7D v 5D Mlll

[S_J_P]

Re: 7D v 5D Mlll

Thanks SJP a very interesting article.

In real life I wonder whether any one on the Forum has cropped a 5D image to the equivalent of a 7D's 1200 x 800 or even less and what the IQ looks like. Mind you its beyond me how you calculate the equivalent - but there must be a man/lady who does

Dave

Not quite 5Diii vs. 7D, but you piqued my interest with your comment, so here's a series of 60D vs. 6D images shot in my garden this morning.

They're all shot from the same position, with the same lens, evaluative metering, AV, 2-sec timer on a tripod, manually focussed in 10x live-view on the pots lip, a few minutes apart. I went up through the ISO range with each camera and then imported the images into LR. I cropped the 6D images to give as near equal FOV as the 60D native images, but did not apply any other image processing (NR was set to zero). I then exported the images as 7.5Mp JPEG's with 80% quality.

Here's the series....

ISO100
Camera 1 ISO100.jpgCamera 2 ISO100.jpg

ISO400
Camera 1 ISO400.jpgCamera 2 ISO400.jpg

ISO800
Camera 1 ISO800.jpgCamera 2 ISO800.jpg

Higher ISO to follow....

[S_J_P]

Re: 7D v 5D Mlll

...and now the higher ISO images....

ISO1600
Camera 1 ISO1600.jpgCamera 2 ISO1600.jpg

ISP3200
Camera 1 ISO3200.jpgCamera 2 ISO3200.jpg

ISO6400
Camera 1 ISO6400.jpgCamera 2 ISO6400.jpg

[DP the Snapper]

Re: 7D v 5D Mlll

Thanks a lot for your trouble Steve.

I assume the top image in each set is the 60D?

Viewing on my screen it looks as though the top image in each set is slightly sharper and maybe a little darker - but really there isn't much difference.
Both cameras perform superbly at high ISO.

What is your conclusion?

Dave

[DP the Snapper]

Re: 7D v 5D Mlll

Steve, as a matter of interest could you post a shot from both cameras showing the whole image - to see how much you have cropped?

Thanks

Dave

[S_J_P]

Re: 7D v 5D Mlll

Now that's interesting!

I think the slight brightness difference is down to the wider native FOV on the 6D taking into account more of the background in the evaluative calculation. If I ran another series I think I'd go for Manual exposure.

Looking at the images here, there's little to differentiate them in terms of sharpness, unsurprising I guess as with the re sampling and reduction in pixel count any minor differences in the original sensors may have been lost.

What is interesting to me is that with the original full resolution images, I think the 6D's sharpness surpasses the 60D's despite the 50% reduction in effective pixels as a result of the 1.6 crop, even when zoomed-in. Ok, by a 1:1 zoom there's a significant difference in magnification, but with the 6D at 1:2 and the 60D at 1:3, the 6D does look "better", how it's "better" isn't entirely clear though.

There are other factors at work here too, the shadow detail in the 6D images seems to be retained to a greater degree, and as already mentioned the performance difference at ISO400 and higher makes the 6D's images far cleaner and more pleasing on the eye, and retains the contrast and saturation to a greater extent.

Another factor I've not taken into account is the effect of NR on the images (as I turned it off), this will smear the 60D's images to a greater extent.

One last point of note is that the file sizes of the 6D are smaller than the 60D's even at the same megapixel count, and the difference increases with increasing ISO and thus noise. By ISO 6400 the difference is around 25%. If course noise adds "detail" and so compression isn't as efficient.

Overall I have to say that I feel that a cropped 6D image is not inferior to an uncropped 60D image at low ISO, and the 6D image is superior to the 60D's as higher ISO. I'd therefore not hesitate to use the 6D and crop-in in preference to the 60D. Of course there's a limit to how practical this is in the extremes, and I suspect that at anything greater than a 2x crop, the the greater native pulling-power of the crop sensor model will probably be preferable, it's all a matter of maintaining sufficient pixels for the desired image medium.

[S_J_P]

Re: 7D v 5D Mlll

As requested, two full-frame images, full res, at middle ISO of 800. I'll stick them onto Flickr so you can see them full-resolution.


60D ISO800 by S.J.P, on Flickr


6D ISO800 by S.J.P, on Flickr

[DP the Snapper]

Re: 7D v 5D Mlll

I suspect that at anything greater than a 2x crop, the the greater native pulling-power of the crop sensor model will probably be preferable, it's all a matter of maintaining sufficient pixels for the desired image medium.

Thanks again Steve.

I am afraid I am not very technical but using Lightroom 'sizing' if I cropped an image in a crop camera to 1.2 or even 1.1 and cropped in a full frame to have the same subject size the result would be a better image with the crop camera? For my question the images would only be for use up to computer screen size.

Dave.

PS Different subject .... those are great fox images on your Flickr site.

[S_J_P]

Re: 7D v 5D Mlll

Aha, I see what you're getting it, and that explains a comment you made earlier!

If your images are destined for screen use (at say 1,920 x 1,200) then you could crop-in to 2.3MP-worth of image area before you'll see a degradation in image quality as you would need to interpolate (fill-in) the missing pixels. Whether this 2.3MP area is from a full-frame or a crop sensor camera won't affect it's perceived quality, as you're merely displaying pixels. The improved IQ from the full frame camera sensor would improve the quality of the displayed result though if the crop and FF images are on a like for like basis (at the same number of pixels).

For example, let's assume you have an image shot with an 18Mp crop body and the same image shot with a 20Mp FF body. If you wish to crop the 18Mp image to 1/2 of the original size (9Mp, in effect a 2x "magnification") then you'll have to crop the full-frame 20Mp image to 20% of it's original size to achieve the equivalent "magnification" (there's a 2.6:1 difference in sensor area between drop and FF sensors, the usual "1.6 crop" measure is relating to each edge measurement, so the area difference is 1.6^2, or 2.6:1), that is equivalent to around 3.8Mp which is still more than you require for screen display or for a decent 300dpi 5x7 print.

Once you have to crop either image to such an extent that you have less pixels than you need (2.3Mp for display, 3.2Mp for 5x7 print, 2.2Mp for a 6x4 print) then I would expect the crop sensor image to look better than the full-frame image, but that's not until a hefty magnification (equivalent to a crop to 13% of the 18Mp image 12% of a 20Mp image for display purposes) that you'll notice degradation.

Overall I'd reiterate that I'd rather crop-in a FF image than take a native crop-sensor image, unless the cropping is extreme (or I'm printing at a large size).

Thanks for the fox comment Those images are a case in point, they're taken full-frame and cropped to around 50% of the original size, and I consider that to be preferable to using a crop-sensor camera.

[DP the Snapper]

Re: 7D v 5D Mlll

Thanks Steve that is just what I have been wanting to know for ages

And thanks also to Dave S for starting the thread - good luck and shots with the new camera

Dave

[Riggers]

Re: 7D v 5D Mlll

Great explanation Steve. Thanks.

[S_J_P]

Re: 7D v 5D Mlll

Thanks for bringing it up!

I'd often thought that cropped FF would be similar to uncropped crop-sensor, but this spurred me on to actually test it

[Riggers]

Re: 7D v 5D Mlll

Could have tested it myself but too lazy.... I mean busy :)

[Nathaniel]

Re: 7D v 5D Mlll

Steve,

I notice that the algae/moss on the side of the plant pot is a bit more green in the all of the second pics as you move along increasing the ISO. I assume that the second image is the 6D.

[S_J_P]

Re: 7D v 5D Mlll

Steve,

I notice that the algae/moss on the side of the plant pot is a bit more green in the all of the second pics as you move along increasing the ISO. I assume that the second image is the 6D.

Yes, the 6D images are the second ones Nat. I took the images using evaluative metering and the wider FOV of the 6D led to a slower shutter speed due to taking into account more of the dark background - I should have used manual.