dark current noise

Astrotripper wrote:

alexisgreat wrote:

To any of you who have any of the larger Oly M43 bodies like the EM1,5, or 10, how does the EPL-6 handle dark noise at longer exposures at higher ISO compared to those cameras?

It's the same for all Olympus cameras with the 16mp Sony sensor. E-M1 is an exception, as it uses Panasonic sensor, and thus exhibits higher levels of dark current noise. See here , here and here for example. You can use google translate, but even looking at the table and histograms at the bottom of those should be enough. An E-PL5/6 might heat up a bit faster, but I kinda doubt that the difference would be significant enough to be noticeable.

Edit: I did not scroll down to the histograms, i looked at the ISO graph first. Thanks for the link, I will read it later ...

(Caution, you are mixing high ISO and long exposure here. While EM1 is king with high ISO and dynamic range, it fails partially with long exposures. At least if you don't enable darkframe subtraction.

And I doubt that the modern sonsors heat up significantly: My imporession is that during long exposures, the sensor does nothing. It sits there and waits for exposure.

Where the sensor might heat up is video readout, EVF readout with frame rate high and during live composite/live bulb when fast 1 second reradout is going on.)

It's still been mostly cloudy around here so I haven't been able to test it- but basically what I'm looking for is if the noise is fairly low up to ISO 3200 and 1 minute exposures for pictures of constellations like Orion?

ISO 3200 might be too much for 60 second exposures. You might end up with the core of the nebula blown, at least when using a fast lens. For exposures that long, you can use lower ISO. Here's a shorter, 40s exposure at a very small aperture:

See it on flickr

And does dark frame subtraction help remove a lot of it in camera?

Dark frame subtraction is for removing hot pixels, not dark current noise. It actually increases noise levels. That's why for astrophotography, dark frame subtraction is done with master dark frames that are a result of averaging many individual dark frames. This process makes sure all you are subtracting is hot pixels, and nothing else.

This is the theory, but for the average OMD user who compares sensors and who does not fiddle with averaged dark frames ...

First, hot pixel handling is roughly done by "pixel mapping" in the utility menu. Most of what we are talking about is gone after doing this step...

Second: At least the EM1 behaves more complicated: While it has no master and average frames, it actually reduces the noise by applying a single darkframe. Either it does some extra magic algorithm/noise cancelling thing (the rumored "RAW noise filter") or some low level dark current noise is still prominent even after hot pixel removal. It seems to be much better to remove the predictable noise pattern even for short exposure times.

Oh, and obviously, temperature of the sensor makes a difference. Just leaving the camera on for a few minutes will make for increased noise levels at long exposures. Cooking a sensor for an hour in Live Composite mode will make high ISO completely useless for long exposures (talking from experience here). So a cold winter night is actually a pretty good time for long exposure photography

Yes, unfortunately, cold weather seems to help