Dark frame diagnosis...?

[Aenima]

I am hoping someone can assist with an issue I've been having when taking calibration frames on my 300D - only noticed the problem after having the filter removed and to a lesser extent with longer exposure times.

I first realised something was wrong when I tried to take the darks indoors with the scope's cap on, and saw that a big part of the frame was illuminated by the lightbulb in the room. I removed the camera from the focuser and tried with just the cap on. Still got light leakage, it seemed from either the viewfinder or the rubber side cover where the sockets are.

I have heard of light leakage and covering the viewfinder but with my 300D it hadn't needed it before, so assumed it might have been the filter mod changing the internal parts slightly. I havent really got detailed knowledge of where the light might be reaching the sensor via the openings in the camera casing and haven't heard much on the subject either, so i'm hoping someone might have had the same thing happen or be able to tell from the picture where the light might be coming from.

This is a dark from outside with scope cap on, but I think the blue tint is because the sky was getting light near dawn time. I might not have noticed if it was totally dark out and used the darkframe as normal in DSS, but this is now a constant feature in my darks - and less noticeable but present in all my shots.

Can anyone help with diagnosing the source of this problem? Maybe as its always in the same place (right hand side, lower right mostly) it might correspond to a certain area of the camera?

Any info and all suggestions gratefully appreciated,

Regards

Aenima

[rfdesigner]

In your image, my first reaction was "amplifier glow". That doesn't mean it is amp glow, just that it looks like it. Normally sensors have an amplifier in one corner that gets a little warm, and this increases the leakage and noise in that corner of the sensor.

Dark frames must be at the same temperature as your lights.. taking the camera indoors will likely change the temperature (certainly will during winter), many people don't use darks on DSLRs because of the difficulties of matching temperatures.

Derek

[WaveSoarer]

Aenima,

Have you tried placing black electrical tape over any potential light paths in to the camera? It may help identify a location if nothing else.

I hope you find out what the issue is.

Dave

[Aenima]

Thanks for the replies,

I did wonder about amp glow but the same length exposure done with my bino case over the cam can (*can* but not 100% sure of this conclusion) be free from the trouble, but ive not had much more certainty trying to cover areas either. Mostly i'm putting a bag over the camera and hoping it will work, but really want to get to the root of the issue

Tape is a good idea and worth a try , i'm still hoping someone will recognise the picture or the problem itself.

Many thanks,

Aenima

ps. Hi Dave, hope the imaging is going well.

[WaveSoarer]

Aenima,

I'm enjoying doing occasional imaging - though I haven't posted anything recently (wouldn't stand up to much scrutiny really). I'm using Mrs WaveSoarer's DSLR which has an issue with the memory card port. Niggely problem, sometimes it works sometimes it doesn't, but hopefully we can get it fixed.

I really hope you resolve this but at least the sensor doesn't show any signs of damage or electronics issues.

all the best

Dave

[Aenima]

In your image, my first reaction was "amplifier glow". That doesn't mean it is amp glow, just that it looks like it. Normally sensors have an amplifier in one corner that gets a little warm, and this increases the leakage and noise in that corner of the sensor.

Dark frames must be at the same temperature as your lights.. taking the camera indoors will likely change the temperature (certainly will during winter), many people don't use darks on DSLRs because of the difficulties of matching temperatures.

Derek

I think the main reason that I didn't stay with the idea of amp glow was the colour change when different types of light are near (eg blue outside, yellow indoors).

I know that filter removal will inevitably leave a small gap where it used to be, but knowing nothing about the internal workings I wouldn't like to guess as to where the light is coming from.

Regards

Aenima

[rfdesigner]

I think the main reason that I didn't stay with the idea of amp glow was the colour change when different types of light are near (eg blue outside, yellow indoors).

You don't need to guess. look at the pre-debayed image, if there's light ingress you'll see a checkerboard pattern.. zero pattern means no light ingress.

You can load raw images in IRIS and it shows them in mono form, other packages will do the same

Derek

[Oddsocks]

Hi Aenima,

It might be possible that the light-shielding gasket surrounding the CCD chip has not been set back correctly after the IR filter was removed.

See Gary Honis's excellent instruction web site for self modifiers of the 300D here: http://ghonis2.ho8.com/rebelmodnew.html

Pages 4 and 5 show images of the gasket and in the user comments section some self-modifiers reported light leakage at the edges of the CCD if the gasket had been damaged during disassembly.

This seems the most likely cause from your description of the problem.

You may have to return the camera to your modifier and have the gasket checked.

Regards,

William.

[Aenima]

Hi Aenima,

It might be possible that the light-shielding gasket surrounding the CCD chip has not been set back correctly after the IR filter was removed.

See Gary Honis's excellent instruction web site for self modifiers of the 300D here: http://ghonis2.ho8.c...ebelmodnew.html

Pages 4 and 5 show images of the gasket and in the user comments section some self-modifiers reported light leakage at the edges of the CCD if the gasket had been damaged during disassembly.

This seems the most likely cause from your description of the problem.

You may have to return the camera to your modifier and have the gasket checked.

Regards,

William.

Thank you William, while I hope this is not the problem, and if it is then hope its not a serious one?

It was modded by a reputable company and if this needs to be looked at i'm sure they would be fine with me sending it in for checks.

I can think of two places - as mentioned - the viewfinder and the socket area on the side. If it is the gasket would these two places still be where the light actually gets in? or would it mean it gets in via somewhere Ive not even thought of yet?

Many thanks for the advice

Regards

Aenima

[Oddsocks]

HI Aenima,

Looking at the stripped down camera pictures in Gary Honis's site there seem to be quite a few places that light can leak in, the viewfinder is an option but usually when the mirror flips up during exposure this light path is closed.

You can prove that simply by fitting a standard lens with the lens cap fitted and covering the viewfinder with piece of opaque material (cardboard, Sellotaped over the viewfinder works fine) and then taking a long-time exposure in bright daylight at high ISO setting. If the artefact is still visible then you can discount the viewfinder route.

It can be really difficult to pin down the source of light leakage as the light can be reflected around the CCD chamber and may not be entering the body anywhere near the edge of the CCD affected.

If it were me with the problem I would send the camera back to the modifier...

Regards,

William.

[Aenima]

HI Aenima,

Looking at the stripped down camera pictures in Gary Honis's site there seem to be quite a few places that light can leak in, the viewfinder is an option but usually when the mirror flips up during exposure this light path is closed.

You can prove that simply by fitting a standard lens with the lens cap fitted and covering the viewfinder with piece of opaque material (cardboard, Sellotaped over the viewfinder works fine) and then taking a long-time exposure in bright daylight at high ISO setting. If the artefact is still visible then you can discount the viewfinder route.

It can be really difficult to pin down the source of light leakage as the light can be reflected around the CCD chamber and may not be entering the body anywhere near the edge of the CCD affected.

If it were me with the problem I would send the camera back to the modifier...

Regards,

William.

Found it!

The side panel where the rubber flap opens to fit plugs like remote shutter release.

Glad the entry point is solved, the eyepiece is nicely sealed to outside light, so its just the side panel. This gasket thingy, what is it and should it be what stops light from the side socket panel reaching the sensor?

Because thats more serious than if it just a case of sealing the side up the area is pretty exposed on my model so room for improvement there!

So whats a gasket and how does it block light to the sensor? I cant figure it out sorry

Regards

Aenima

[WaveSoarer]

Great news! Looking forward to seeing some great images!

[Aenima]

Great news! Looking forward to seeing some great images!

I just uploaded a new M3 image to my gallery as it goes.

And thanks!

Regards

Aenima

[Oddsocks]

Good news Aenima,

Glad you found the answer...

Regards,

William.

[Aenima]

Appreciate the ideas folks, ta.

My easy solution of blocking the gaps in the I/o panel (learnt its name yesterday, I/O something.... )

The one thing I thought about though is if the light is getting in there and showing up on sensor, and I block it up - is that the end of the problem? or will stray light from the rest of the camera and its openings (including via the optical path from through the telescope) end up still reaching the chip?

I mean if the light is reaching the sensor from the one place - then is that likely to still be exposed after I block the side panel. from other less obvious sources? as it's only the side that got sealed rather than the edge of the sensor itself?

This is what I mean when I say I dont know much about cameras

Regards

Aenima

[rfdesigner]

The only way to know is to fit the camera to the scope, leave the cap on the front of the scope and take some "images" in daylight.

Keep some duct tape or similar to hand.

[Aenima]

Well, after some bodge-jobbing with a bit of modelling clay I'm getting big reductions in blue patches as in pic above. It seems to be mostly the i/o port area thats leaking and covering all but the remote socket has improved it almost completely.

Still, I recommend to those with similar cameras after modding its a good idea to take extended darkframes, with (and without) a light source checking for leaks - I only noticed because guiding allowed longer exposures else it would've gone unchecked possibly affecting image quality.

Thanks to the above posters for the suggestions and input,

Regards

Aenima