" />

CANON 5D MARK II - LONG EXPOSURE NOISE TESTS (PART 1)

[LINK TO PART 2]

I’ve had the 5D Mark II for a few days now and I finally have all the pieces in place to actually take some pictures. Last night the weather wasn’t very exciting so I decided to put together an initial round of “studio” long exposure noise tests. My “studio” in this case is an old wooden wine case filled with toys sitting on a table in my living room. Not the perfect set up but I don’t see any reason that these tests wouldn’t be valid as most long exposures are not made in strictly controlled studio settings.

Here are the basic specs:
- Ambient air temperature was about 70 degrees F
- ISO 100
- f8 for 2.5 second and 5 minute exposures, f16 for 10 minute exposure
- In camera noise reduction OFF
- Processed with DPP 3.5
- DPP noise reduction and sharpening OFF

Keep in mind that there was no sharpening and no attempt to extract the most detail from the images (although I can tell you they sharpen up wonderfully) so evaluate only on how the image holds together over the longer exposures.

Here’s a picture of the overall setup…
5D2 "Studio" setup

I chose four areas from which I took 100% crops for each exposure. For this test I used a 2.5 second exposure, a 5 minute exposure, and a 10 minute exposure. The histograms were roughly equal (as expected). Each crop is below, the first image is 2.5 seconds, the second 5 minutes, and the third is the 10 minute exposure.

My comments are at the bottom…I will probably make these rollover images instead of in line like this but didn’t have time today.

Crop 1

2.5 second exposure

5 minute exposure

10 minute exposure

Crop 2

2.5 second exposure

5 minute exposure

10 minute exposure

Crop 3

2.5 second exposure

5 minute exposure

10 minute exposure

Crop 4

2.5 second exposure

5 minute exposure

10 minute exposure

So as you can see for the most part the image holds up amazingly well!

My only complaint is this - and I saved this sample for the end of the comparison because chances are pretty good that if I posted it at the beginning everyone would have been looking for it but if I posted it at the end no one would even notice the issue until they looked at this sample and then went back to review the images a second time. In any case, in the longer exposures and in particular the 10 minute exposure some small purple speckles begin to appear (if color noise reduction was on they would appear white).

While I would have liked to provide a better example than the one below I only have access at the moment to the JPG files I saved and not the original RAW file. So, to better show the noise that is produced I took one of the JPGs, resampled it to double it’s size then sharpened the hell out of it and cropped it back to 500px. All this using an old version of Paint.NET. So don’t evaluate it for quality, only use it to identify what I’m talking about and then see if you can identify it in the other samples.

Here is the sample:

These specks are not really visible when viewed at “print size” or anything under 100% so I don’t really think they’re a big deal. The original 5D and 1Ds Mark II had this issue as well without using the in camera noise reduction for exposures of longer than about 8 minutes (which is where it becomes noticeable in these images.

I am going to re-do this test to add the following:
1. 15, 20, and 30 minute exposures with and without in-camera noise reduction - this will take a little while because I want to give the camera some time to cool down between the longer exposures just to equalize things.
2. 50% views of the images to better show when the specks would actually be visible in a print.

I will also post some outdoor images that I hope to take this weekend on a short trip to the coast (likely no internet so if I don’t respond to questions, etc, I will sometime Sunday when I return).

Leave a Reply


Comment moderation is in use. Please do not submit your comment twice -- it will appear shortly.