Extreme High ISO, low light test: 200-25600 on Canon 5D mark III and Sony Nex-7 up to 3200 ISO

Extreme High ISO, low light test: 200-25600 on Canon 5D mark III and Sony Nex-7 up to 3200 ISO

Four candles in a pitch dark room. I wasn’t sure what to expect and if this would work at all. I wont spoil it for you, sit back and enjoy :)
if you like what you hear, “meltingclocks” twistedtracks.com
more after the break (including 100% screengrabs)

The most challenging part during this test was keeping focus on the eyes, Cineroid EVF attached to the Canon 5D3 with Kit lens, shot at 105mm at f4, wide open,
a vintage Minolta 35-70mm f4 wide open on the Sony Nex 7 (via LA EA2 adapter). Ungraded footage, Color style, Neutral, Color tone -4, Saturation -4, contrast -4, sharpening 0, noise reduction and highlight priority turned off. no adjustments in post – footage looks even better at 50mbps, highest output for H264 in pp, however, the result would be a +1GB file. I am really impressed with the level of detail all the way up to 25.6k and the lack of noticeable color shift up to ISO 20k, only at 25.6k reds turn magenta! I can’t wait to start shooting with this camera!

Some of you already emailed asking about the softness issue on 5D3 files. I haven’t had a chance to shoot much yet. Based on this test I think the files will be perfect for glamour, beauty and fashion work, landscapes, cityscapes or anything detail rich, will need post sharpening, probably lots of it. I did notice focus is very critical and had to reshoot most of the clips, missing the mark by a few millimeters is unforgiving, take a look at the eyes, we have very good definition between pupil/iris and on eyelashes (and we are looking at un-sharpened files).

As for the Sony Nex 7, there is not much to be excited about when your only light source is a candle :) I love the camera and was curious how it would perform at high ISO, the clips and screen grabs below confirm what we already knew, not very good and looking at ISO 3200 we have a nice example of what high compression codecs do to the image……..

If you find this information helpful, please consider helping me maintaining this blog by purchasing your gear through my product links to B&H. It will cost you nothing and allows me to keep adding! Or consider making a direct donation using PayPal

19 thoughts on “Extreme High ISO, low light test: 200-25600 on Canon 5D mark III and Sony Nex-7 up to 3200 ISO

  1. Pingback: [Video] Moiré Comparison: Canon EOS 5D Mark III vs Nikon D800 vs Sony NEX-7 | CanonWatch

  2. mike.kobal Post author

    Hey Rock, take it easy, buddy, otherwise we’ll have to start calling you Angry Rock- motion in video is important to check for motion artifacts, usually high compression codec’s really struggle with motion

  3. Rock

    Boring? I guess the purpose of ur test is to entertain rather than scientific or simply just being ‘realistic’

  4. Beachrider

    Why don’t you just compare the Canon to the Sony pro video stuff? If price ratios aren’t very important, your readers would learn a lot about what separates pro video stuff from the Canon’s SLR-adaptations.

  5. mike.kobal Post author

    shooting this with stationary candles would be too boring and movement in a video clip is crucial, it can tell you a lot about motion artifacts, you can click though to vimeo and download the video file for your own frame to frame comparison.

  6. Rock

    I’m not at all trying to say such comparison is useless/invalid. I’m simply trying to point out it’s really hard to see the “real” results when you have screenshots with the two candles being in the different spot hence the difference in the amount of light that’s shining on the face. Maybe if you did the comparison with the candles being stationary we could see a more accurate result.

  7. mike.kobal Post author

    Hey Rock, most readers are aware of the fact that I test the gear I use for personal and professional work and that my approach is unscientific, looking at sensor data and mtf charts can be helpful but wont show me how it effects a challenging real life situation, I need to know how far my cameras can be pushed and how good the results are relative to much more expensive gear.

  8. Rock

    Despite the class&price difference of the two cameras, I also like how the comparison was made by screenshots with the candles constantly moving…very scientific I must say lol

  9. squig

    Hey Mike, I’ve been doing my own low light tests, the MKIII native ISOs are cleaner just like the MKII. 8000 ISO is particularly bad and noisier than 12800.

  10. mike.kobal Post author

    and the funny thing is many people are interested to see how these two cameras hold luggage, uh, I meant compare :) this comparison is a lot less dramatic then comparing a hacked Panasonic GH2 against a RED

  11. Jim

    I liken that test to being something akin to comparing a station wagon to a motorbike and concluding that the wagon holds more luggage.

  12. Pingback: sonyalpharumors | Blog | New Sony NEX-7 reviews (Fuji main compeitor now for sale)

  13. Pingback: More EOS 5D Mark III ISO performance tests and comparisons (NEX 7, D800) | CanonWatch

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>