Canon 5D Mark II Vs. Canon 7D Vs. Fujifilm X100 - 100% Crops.

[Marloon]

Disclaimer: PLEASE BE AWARE THAT THIS IS NOT A SCIENTIFIC TEST! THE FOLLOWING ARE JPEG OUTPUTS FROM THE MENTIONED CAMERAS AS THE X100 RAW FILES ARE NOT YET SUPPORTED BY ADOBE LR AND CS5. RAW PHOTOS WILL BE POSTED ONCE ACR IS RELEASED.

ORIGINAL IMAGE




ISO 200


ISO 400


ISO 800


ISO 1600


ISO 3200


ISO 6400

My notes:
- The buffer on the X100 is definitely horrible. I am using a class 4 SDHC card (15mb/s | 16gb) and it takes the camera 7 seconds to store the ONE raw + ONE large jpeg file onto the card. I am still able to take photos however, the buffer freezes the camera up and prevents me from changing ANY settings. This has become an issue in the field. A faster card is on it's way (45mb/s | 16gb) and that supposedly lowers the time down to 2 seconds - which is faster, but is still a slouch.
- The 7D and the 5D are definitely a lot faster with bigger buffers and of course, I didn't experience any slowdowns - which is of course expected.
- The X100 does have a significant amount of in-camera sharpening when it outputs jpegs but all of the settings were set to it's standard.
- It appears to me that the X100 does not have chroma noise but it's more grain like in comparison to the other cameras.
- After looking at these files for about an hour, I still don't know how many stops of advantage in terms of noise you'd get from an X100 compared to the 7D - to me, it looks like one stop.
- Oddly enough, the white grate behind the light stand starts to disappear ISO 3200 on the Fuji X100 and detail is lost by ISO 6400

I hope that you've enjoyed this test as much as I have. It's definitely feels nice to have a pocketable, lightweight that offers high image quality - one that you don't have to worry about/that is comparable to something like a 7D/5D, and has both manual and automatic features. I hope that Fujifilm will continue to develop this lineup - one that offers interchangeable lenses as there definitely is a market potential for these types of cameras.

[jwcdds]

Interesting indeed. First off. Thank you. A very gracious thank you for going through the trouble of doing this. Secondly, I think you mislabeled the image. I gather the right-most image is the 7D/24L/f4. You mislabeled it as the 5D2.

Since I have no qualitative way to measure ISO level, I can't comment on whether it's actually better or worse. My eyes see more of the lightstand in the cropped image due to the 12mp resolution, so my mind tells me there's more contrast (but I also wonder what my mind would think if the 5d2/7d images were downsized to the 12mp resolution, or the x100 uprez'd).

You do point out an interesting thing happening, with regards to the closet rack disappearing at iso6400. I suspect the Fuji has additional contrast/sharpening at work here. Since my understanding is that the 5D2 is pretty darn sharp at the pixel-level, yet it looks terribly soft (as the 7D) compared to X100 image. It's either that or both the 35L and 24L are inferior to the lens on the Fuji (which is a possibility).

[ICBuyer]

well, the focus on the right most image is off.

[Marloon]

Originally Posted by jwcdds

Interesting indeed. First off. Thank you. A very gracious thank you for going through the trouble of doing this. Secondly, I think you mislabeled the image. I gather the right-most image is the 7D/24L/f4. You mislabeled it as the 5D2.

You are most welcome. It took me about an hour and a half to complete testing, cropping, compiling and posting. Phew! Thank you for pointing out that I've mislabeled one of the images - that has now been corrected.

Originally Posted by jwcdds

Since I have no qualitative way to measure ISO level, I can't comment on whether it's actually better or worse. My eyes see more of the lightstand in the cropped image due to the 12mp resolution, so my mind tells me there's more contrast (but I also wonder what my mind would think if the 5d2/7d images were downsized to the 12mp resolution, or the x100 uprez'd).

That's a possible request that can happen. Since I can't really make NEW pixels, I will downrez the other files. That'll have to wait though, I have to study for my finals (it'll end on the 8th of April)

Originally Posted by jwcdds

You do point out an interesting thing happening, with regards to the closet rack disappearing at iso6400. I suspect the Fuji has additional contrast/sharpening at work here. Since my understanding is that the 5D2 is pretty darn sharp at the pixel-level, yet it looks terribly soft (as the 7D) compared to X100 image. It's either that or both the 35L and 24L are inferior to the lens on the Fuji (which is a possibility).

It is rather interesting. The high resolution sensors on both the 5D and the 7D push the 24L and the 35L to it's maximum resolving potential.

Originally Posted by ICBuyer

well, the focus on the right most image is off.

As for the focusing, I HAD to make sure that I was focusing properly. I let the camera focus itself and then adjusted using live-view at 10X. I switched the lens to MF mode, and left it at that as I kept pushing the iso and the shutter to keep the focus consistent.

I evaluated the DOF using the DOF calculator. The subject was 1.5 m away. The 7D was set with a 24mm FL and a f/4.0 Aperture.

Depth of field
Near limit 1.26 m
Far limit 1.86 m
Total 0.61 m

In front of subject 0.24 m (40%)
Behind subject 0.36 m (60%)

Hyperfocal distance 7.6 m

So I am pretty sure that I am fine in this aspect.

[arentol]

Originally Posted by Marloon

That's a possible request that can happen. Since I can't really make NEW pixels, I will downrez the other files. That'll have to wait though, I have to study for my finals (it'll end on the 8th of April)

I would upsize the X100 (a lot) and the 7D (a little), to match them to the 5d2. This is the only fair way to do it because you are comparing cameras here, and penalizing a camera for being better at something (more MP) is completely contrary to the goal of the test. Since the X100 provides the least MP it should have to raise itself to the level of the camera that provides the most.

[jwcdds]

Originally Posted by arentol

I would upsize the X100 (a lot) and the 7D (a little), to match them to the 5d2. This is the only fair way to do it because you are comparing cameras here, and penalizing a camera for being better at something (more MP) is completely contrary to the goal of the test. Since the X100 provides the least MP it should have to raise itself to the level of the camera that provides the most.

But on the flip side, downsizing the files from the 5D2 & 7D will offer great improvement on actual noise level of the image itself. The comparison was originally triggered by a discussion of ISO/noise performance between the 3. Up-sizing the x100 will expose the loss of detail (as expected). Down-sizing the 5D2/7D files will reduce the perceived noise in the overall image.

[Stamp]

Thanks for the test, Marlon.

Why does it matter that there is no aftermarket RAW support for the x100? Surely it came with Fuji RAW converter, right? Why aren't we taking RAW images, and simply converting them with no PP'ing? And no, smarty pants, I don't consider converting PP'ing, so keep your trap shut

[bohdank]

Different RAW converters will give different results. Ideally you want to use the same software for all the images.

[Stamp]

Originally Posted by bohdank

Different RAW converters will give different results. Ideally you want to use the same software for all the images.

Didn't know that. I figured if no setting were tweaked, and the were just converted, they'd be relatively the same.

[bohdank]

Different software, different algorithms used in converting RAW to something useable.

[JBF]

The Fuji looks to me that it has in camera sharpening on. It (fuji) looks to me to win the high ISO test.

[davidfig]

Originally Posted by JBF

The Fuji looks to me that it has in camera sharpening on. It (fuji) looks to me to win the high ISO test.

That's what it looks like. Can't wait for the all the fanboys to tear this apart.

Background grate aside, this looks pretty amazing considering a $1000 camera is pitted against a $3000(7D) and $4000(5Dii) setup. But that will be lost on everyone.

Would be nice if the OP would comment on white balance.

[eelnoraa]

I think focus is fine. Today's high pixel count SLR always give a little soft image at pixel leve if no sharpening applied. To me, 5Dii is only on par with X100, or sligtly ahead (at most) up to ISO1600. Beyond that, X100 is noticeable better. The noise charasteristic of the x100 is also better than that of the 5Dii. The pattern seems finer, less blotchiness. And both are better than 7D.

Yes, I agree, in term of low light performance, X100 wins here. Again, Canon is really losing ground in low noise department. I am just so disappointed at Canon for making so little improvement to its sensor compared to there camera makers.

[yourdoinitwrong]

The results from the X100 at 1600 and above are very impressive for sure. I'm still happy with my 7D but it's interesting to see what can be accomplished with an APS-C sensor. Hopefully ACR will bring out support for the X100 soon so we can see the RAW comparison.

[Marloon]

Originally Posted by davidfig

That's what it looks like. Can't wait for the all the fanboys to tear this apart.

Background grate aside, this looks pretty amazing considering a $1000 camera is pitted against a $3000(7D) and $4000(5Dii) setup. But that will be lost on everyone.

Would be nice if the OP would comment on white balance.

To give you the straight answer, the AWB on the X100 is fantastic! Under mixed lighting, the can pick out the perfect WB setting. To make sure that I wasn't delusional, I went over to AlanU's house so that he, too, can play with the X100. Here's what he had to say about AWB (and more) - I swiped this from an earlier thread.

Originally Posted by AlanU

For my style of shooting (non sports) I will have to say the x100 has better/superior ISO performance than a 7D and it completely blows away my 1dmk3 in terms of IQ, ISO performance,AWB and destroys the 1dmk3 when it comes to metering. The 1dmk3 has horrible metering for a "pro body". I just did a quick snap of my daughters using a 1dmk3 with cheap (sarcasm) 35L vs x100. The x100 metered perfectly and as far as dynamic/tonality the x100 was incredibly better. I shot ISO 2500, f4 and the noise (jpg set to low NR) is as clean as my 1dmk3 at ISO 1000-1250!! ISO 1600 in demanding light the 1dmk3 is useable hitting it hard with NR.