The noise produced by a Canon 1100D at various ISO settings and temperatures

[Gina]

Cheers

Had a bit of a problem last night. I had been imaging and forgot to turn the mount off when I set up the darks run and left it to do it's stuff overnight. The result was that during the nght the mount continued tracking in sidereal mode and the camera crashed into the pier and the focuser parted at its attachment to the adapter This resulted in the connections pulling out and the camera box and focuser fell onto the floor I've put things back together and pointed the scope at the far hill - seems to be working alright but I won't know if everything is all well until I get imaging again. I shall need a full star field to check that everything is still in alignment. Fortunately the forecast is good for tonight.

I really must be more careful Maybe I should arrange a safety switch that operates when the camera approaches the pier.

Anyway, the outcome as far as this thread is concerned is that it didn't complete the darks test run last night.

[Photosbykev]

If you are running EQMod that can be set up to limit the travel and I believe there is a timer function in EQMod for some related activity, maybe park?

[Ags]

Cheers

Had a bit of a problem last night. I had been imaging and forgot to turn the mount off when I set up the darks run and left it to do it's stuff overnight. The result was that during the nght the mount continued tracking in sidereal mode and the camera crashed into the pier and the focuser parted at its attachment to the adapter This resulted in the connections pulling out and the camera box and focuser fell onto the floor I've put things back together and pointed the scope at the far hill - seems to be working alright but I won't know if everything is all well until I get imaging again. I shall need a full star field to check that everything is still in alignment. Fortunately the forecast is good for tonight.

I really must be more careful Maybe I should arrange a safety switch that operates when the camera approaches the pier.

Anyway, the outcome as far as this thread is concerned is that it didn't complete the darks test run last night.

Horrifying! I hope everything is ok.

[Cath]

oops!

Can you not set slew limits for the mount Gina? .. the Celestron mount I have has limits I can set, although its done via the mounts handset. Maybe whatever software you use to control the mount has the option, I don;t know.

[Gina]

I must investigate setting it up in softwere - EQMOD or whichever part of the EQASCOM system does it Thank you folks

I varied my usual procedure last night - bad move! I can close the roof with the scope pointing in any direction but the east side fold down flap is obstructed in some positions so I usually park the scope first. The object I was imaging last night was not far from the celtial pole and I didn't notice the scope wasn't parked when I closed the flap.

[Gina]

I think everything is alright after last nights fiasco. I refocused using my Bahtinov mask and was blessed with a gap in the clouds long enough to capture a full field of pinpoint stars so the sensor still seems to be well at rightangles to the optical path. There wasn't a long enough gap between clouds to get set up and do serious imaging so I've set up for a repeat of last night 's aborted darks test run and it's all running happily at +6C EXIF T. Yes, I HAVE parked and switched off the mount

[Chris]

Thats a relief:), you've had more than your fair share of obsticles e.g. your RoR blowing away etc,

[Gina]

Thank you Chris

[Luis Campos]

Hi there,

Those are very interesting results and totally confirm what you can experience in the field when imaging with a DSLR, in my case an "oldie" 350D

I never use ISO above 400 ISO, even with the cooled sensor, using High ISO destroys a very important thing in astro imaging, dynamic range...also the stars get saturated very quickly and color get washed out.

Latelly I've doing some experiments with 100 and 200 ISO and very long subs, 900 sec. at least, the noise level is dramatically reduced, the colors are much richer, even with bright stars, and there's so little noise that you dont need a lot of subs to get a good S/N ratio, on this case maybe 5 subs of 900 sec. are better then 15 subs of 300 sec. at 800 or 1600 ISO.

The only down side is that the mount needs to be very well polar aligned or field rotation starts creeping on your subs :/

My advice is if you have a permanent mounted and well polar aligned setup, go for very long subs at low ISO 200 or even 100 ISO give the best results, make your experiences!

When the skies clear I'll post some test results with my setup.

Greets,

Luís

[dph1nm]

go for very long subs at low ISO 200 or even 100 ISO give the best results

Hmm - for Canon cameras, ISO100/200 are well below unity gain (1 photon == 1 ADU) so you will risk loss of depth due to quantisation. The read noise at ISO100 is also nearly a factor 10 higher than at ISO1600, so you really do need to make sure your exposures are long enough not to be read-noise dominated.

NigelM

[Gina]

I have now completed dark test runs at EXIF T temperatures of -10C, -4C, 0C, +6C and +12C, with all ISO values and subs lengths of 1m, 2m, 4m, and 8m.

Unfortunately, I have a problem now I was imaging last night - first really clear night for some time and got all set up on M33 with guiding working well, taking 3 4m subs and then a proper run of 2m subs at ISO 3200 and 0C but after 11 subs the image disappeared - just went black! This is the camera I'm using for the darks tests (the Peltier TEC cooled with set point temperature control). USB control is working but there's no image I'll have to strip the camera down and see if I can do anything with it.

Meanwhile, I'm afraid there will be no more test dark runs I was planning to take a series of runs from +10C reducing by 2C until I'd completed the job. Now the project is on hold until I can sort something out I'll post any progress as I make it - whether positive or not.

[Chris]

oh no, thats bad news! I hope its something simple like a ribbon cable connector come loose? Am I right in thinking you've got a fair few 1100D bits laying around? If so you could try swapping suspected problem components with ones that work from another camera? it might be pretty tedious though:(

[Shibby]

Not good Can you hear the shutter working? You must have racked up quite a shutter count by now, but fingers crossed that's not the problem!

[Gina]

oh no, thats bad news! I hope its something simple like a ribbon cable connector come loose? Am I right in thinking you've got a fair few 1100D bits laying around? If so you could try swapping suspected problem components with ones that work from another camera? it might be pretty tedious though:(

Yes, I've got a few bits but no longer have a spare sensor chip after trying to debayer one. I've taken the camera out of the box but gone no further yet.

[Gina]

Not good Can you hear the shutter working? You must have racked up quite a shutter count by now, but fingers crossed that's not the problem!

II've checked the shutter and it's working fine by the seems of it. I've also run live view from EOS Utility - that opens the shutter and exposes the sensor to full view - still just a black image. One thing though - at one point I got several pinpoints of light like hot pixels. But that was only once and nothing later. That could point to an intermittent contact.

[Gina]

Brought camera and netbook indoors and set up on the dining table. Powered camera with a battery and connected to netbook with USB cable. Switched on and EOS Utility started up as usual, set it to image and operated the shutter from the computer - viola! an image - a splodge of light. Set to live view and the light leakage showed on the screen So I replaced the body cap with a lens and - wayhey!! a picture

THE CAMERA IS WORKING! :) Phew!!

Next thing is to try waggling things to see if I can find a dodgy connection. There is a reason it wasn't working in the box and powered from my main supply etc. I need to find it

[themos]

Well, I've had a dodgy power supply (from astronomiser), it failed on me after about a year and hasn't worked since so you may consider that as the cause.

[Gina]

No it's not the power supply - I have two other cameras runniung off the same supply working perfectly.

[Ags]

What alarming news... I hope you figure it out without too much trouble

[Gina]

I'm doing tests now. I have the camera back in it's box and just about to connect up a bench PSU.

[kirkster501]

So what is the outcome of this work guys? Avoid ISO800 is the vibe I am getting.....

???

[Gina]

So what is the outcome of this work guys? Avoid ISO800 is the vibe I am getting.....

???

It's work in progress ATM as we gather much more data for more detailled processing and reduction of random errors.

[Gina]

Camera testing is progressing with no problem found. It's back in it's box with the usual external supply obtained by switch mode regulator from nominal 12v (13.8v) supply. I've yet to add the silica gel desiccant bags and apply cooling. I think it possible that a gel bag could have got lodged somewhere and caused a problem when the unit fell on the floor a little while back.

[Gina]

Camera testing still progressing... Silica gel bags in box and lid fastened down, cooling is now being applied to run the cold finger at just above Dew Point. Running APT on a darks schedule. All fine ATM

Other readings :- ambient 20.8C. Inside box - temp 22.4, humidity 29%, dew point 3.6C. Cold finger 4.2C

[Merlin66]

There's an active discussion on ISO v's noise going on over on the yahoo group Digital_astro:

http://tech.groups.yahoo.com/group/digital_astro/message/156489

I quote:

"" The problem with many digital cameras is fixed pattern noise, which

tends to be stronger at lower ISOs, balanced by lower dynamic range

at higher ISOs. The sensors are linear, so color is the same at all

ISOs, just noise will limit color information. So the best balance,

in my opinion, is the highest ISO where banding noise is low

(at least 10x lower than read noise). This is described here

(see Figure 2):

http://www.clarkvision.com/articles/nightscapes/

Look at the sensor analysis results for several cameras:

http://www.clarkvision.com/articles/index.html#part_4

Banding noise is shown for the Canon 1DX, 11DIV, 7D,

5DIII and 5DII. (click on the sensor analyses for each of

these cameras and look at Table 2 which shows banding noise

as a function of ISO.)

I ordered the list above, Canon 1DX, 11DIV, 7D, 5DIII and 5DII

in order of increasing annoyance of the banding noise,

with the 5DII coming in last (among these cameras).

So, the best ISO is the ISO where banding noise does not detract

from the image for the way you process images. For example,

if you stretch the image a lot, banding noise will become objectionable,

so use a high enough ISO for that not to be a factor. Second,

a long enough exposure (if you are able to) to gather as many photons

in each frame. Also consider noise from the sky in this evaluation.

Roger ""

Hmmm.

Ken