Hot pixels. How to test? Any help appreciated.

[JamieNeb]

Hello, 

I am posting to request some help in checking the hot pixels on my dslr camera sensor. 

I have read that to test the sensor for hot pixels you should remove any lens and replace with the body cap, place the camera in manual mode, set iso (800+), place under a dark t-shirt (stop any stray light in viewfinder, make sure lcd is not showing (this step is to east potential amp glow), set to bulb, run long exposure (2mins+). 

Is this the correct practice for testing the sensor for hot pixels? 

If this is the case I don't suppose anyone would take a look at one of the test files for me and help identify if my sensor has an issue or not. 

Any help would be much appreciated 

[happy-kat]

Hi

If you already own the dslr amd are going to use for example DSS to stack the images taken then I wouldn't personally be worried about hot pixels as the software will deal with them.

[Alien 13]

With Canon cameras there is a process that re maps the sensor and removes a lot if not all of them, cant remember off hand but it uses the sensor clean menu.

Alan

[wimvb]

The body cap on most cameras lets light through. Better is to leave the lens on. Put the lens cap on and cover with aluminium foil, to keep any light out. Then make 1 - 5 minute exposures. The longer exposures will show hot pixels, as well as any amp glow. Amp glow may be more of a problem than hot pixels. The latter can be taken care of in calibration and stacking, as already mentioned.

[JamieNeb]

Thank you all for the replies. 

1 hour ago, happy-kat said:

Hi

If you already own the dslr amd are going to use for example DSS to stack the images taken then I wouldn't personally be worried about hot pixels as the software will deal with them.

Yeah I have read that calibration frames and post processing should take care of these hot pixels but having just brought a used astro modified dslr my paranoia has kicked it and gotta just makes sure its OK.

 

1 hour ago, Alien 13 said:

With Canon cameras there is a process that re maps the sensor and removes a lot if not all of them, cant remember off hand but it uses the sensor clean menu.

Alan

Literally just read about this. You have to keep the body cap on and run the manual cleaning like you say for 30 seconds, then turn off camera and capture another exposure unfortunately hasn't seemed to of done anything.  

 

21 minutes ago, wimvb said:

The body cap on most cameras lets light through. Better is to leave the lens on. Put the lens cap on and cover with aluminium foil, to keep any light out. Then make 1 - 5 minute exposures. The longer exposures will show hot pixels, as well as any amp glow. Amp glow may be more of a problem than hot pixels. The latter can be taken care of in calibration and stacking, as already mentioned.

I have just taken three 5 minute exposures at iso 800,1600,3200. I didn't use aluminium foil just covered with a black t-shirt will post raw files on a new post. 

[JamieNeb]

These are the test shots of the camera. Not sure how much of a test it is as only 15mins exposure time in total and surely there would be more thermal noise the longer in to a imaging session but hopefully there is enough information for someone to gauge if this abnormal in anyway.

Amp glow is also something that concerns me so does these images show any bad signs?

IMG_1245.iso3200.CR2

IMG_1246.is1600.CR2

IMG_1247.iso.800.CR2

[wimvb]

Lens caps aren't opaque to ir light. Neither are t-shirts. Metal is. When I take darks, I do it as I described, in a closed box, at night. I've learned the hard way.

My box brownie (aka pentax k20d), has plenty of hot pixels. But if I dither my light frames, they're not a problem. I just use aggressive pixel rejection.

[wimvb]

Can't open your raw images, so can't comment on amp glow. But if there are light areas along the edges or in the corners, this is most likely amp glow. It increases for longer exposures, or later in a session, as the electronics geat up. A possible cause is the battery, which can get warm. Solution to this problem is a power adapter and removing the battery. This also solves the problem of premature ending of an imaging session due to low batteries.

Btw, did you use the view finder cover? You can get light leakage without it.

[Stub Mandrel]

51 minutes ago, JamieNeb said:

my paranoia has kicked it and gotta just makes sure its OK.

It won't be... there's no such thing as a sensor without hot pixels...

[Alien 13]

Just now, Stub Mandrel said:

It won't be... there's no such thing as a sensor without hot pixels...

None on mine 

Alan

[ollypenrice]

Light does get into the most improbable places. I only ever do darks and bias off the scope with the metal chip window cover screxed in place.

Olly

[Stub Mandrel]

One could argue that any stray light that gets in also needs to be compensated for by your darks, so perhaps you should generate them simply by capping the scope...

There's a contentious suggestion!

 

At the very least we should be taking precautions to keep out as much extraneous light as possible...

[wimvb]

25 minutes ago, Stub Mandrel said:

One could argue that any stray light that gets in also needs to be compensated for by your darks, so perhaps you should generate them simply by capping the scope...

There's a contentious suggestion!

 

At the very least we should be taking precautions to keep out as much extraneous light as possible...

Light leakage can change with the direction the scope is pointing, depending on the light source. Can be difficult to calibrate out. Always better to fix problems at the source, as suggested by your last remark.

[JamieNeb]

Thanks again for all the replies very much appreciated.

If I was to go about testing the camera for amp glow without putting it on the scope what would be the best way to do so? 

Am I right at all in thinking that with the body cap or lens on (cap with foil) ,view finder covered and not using live view this is the bare minimum needed for a "good" dark frame and testing amp glow? 

My main concern for testing for amp glow is how do you produce a "true" representation of the potential glow if you just fire off a handful of exposures indoors at room temperature? Surely this isn't enough to heat the sensor to show the full extent of the glowing. Also, what would  be the preferred iso camera settings?

Ross

[ollypenrice]

9 hours ago, Stub Mandrel said:

One could argue that any stray light that gets in also needs to be compensated for by your darks, so perhaps you should generate them simply by capping the scope...

There's a contentious suggestion!

 

At the very least we should be taking precautions to keep out as much extraneous light as possible...

This might be arguable in the case of taking your darks in the dark....

Olly

[wimvb]

1 hour ago, JamieNeb said:

Thanks again for all the replies very much appreciated.

If I was to go about testing the camera for amp glow without putting it on the scope what would be the best way to do so? 

Am I right at all in thinking that with the body cap or lens on (cap with foil) ,view finder covered and not using live view this is the bare minimum needed for a "good" dark frame and testing amp glow? 

My main concern for testing for amp glow is how do you produce a "true" representation of the potential glow if you just fire off a handful of exposures indoors at room temperature? Surely this isn't enough to heat the sensor to show the full extent of the glowing. Also, what would  be the preferred iso camera settings?

Ross

You test for amp glow by taking dark frames. Dark frames should be taken at the same iso and exposure time as lights. But as this is just for testing, I would suggest start with 5 minutes exposures at your normal iso setting (800 - 3200). Take some 15 - 20 exposures with very short wait time inbetween. The long exposure time and fast sequence will heat the sensor enough to show any amp glow. Especially compare the first ('cold') and last ('warm') exposures. There should be a difference.

Btw, 'real' dark frames should also be taken at the same temperature as the light frames they are to correct. But that is not necessary in this test.

One cause of amp glow may be the battery. If you have a mains adapter for your camera, you could do a test with and without battery in the camera.

Just experiment and test, it's a great way to get to know your gear.

[spillage]

HI

I take it you are going to use this with a scope and motorized mount. If this is the case I would just suggest you dither between your light frames and just use bias and flat frames without any dark frames This should remove any issue with hot pixels.

spill.

[JamieNeb]

1 hour ago, wimvb said:

You test for amp glow by taking dark frames. Dark frames should be taken at the same iso and exposure time as lights. But as this is just for testing, I would suggest start with 5 minutes exposures at your normal iso setting (800 - 3200). Take some 15 - 20 exposures with very short wait time inbetween. The long exposure time and fast sequence will heat the sensor enough to show any amp glow. Especially compare the first ('cold') and last ('warm') exposures. There should be a difference.

Btw, 'real' dark frames should also be taken at the same temperature as the light frames they are to correct. But that is not necessary in this test.

One cause of amp glow may be the battery. If you have a mains adapter for your camera, you could do a test with and without battery in the camera.

Just experiment and test, it's a great way to get to know your gear.

 

Do you think it would be a good idea to vary the iso level? or doesn't this really matter as amp glow is more to do with temperature.

Is it best to place the camera in a dark cupboard? 

Would 2 minutes be a sufficient amount of time between exposures?

Apologies for the very basic understanding, still completely green with all this stuff.

 

48 minutes ago, spillage said:

HI

I take it you are going to use this with a scope and motorized mount. If this is the case I would just suggest you dither between your light frames and just use bias and flat frames without any dark frames This should remove any issue with hot pixels.

spill.

 

Yes the end goal is definitely to be fully up and running with a guided tracking mount setup.

I have only just purchased a used astro modded camera so trying to put it through its paces to see if its up to the job of imaging. It's just I still new to it all so trying to do this quickly as still have the possibility to return the camera if there is an issue. Amp glow seems the last thing on the list to check out now so I am deferring to the people in the know. 

Dithering is something I've heard mentioned a number of times now but completely clueless as to what it is really, sounds great though!! 

[wimvb]

Here's a blog post on dithering

https://wimvberlo.blogspot.de/2016/07/the-effect-of-dithering_21.html

On my blog there are more entries on dithering, just check. Basically, you move the camera (mount) about 15 pixels between exposures. During stacking, any hot pixels will end up in different locations in the final image, and be rejected by the stacking software. This by far the best hot pixel fix you can use.

As for time between exposures, I usually have a pause of 10 seconds. Just enough to save the image on the cameras sd card.

If you use longer times, the sensor cools down again. This is a good thing in itself, but you lose imaging time. Unless you let the sensor cool down completely, it will take longer to reach a stable temperature. For this test, I would take a short time between exposures. ISO is of less importance here. You need to stretch the final image anyway, so iso 1600 should be fine to start with.

 

[spillage]

I know the most popular software may be sgp but give apt a go as its mainly free to use. You can also use this to dither and set up you sessions. If you are using a canon I would suggest 800iso as starting point but it depends on the camera so I would just take some light at 400 800 and 1600 to see which suits you and your gear.

[JamieNeb]

1 hour ago, wimvb said:

Here's a blog post on dithering

https://wimvberlo.blogspot.de/2016/07/the-effect-of-dithering_21.html

On my blog there are more entries on dithering, just check. Basically, you move the camera (mount) about 15 pixels between exposures. During stacking, any hot pixels will end up in different locations in the final image, and be rejected by the stacking software. This by far the best hot pixel fix you can use.

As for time between exposures, I usually have a pause of 10 seconds. Just enough to save the image on the cameras sd card.

If you use longer times, the sensor cools down again. This is a good thing in itself, but you lose imaging time. Unless you let the sensor cool down completely, it will take longer to reach a stable temperature. For this test, I would take a short time between exposures. ISO is of less importance here. You need to stretch the final image anyway, so iso 1600 should be fine to start with.

 

 

Thanks for providing the link I will definitely check that out. Dithering really sounds like a must then if you're wanting a one size fits all solution to completely eradicate hot pixels, hopefully its not something to complicated to setup.

I have an intervalometer being delivered tomorrow if amazon are true to their word so can give them test shots a go then. I will try to replicate the "real life" expierence as much as possible by placing camera in a dark space, not using LCD, Covering viewfinder and your foil tip in the lens and see where I am after.

Also don't suppose you could explain what it is to 'stretch' a picture? This is something I would need to do to my dark frame to see if there's glow?  

 

1 hour ago, spillage said:

I know the most popular software may be sgp but give apt a go as its mainly free to use. You can also use this to dither and set up you sessions. If you are using a canon I would suggest 800iso as starting point but it depends on the camera so I would just take some light at 400 800 and 1600 to see which suits you and your gear.

 

Thanks for the software suggestions will definitely look in to both when the time comes just wanna get this camera tested and then the wallet can be somewhat given a little creative licence....!

[wimvb]

At about this point, it's time to point you to a good read:

'Making every photon count', by Steve Richards (Steppenwolf on this forum)

https://www.firstlightoptics.com/books/making-every-photon-count-steve-richards.html

It explains everything you'll need to know about astrophotography.