and Burnell recommend dithering DSLR frames by about 12 pixels in their handbook of image processing. It works wonders. Inceases SNR (noise reduction), improves flat fielding and as mentioned masks calibration issues.
Try a debanding routine. Startools and Pixinsight have this function.
Also, if using DSS, try a Median Kappa-Sigma, with 50 iterations and at least a 2 x drizzle, or perhaps consider super pixelation. It CAN have a big effect.
I always found APT better for DSLR capture than BackyardEOS. And APT shows the sensor temp after each frame as well - I have seen alarmingly high figures (once topped over 40°!!!)
thanks lewis!
always wondered what all that kappa stuff was for....... and yes it was said earlier by Brendan and dave that very high sensor temps can happen
pat
Dithering will eliminate this effect. The frames are too closely matched.
My basic understanding of the problem (among other problems) is that the addition (average, median as the case may be) of pixel values during stacking increases the value of noise, where pixel x, y position is closely matched or coincident and creates this striated appearance. Calibrated frames will do this too. Light to dark frame temperature mismatch is common with DSLRs.
It is easier in the long run and more effective to space your light frames while imaging by 10 - 15 pixels. There will always be a little dark mismatch and dithering will hide this as well.
Berry and Burnell recommend dithering DSLR frames by about 12 pixels in their handbook of image processing. It works wonders. Inceases SNR (noise reduction), improves flat fielding and as mentioned masks calibration issues.
Virtually essential for DSLR's.
Added comparison image. Same camera.
OMG roland that is amazing! I have only read about dithering on this and other forums and, although I know the reason why,........ how is another kettle of fish
some serious reading methinks on this
how does one "dither?", in between each light frame do you put your mount on a guiding speed and move the image a smidgen?
pat
Pat my setup is unguided with only a hand controller. There are programs that have a dither facility, but this is how I do it.
Oops! the DEC after image 3 should be RA.
ha ha Rowland that's fab but its just double dutch to me!
unguided you say? you are a genius, I have trouble guiding with a bloody lodestar ha ha thanks mate!
pat
oh and by the way guys can my data be salvaged and improved despite the lines?
pat
Pat, the advice you've had is very technical and spot on, AFAIK.
But there's a biggie here: it's never too late to get your dark fames. As far as I can see, no one has addressed the fact that you haven't added dark frames.
Take some (e.g maybe 20) exposures at the same ISO and the same duration as your lights. Leave the lens cap on the camera: you can take them anywhere, anytime. The point is, you are capturing a record of the way the pixels in your sensor heat up and create noise while your lights are exposing. These are your darks.
Your stacking program will then "subtract" this heat-generated noise from your light frames. This makes a *huge* difference. The good news is, you can capture these dark frames anytime.
Try this and let's see yor results. I think your image is amazing given there were no darks; well done!
Also, Pat - see the darkness around the edge of the image, surrounding a brighter central circle? This is "vignetting", a consequence of the optics of your telescope.
You can eliminate this by taking "flats" - short exposures (maybe around 1/50s given your setup) of a uniform light surface - a lightbox, or a white cloth stretched across the scope, or the twilight/dawn sky. These have to be taken before you touch the scope/camera setup - everything has to be *exactly* the same as when you took your lights.
It's too late to get these now for your M83 image. But you can crop the dark edges out of the image pretty well.
Pat, the advice you've had is very technical and spot on, AFAIK.
But there's a biggie here: it's never too late to get your dark fames. As far as I can see, no one has addressed the fact that you haven't added dark frames.
Take some (e.g maybe 20) exposures at the same ISO and the same duration as your lights. Leave the lens cap on the camera: you can take them anywhere, anytime. The point is, you are capturing a record of the way the pixels in your sensor heat up and create noise while your lights are exposing. These are your darks.
Your stacking program will then "subtract" this heat-generated noise from your light frames. This makes a *huge* difference. The good news is, you can capture these dark frames anytime.
Try this and let's see yor results. I think your image is amazing given there were no darks; well done!
thanks again jon
does your camera need to stay at prime focus for darks and flats and also the same focus?
pat
i have just found the two finished stack photos i combined..... i dont know if it would make any difference but i purchased an eos clip in LP filter for the first image, but i did not like it because i could not get any blue in the spiral arms or nebulae in the red portion unlike the second image which has a 2" idas attached instead........ what a huge difference in the images
by the way all the lines seemed to have been manifested in the clip in version only
incidentally i had submitted the first image on the beginners forum a couple of months ago and david (houghy) run it through pixin... and it came out much better
pat
no worries Pat. What's a Lodestar? 
