Noise in long exposure

[sai]

I know that you get more noise when taking long exposure shot, because the

sensor heats up. Now many people have told me I could take several short

exposure shots and then put them all together in PS. But if I shoot for say 10

minutes, stop and then instantly shoot again for 10 minutes and so on so forth,

woul't the sensor be as hot as if I had just taken a normal long exposure shot?

I mean 3 10 minutes shots, taken continusly would be the same work/heat for the

sensor than a 30 minutes one. Or am I mistaken?

thanks for your comments

[bobatkins]

You are not mistaken.

 

However you will reduce random noise by the square root of the number of frames you stack just because of the statistics of random noise. You will not reduce hot pixels, stuck pixels or pattern noise.

 

I'd assume most thermal noise shows up as random noise, but I don't know that for sure.

[dave_holland]

I think you're right, for the short stitched shots to give less noise you need to allow time gaps between exposures, to allow sensor cooling. Remember also the value of a darkfield view to partially negate sensor noise difficulties. You can do a sinlge darkfield view at the end, or use the CF utility in the camera. But many astrophotographers don't use the latter CF function, because it takes so much longer to capture an image. It's quicker to do the darkfield subtraction later.

 

Dave

 

 

Dave

[colinsouthern]

Hi Simon,

 

There are quite a few things that cause noise - sensor heat being one of them. The joy of multiple exposures though is that you're averaging out the noise (it's pretty much a random thing).

 

You simply stack the images as layers in photoshop, and change the opacities: 100% for the bottom layer, 50% for the one above, then 33%, 25%, 20% etc.

 

I did this with a series of 32 noise images taken at ISO3200 on a 20D - by the time I'd stacked them and flattened them I had an image with the equivalent noise of a 100 to 200 ISO shot.

 

Of course, in this example the noise was caused by amplification of the analog signal as part of upping the ISO, but - in general terms - it's the same as what you're talking about.

 

Hope this helps,

 

Cheers,

 

Colin

[sai]

Great! So although I'm not getting rid of the noise per se on each picture I do get rid of it randmly stacking them together in PS. That's good news, also to be able to manage the battery consumption better. I'm thinking on star trail photography, where no pauses can be taken.

Regarding to the PS technique, Colin, " You simply stack the images as layers in photoshop, and change the opacities: 100% for the bottom layer, 50% for the one above, then 33%, 25%, 20% etc." is that all? Yippie! =)

[lancemcvay]

<i>It's quicker to do the darkfield subtraction later.</i>

<p>

How is this done? How variable is the noise produced by the sensor, say, outside over a period of six hours, with exposures from three to six minutes? Would you just need one dark frame for each exposure time? Because that would save me an awful lot of time on some work I've been doing lately.

<p>

<center><img src="http://abandoned-alaska.com/coppermine/albums/10-04-2007birthofwinter/IMG_2292.jpg"></center>

[sai]

And how long should the exposure on the dark frame be?

[colinsouthern]

Hi Simon,

 

My technique was specifically for identical images - to be honest, I haven't thought through the implications (if any) of using it for subjects with apparant motion (eg "star trails").

 

Probably easiest thing is just to try it :)

 

Cheers,

 

Colin

[colinsouthern]

Simon,

 

Just thinking briefly about this some more, I'd say that you're going to need to change the blending mode to something along the lines of "Lighten" - not sure what will be needed by way of opacities.

 

You might get some more difinitive answers in the digital darkroom forum?

 

I'll try it for myself when I get a chance.

 

Cheers,

 

Colin

[zigzag]

Just some thoughts:<br>

over these long periods there are a number of variables that will affect the frame:<br>

1. Thermal noise ('drift'). With the camera in a constant ambient temperature and having been on for a while, the drift will probably vary around a mean as the A/D feedback circuits compensate.<br>

2. Quantisation noise should be almost random<br>

3. The luminosity of the sky will vary over time and will form its own 'drift' that could be significant. (Possible WB also, battery voltage reducing, ...) <br>

4. If random noise cancels to the square of the number of samples, the same may not be true of these other sources which could be cumulative as they are not random - for these the law of diminishing returns would say that there comes a point when more samples will add rather than reduce noise.<br>

5. The thought behind dark frame subtraction is that the dark frame is done right after the first as this is the closest possible time when all of the variables are as close as possible to what they were when the picture was taken.<br>

6. If you allow the sensor to cool, heat, cool, heat you are introducing more variability than if you do a drak frame reduction directly after the shot, not letting the electronics cool to a lower temperature - you want it to be at the SAME temperature.<br>

7. For Colin's suggested method - it is true that straight addition would dim the star trails and 'screen' or 'lighten' could be worth a try - however this will exacerbate light interference as it always the lightest thing that will get emphasised.<br>

8. The exposures on the dark frame should be the same duration as the original exposure.<br>

<br>

Agree with Colin - try it. If you expect to take 100 frames then maybe do 1 in 10 dark frames equally interspersed - with one at each end = 11 dark frames - use these for your cancellations.<br><br>

 

This is not any definitive answer - just a suggestion.

[zigzag]

Or you can buy a film camera: <a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00MGrZ">Canon EOS</a>, mine cost me $27.

[obakesan]

Hi

<P>

Hey, this is interesting, I hadn't been doing astro stuff with my 10D, but it never occured to me that the sensor would heat up on exposures. I guess that it must!

<P>

So, does this mean that I can have <I>some sort of benefit</I> here in Finland when doing long exposures at night when its below -10 deg C

<P>

I've been wanting to do some 'northern lights' shots this winter and was not looking forward to the usual trials of 'battery failure' and freezing my nu#$ off in the cold when I should be in beside the fire.

<P>

If so, then that <B>great</B> ... now I'm keen to go out n try!

<P>

:-)

[colinsouthern]

Chris Eastwood wrote: "I've been wanting to do some 'northern lights' shots this winter and was not looking forward to the usual trials of 'battery failure' and freezing my nu#$ off in the cold when I should be in beside the fire. "

 

That's where the TC-80N3 comes in - you can program it to wait xx minutes and then take xx number of shots or xx duration, xx minutes apart - whilst you enjoy all the benefits of sleeping bags and hot water bottles (at worst) :)

[obakesan]

geez I dunno ... tempting as that sounds it would make me feel like I'm cheating somehow.

<P>

Where's the rugged endurance, the will to overcome the obstacles in pursuit of that image. It lacks something a <I>je ne sais quoi</I>

<P>

It would be like my friends finding that I stayed in a hotel rather than camping.

<P>

Still ... it has attractions ... Perhaps I could keep it 'our secret'

[zigzag]

"the sensor would heat up on exposures ... when its below -10 deg C"<br>

 

No such luck, it's a 'well known fact' (sic) that noise becomes less random as you approach the arctic circle - something to do with the aurora borealis. The best thing you can do is put your camera in the sleeping bag with the hot water bottle, making sure that the lens itself is at a slightly higher temperature so it does not fog up. This will help the batteries also. Rather than freezing your nu#$, you may want to spend some time roasting them over a hot fire - and pop out at a frequency no less than every 6 minutes to make your latest exposure, pointing towards Polaris.

[obakesan]

Glen

<P>

thanks for the advice ... this thread is going downhill rapidly, I wonder if I should push it along (like my Finnish friends did on my first downhill sled ride)

<P>

btw ... how's the 10 goin? I've been tempted in that area and now (being OS n all) that I don't have a film body with me am wondering if I should look for a 10 on ebay or for a 600 (called 630 at home) as I just like them (I have 2 630's).

<P>

but seriously, will the ambient temp make any difference? I can't imagine that the astro fellas can cool their sensors as well as mother nature keeps things cool here

<P>

thanks

[zigzag]

My EOS 10 is going well - I just bought a Hong Kong Remote Control gadget that I need to try out with it. It's nice to have a film camera that appreciates my 50mm f/1.8 without turning it in to an 80mm. Either of those bodies are fairly chea - I carry my 10 around in my bag and catch some interesting shots (we've had night lightning storms this week)<br><br>

 

I should imagine that your ambient would only work in your favour - noise wise. battery and/or camera failure would be likely to be a more significant issue as you must know - commercial grade ICs are commonly rated 0 to 70 degrees C, and as Canon claim no weather sealing or protection for many of their cameras, leaving it exposed for up to 6 hours at -12 degrees could be an interesting experiment - unless the camera body was wrapped in a thermal protector (that body heat could keep it alive). (See reports of equipment failures on Antarctica trips at www.luminous-landscape.com for M. Reichmann experiences.)

[obakesan]

Glen

<P>

thanks for pointer to LL review on Antarctic trip, I hadn't thought of searching for this ... how odd. Yes I'm well familiar with equipment failure in the low temp. Last year we were at a place in Finlands most westerly border and it was -25 or something stupid. The only camera that kept working (from our group) was my girlfriends A540 - my mates Nikon and my 20D shut down. We were sled / ski traveling so no option but for gear to be out in the cold all the time.

<P>

I gotta tell you it had me wishing for a mechanical camera such as my OM-1 again!