Vertical lines below stars

[mrjolly]

Hi,

Was out getting some pics of the supernova and M81 last night and towards the end, I tried taking x10 60s shots of the Horsehead just to see what, if anything would come out.

Surprisingly, I got a bit of the HH nebula for even that small amount of data, unfortunately, I've got a strange problem with bright stars, Alnitak being the most obvious, in that the bright ones have a dark vertical line running directly down from them.

Taken with an un-modded EOS1100 - is it just that these stars are too bright and have overloaded the sensor somehow? I've tried stacking in DSS and PI but get the same results from each.

Thanks!

[davefrance]

Try shortening the exposure time and see if you still get the lines.

[Adamar98]

I think a few calibration files (darks, flats and bias) might sort it out. Nice image by the way.

[mrjolly]

Try shortening the exposure time and see if you still get the lines.

I'll give that a go next time I'm out, I'm tending to think it's just overloading the sensor as the lines are completely vertical yet the camera was at about a 45ish degree angle so the chances of it being something optical seem extremely remote!

[mrjolly]

I think a few calibration files (darks, flats and bias) might sort it out. Nice image by the way.

Haven't plucked up the courage (found a good way) to do flats yet, definitely need to sort something out though as even after background extraction I'm getting some wacky viginetting! Luckily, nothing a little cropping hasn't sorted so far!

[Adamar98]

Im no expert but I don't think its anything to do with your exposure times. If anything I would say up your exposure times. I would try a few bias frames and see what that does. When I take flats with the DSLR, I use a white screen on my ipad or laptop. its seems to work a treat.

[Leveye]

Try the same length shots again using a different ISO setting. Each one has a different effect on an image with Canon products i have found. 400-800 seem to work the best but with short exposures 1600 works in a pinch.  My guess it goes away. 

[davefrance]

Im no expert but I don't think its anything to do with your exposure times. If anything I would say up your exposure times. I would try a few bias frames and see what that does. When I take flats with the DSLR, I use a white screen on my ipad or laptop. its seems to work a treat.

Not an expert either and agree that darks, flats and bias could help, but still feel that as the brighter the star the stronger the line, reducing exposure time would seem logical. Willing to be proven wrong.

[lensman57]

Hi,

Was out getting some pics of the supernova and M81 last night and towards the end, I tried taking x10 60s shots of the Horsehead just to see what, if anything would come out.

Surprisingly, I got a bit of the HH nebula for even that small amount of data, unfortunately, I've got a strange problem with bright stars, Alnitak being the most obvious, in that the bright ones have a dark vertical line running directly down from them.

h_small.jpg

Taken with an un-modded EOS1100 - is it just that these stars are too bright and have overloaded the sensor somehow? I've tried stacking in DSS and PI but get the same results from each.

,

Thanks!

Hi,

I am afraid that this is a sensor issue as I had the same problem with a new Astro modified 1100d from Juan of Cheap Astrophotography, I sent the camera back and he changed the sensor and the board at the same time at no cost. The second sample still shows faint lines such as yours under very bright stars, M45, but with extreme stretching so in effect it is fine for what I want it for. It also  seems that it is an issue with 1100d as I have not heard of it with any other model. If you did find a solution for it please drop a post . I have attached  a capture of M45 from the first faulty sensor.

Regards,

A.G

[Leveye]

Thats a new one on me. Hope you can get it all straightened out. This is from a modded camera? Not good.Over the pond here modded older T2i and T3i's are becoming very very popular. The images i'm seeing are amazing.

[ollypenrice]

Darks, flats and bias will have no effect on this. In a camera without an anti blooming gate you'd get bright white bleed lines coming down where now you have the dark lines. So, though I don't know the electronic ins and outs, this will be an effect related to blooming and its control, I expect.

Alnitk is one of the astrophotographic stellar monsters so if you only get it on a handful of stars as bright as this I'd just fix it with the healing brush in Ps. If it's widespread then you have a problem, certainly.  On the other hand your optics have controlled Alnitak incredibly well. It is showing little bloat and you have background data close in to it, which is unusual and impressive.

Olly

[ChrisLX200]

I have seen this before (not on my 70D) and it's described as a sort of negative-blooming effect. Very difficult to process out. Not all processors seem to suffer from it though, even using the same model of camera.

ChrisH

[mrjolly]

Thanks all for the replies and help! Looks like it's just an issue with the 1100d. Next time I'm out, I'll try the same region at a range of different ISOs to see if that has any effect at all (the one above was iso800) and I'll also try with an older unmodded 350D to see if that does something similar.

It's not the end of the world, the fainter stuff I've been interested in up to now has come out okay and if I'm going for something with a bright star nearby I'll just have to be a bit more careful with the framing and cropping!

[ollypenrice]

I have seen this before (not on my 70D) and it's described as a sort of negative-blooming effect. Very difficult to process out. Not all processors seem to suffer from it though, even using the same model of camera.

ChrisH

This was my thinking too. Lines this can sometimes be processed out quickly in Ps by making a copy layer and in Filter-Other-Offset you can shift the bottom layer right or left by more pixels than the width of the line and then erase the line on the top layer. If there aren't too many stars its effective.

Olly

[davefrance]

CMOS sensors should not be subject to blooming or smearing, only CCD.

[lensman57]

Thanks all for the replies and help! Looks like it's just an issue with the 1100d. Next time I'm out, I'll try the same region at a range of different ISOs to see if that has any effect at all (the one above was iso800) and I'll also try with an older unmodded 350D to see if that does something similar.

It's not the end of the world, the fainter stuff I've been interested in up to now has come out okay and if I'm going for something with a bright star nearby I'll just have to be a bit more careful with the framing and cropping!

Hi,

I also have a modded 1000d and an unmodded 1100d and neither of these show this effect, so I guess that it is probably limited to some 1100ds, I wonder if it was a batch of processoes and sensors that suffer from this or if the gain or processor settings have an effect, in anycase there is not a lot that could be done with this .

A.G

[ollypenrice]

CMOS sensors should not be subject to blooming or smearing, only CCD.

Indeed. Sorry, I have a very CCD head!!

Olly

[davefrance]

Indeed. Sorry, I have a very CCD head!!

Olly

Regrettably I can't afford a CCD head, or so the wife informs me. Cheers

[mrjolly]

If you're trying to get me to convert to CCD, it's not going to happen. Never. At all. Ever.

It's tempting though....!

[gajjer]

Hi MrJolly

I too have a 1100D which has been modified and it shows the same signs of streaks from very bright stars.

I would be interested to hear if others have the same problem.

cheers

gaj

[davefrance]

If you do a search on the internet this problem seems to have been around since a least 2005 and mainly relating to 1100D's but others are mentioned. I can find no concrete cause or solution, some have suggested a firmware update may solve it. Worth emailing Canon? they must be aware of it.  

[mrjolly]

I've since found a couple of references to the problem on the internet so it definitely looks fairly widespread with the 1100D:

http://www.startools.org/forum/viewtopic.php?f=10&t=289

http://www.astronomyforum.net/astro-imaging-forum/171651-vertical-banding-canon-1100d-rebel-t3.html

http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/6211233/page/2/view/collapsed/sb/5/o/all/fpart/1

Apparently one solution might be to shoot at 1600iso, another forum mentioned it could be heat related so a few things I can play around with next time I get the chance to go out.

[davefrance]

I've since found a couple of references to the problem on the internet so it definitely looks fairly widespread with the 1100D:

http://www.startools.org/forum/viewtopic.php?f=10&t=289

http://www.astronomyforum.net/astro-imaging-forum/171651-vertical-banding-canon-1100d-rebel-t3.html

http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/6211233/page/2/view/collapsed/sb/5/o/all/fpart/1

Apparently one solution might be to shoot at 1600iso, another forum mentioned it could be heat related so a few things I can play around with next time I get the chance to go out.

Good luck hope you get it sorted.

[mrjolly]

Thanks!

I call it an issue but it's more of an inconvenience! So far I've been using it for fairly faint stuff & it's worked perfectly for that. If nothing seems to help I'll just have to remember not to get anything too bright in frame or rotate the camera so that the lines don't obscure the target & crop out any offending lines.

Or pretend the stars are on sticks or something.