A temperature question on Darks

[Notty]

So I understand that the sensor temp goes up with the length of the exposure.  My normal for galaxies is 10min subs.  I know we have to match the sensor temp during the darks and lights.  My question is (for DSLRs) how much of the temperature of the sensor is a function of the exposure time and how much is it affected by ambient temperature (within the usual temperate UK latitudes anyway).  If I went out now and started blasting off 10m long Darks (all other things setup wise being equal) would my DSLR sensor heat up to about the same temp as when I was taking my lights a week or so ago at night?

Strange question I know, but I want to investigate the possibility of building up a library during the day so as not to waste time during darkness.

Any ideas?  Thanks as always.

[PhotoGav]

The ambient temperature will have an impact on the DSLR sensor temperature, so if it's 20ºC outside now compared to 5ºC while imaging, there will be a significant difference in the darks. One solution that is worth exploring is to put the camera in the fridge and do the darks... Most fridges operate at 5ºC, which is near enough to a night time temp for imaging in the UK. Just a few degrees out doesn't seem to make a huge difference.

Hope that helps.

 

[vlaiv]

I don't think it is necessarily direct function of exposure length.

Think of it as dynamic equilibrium - when chip is operating, depending on regime, it is converting some of the power to thermal energy, and some of that thermal energy is being dissipated into surrounding medium. If power input is greater than power output - chip is warming up - if it is the other way around - chip is cooling down. Heat dissipation is in relation to the temperature difference between chip and its surroundings - so if chip is much hotter than surrounding medium - more energy is being transferred - this is why chip temperature stabilizes after some time (when it reaches temperature point where power input is equal to power output). For exposures up to certain length chip does not get hot enough to reach this thermal equilibrium, but for exposures longer than that - chip temp does not depend on exposure length any more - but rather other factors - like temp difference between it and surroundings.

So re question about 10m exposures in different conditions - answer is most likely not - it will be different.

Dark library is something worth building if your camera has set point cooling (so you can be sure that chip temp is always the same when recording both darks and lights - for example always set on -5C).

On the other hand you might give it a go if temperature in your shooting conditions does not vary by much (for example in range of couple degrees of C) - but in this case I would recommend you to use dark frame optimization (option in DSS) - it is an algorithm designed to deal with darks that have been taken in different settings (algorithm basically takes master dark and adjusts it so that calibrated frame has the least difference across it - for example if you have amp glow and subtract correct master dark - background should be uniform with no amp glow area either being lighter or darker than rest of frame - this sort of thing can happen if darks are not 100% correct - they will either have more or less dark signal, and when you subtract DC signal - you are ok, but when it has variations, you will observe that some parts are amplified due to that).

[alexbb]

There's an older topic that might help you:

The second link in the first topic does not work, but you will find a link to the GUI also in the first linked page.
Using this tool, I made myself a library with older darks, taken before I started using APT (which appends by default the sensor's temperature at the end of the filename).
During an imaging session I decide if the current lights have a good match in the library or I take new darks.
After I have enough (more than 50 usually) darks taken in a folder (ISO\exposure\temperature) I delete the raw files and keep only the master dark.

Luckily my skies are very light polluted so taking new darks won't take much longer than the bias frames...

Alex

[Notty]

Great information here, thank you so much for the replies all. I use APT and never thought to look at the file names to see the temp!

[wimvb]

This all may be irrelevant. The exif data shows the cameras temperature, not necessarily the sensor temperature.

Even ZWO who make astronomical CMOS cameras, admitted that the temperature they record is not truly on the sensor, but as Close as they can manage.

If you don't know the accuracy with which the temperature is measured, you may be off in your calibration frames.

During an imaging session, sensor temperature may change in a complex manner, both because of ambient temperature change and Electronics heating up or cooling down. The latter doesn't even have to be uniform, leading to such things as amp glow. Even such things as which materials the camera is made of, determines the temperature of and across the sensor.

In the end you may have to determine by experimentation what works best for your setup. In DSLR imaging, it isn't uncommon to skip dark frames altogether and just use bias and flat frames together with hot pixel removal methods (cosmetic correction). One of the reasons being that it just is too difficult to match dark frames properly.

[ollypenrice]

I'd try dithering on a large scale (12 pixels) and just use a master bias as a dark. This seems to work for lots of DSLR imagers.

Olly