The Grande ISO test - 1Ds II, 5d, 1d II, 20d

I've been promising a high ISO test on my blog. A test to put to rest the chants of some 5D owners how their camera handily beats all non-14 bit cameras. It's also to put to rest any claims by the 1 series lovers that their IQ is the best. My only regret was not having access to a 40D. Pixels on target are insane with that camera.

What are the results you say?

The 5d is not the dragon slayer people think it is. It has high and lows. The 1Ds Mk II? It's a fast, high res colour friendly beast. The 1D Mk.II? Affordable 1series with GREAT IQ. the 20d? Buying used, how can you look away from such great IQ for $500!!

Scientific test? Sure. We had all cameras set to the same setting and metered the light every time with a Sekonic L-358. Tripod, RAW, etc. You can draw your own conclusions from my test. Were there variables? Hell yes I even got one blurry shot on a tripod. Some meters wanted "warm up" time.

Bottom line? My 5d is for sale and I'm keeping the 1Ds Mk. II

...full results here:

http://www.jasonhollis​ter.com/blog

My blog is not setup for comments. It's my commercial/business site and the blog is a traffci generator, not a full on blog in the "blog" sense. You can rip me apart here if you choose.

Gujustud is putting together a great hi-def video of us tackling this at my place. Also made with a Canon HV20, one cool little camcorder for the price. that will be added to the blog Thursday.

Interesting, because I still use the 20D.

Would really have liked to see the 40D since that is my weapon of choice now.

Denny

It just shows how little difference there is between them. In all but the most extreme shooting conditions noise is really not an issue.

"Now we head into enemy territory. ISO 400 and lower.Barbed wire, land mines and fanboy death squads roam the desolate landscape. The clean ISO ranges"

Great line! Thanks for the laugh.

quick, somebody make some popcorn and grab some beer, i have a feeling that it is going to get hot in here once the 5D people read this test

Interesting - thanks for the effort.

In your ISO 1600 shots, you state that the 1D2 "wins on noise" and yet, it is obviously holding more color noise than the 5D.

Otherwise, a respectable test.

One question - what lens or lenses did you use? Did you change focal lengths based on crop size to equalize the framing to take advantage of the higher pixel count of the two full framers? That's where I found the difference between the 1D2 and the 5D back when I compared the two. Pixel-to-pixel, they are quite similar. (I know, crop factor blah blah blah - but it does matter. The more pixels per subject, the finer the grain of the noise and thus the easier it is to reduce it in PP without softening the image).

blonde wrote in post #4702343
quick, somebody make some popcorn and grab some beer, i have a feeling that it is going to get hot in here once the 5D people read this test

Heh Heh.....

Wow, thanks for the time you took to do this, looks to me like the color noise in the 1DmkII shots is about the only difference, this sure makes the 5D look like a screaming deal. Again, thanks for the test...it is very interesting

Heresy! hehehe
Something tells me this one is going to stir up some controversy.

I love my 5D but I never thought of it as a "dragon slayer". I have always maintained that any difference in noise is bound to me minimal and YMMV.

Having said that, I do appreciate the noise performance for what I shoot (available light, lots of indoors)...every little bit that the 5D offers is definitely appreciated.

Cool comparos though...although I get a bit of a sense of bias here. (On both sides!)

Tom W wrote in post #4702344
Interesting - thanks for the effort.

In your ISO 1600 shots, you state that the 1D2 "wins on noise" and yet, it is obviously holding more color noise than the 5D.

Otherwise, a respectable test.

One question - what lens or lenses did you use? Did you change focal lengths based on crop size to equalize the framing to take advantage of the higher pixel count of the two full framers? That's where I found the difference between the 1D2 and the 5D back when I compared the two. Pixel-to-pixel, they are quite similar. (I know, crop factor blah blah blah - but it does matter. The more pixels per subject, the finer the grain of the noise and thus the easier it is to reduce it in PP without softening the image).

As stated in the blog, 50mm - pixel on pixel test. Crop does not play a part in measuring noise. Pixels are pixels. (Funny how this very topic was discussed between myself and others before the test and that people would still not get it - but Tom I know you do)

I see more noise in the critical areas on the 5D to 1D Mk.II image. I'm looking at the wall to table top area and I see BARELY more noise (right now on this computer) on the grey card on top of the books. If you are looking at the glass begind the table it's not smooth. Detail can be mistaken for noise there. There is a bit more chroma noise in the coral shadows and a small part of the book shadowwith the 1D M.kII, but the overall image for noise to detail the 1D Mk.II looks better.

Full size %200 on my dell 2407, the 1D Mk.II image is better.

As I said in the write up - it's open to interpretation. I may add this discussion to that secrtion of the blog perhaps.

Since the 5D is full-frame, you should move the camera closer to get the same FOV to get the same amount of exposure on that sensor. (j/k, going back to that MKIII vs D3 test that PM01 did on the lexusclub forum!)

Interesting test. The 5d really does control red noise better then the others, but it also seemed to do so at the the expense of slightly over exposing and blowing out some of the highlights. Really cool test. Cheers.

TeamSpeed wrote in post #4702448
Since the 5D is full-frame, you should move the camera closer to get the same FOV to get the same amount of exposure on that sensor. (j/k, going back to that MKIII vs D3 test that PM01 did on the lexusclub forum!)

No. This is a pixels on target report. Crop factor does not play a factor.

The 5D and the 1D Mk. II image nearly the same. you just see more of the room with the 5d. There is NO magnification.

I can supply you with the 5D and the 1d Mk.II image - you'll see they are nearly the same for pixels on target.

The 5D and 1Ds Mk.II are not close and neither is the 20D.

Think of it this way. You are both shooting birds from a blind. You and your buddy. He has a cropper and you have a full frame. Same lens, same settings. Who gets the better shot? Just because you have full frame you should move closer to the subject to be fair? No.

Croasdail wrote in post #4702470
Interesting test. The 5d really does control red noise better then the others, but it also seemed to do so at the the expense of slighting over exposing and blowing out some of the highlights. Really cool test. Cheers.

Sharp eye. We noticed that the 5D meters a bit liberally compared to the others is favour of noise control.

The 5d is the winner if ONLY taking into account general noise. Where noise and detail collide such as the 1D Mk.II 1600 shot we discussed it gets to be subjective. The colour of the 5d is not as good. The feature set isn't there.

But if you want to shoot dark closets full of shoes with a 50 1.4, and always rezise for web, the 5d is your best best i guess.