M6 ii

[Kayaker72]

Compare M6 II vs. 5DIV both at 10ft, as this is often the real world situation. You cannot get closer to the subject without one of the following:

a) Falling from cliff, in river, off boardwalk, etc.
b) Scaring "it" off
c) Making yourself "it"'s lunch.
d) View is obscured from further away (trees, crowd, etc.)
... You get the idea.

Considering this 10ft vantage point, what gets me a better image? Possible comparisons - Crop 5DIV to match M6II (which gives best detail at 100%). Downscale M6II to match center of 5DIV (which gives best image for a equally sized print, assuming you're cropping the 5DIV to 1.6).

Your 10ft vs. 16ft test is the opposite of c), and/or reducing the odds of b). It's the "I don't feel comfortable with how close I am, can I move further away if I switch bodies?" / "Can I make my subject more comfortable by moving away?" situation. This is also a good test.

So, I was trying to keep what I was looking at pretty simple: I wanted to assess if the "true" crop factor of modern APS-C sensors was still ~1.2x or if it had improved. My conclusion, it has improved. While I did not show it, the 1.2x clearly has more detail. Yet, I can look at Bryan's comparison of the two cameras with the 200 mm f/2 at f/2 and the 5DIV image has more clarity. Stop down a bit, and this gets a lot closer, especially at the edge of frame, but zoom in on your screen (Ctrl +) and I would slightly give it to the 5DIV. There is some loss with going to smaller pixels. This gets down to the argument of "not all pixels are created equal." So, I wanted to know, within the means of my testing set up, how unequal. Because we can take the resolution and treat pixels as equal and calculate "pixels on target" but that analysis is flawed if pixels are not equal.

So, based on my little test, I am thinking modern APS-C sensors are getting much better, I would put it at 1.4x to 1.5x being the true multiplier for additional "reach," or we are getting 70-80% of the ideal benefit of additional "pixels on target."

But, I understand what you are saying and I did take photos of the target with each camera on 1 ft increments from 9 to 18 ft. So, I do have the comparison you referenced.

Using a different crop, which is a bit bigger:
5DIV at 10 ft

M6II at 10 ft scaled to be the same size as the 5DIV

M6II at 10 ft full resolution

[Joel Eade]

I see better detail in the 5DIV image at 10 ft than I do the M6 II at 16 ft. From my standpoint, I would say the resolving power is somewhere between 14 and 15 ft, or 1.4x to 1.5x being the "true" crop factor for the sensor on the M6 II. This is the same conclusion I reached looking at other crops. So, in my mind, I am just going to call it 1.45x. Still more resolving power than a 1.4x TC (which is actually 1.36x and does also put more glass between sensor and the subject).

Very interesting .... I thought of crop factor mostly as changing the field of view and not in terms of resolution. In that regard the M6II image from 16 ft looks to have similar field of view as the 5DMKIV at 10 ft thus the 1.6 factor. It does, of course get more pixels on target from a greater distance too assuming the crop sensor has smaller pixels or higher pixel density.

[Joel Eade]

Here is an interesting read about crop factor that correlates to what you are finding...

http://arihazeghiphotography.com/blo...g-crop-factor/

[Kayaker72]

Very interesting .... I thought of crop factor mostly as changing the field of view and not in terms of resolution. In that regard the M6II image from 16 ft looks to have similar field of view as the 5DMKIV at 10 ft thus the 1.6 factor. It does, of course get more pixels on target from a greater distance too assuming the two sensors being compared have similar pixel density.

As odd as this sounds, my idea of APS-C has changed over time. Currently, I more think of it as the image that you would get by cropping out the center of a FF sensor. You could do this "in camera" by having a smaller sensor or do it in post-processing if you started with a FF image. But they would yield the exact same image. The resolution of that image is determined by the pixel density. I have always known all of this, but I think I appreciate it more.

So, I really do not think of my M6II as a 32.5MP sensor, which it is. But, especially when comparing it to FF, more often think of it as an 82.5 MP FF sensor that is physically cropped. The benefit of the physical crop being smaller files, less expensive sensor.

Edit---yep, good article by Ari. It is a bit odd that I've always known this, but I do think I am appreciating the simplicity more. Where it gets a bit complicated is considering things like bokeh. For example, if I take the same image of a bird with my M6II and 5DIV, resize the bird to be the same size in both images in post process (mechanically cropped vs post-processing cropped), each image would have the same bokeh.

[DavidEccleston]

I was surprised how close the 5DIV and M6II 10ft shots were. The 5DIV looked to be a bit better resolved, using the 600 marker on the right as my guide. In the 5DIV image I could see the white between the lines continuing, while on the M6 that was the point where it became more of a general grey tone.

So, imagine my surprise when you followed up with the M6II full size, and you can see detail all the way to the 800 mark.

My takeaways are:

1) modern crop sensors are quite an improvement from even 5 years ago.
2) Current resizing algorithms can hide detail that should be visible. I tried every resize option in PS CC. It appears you used Bicubic. While I could get slightly clearer text, the lines blurred to grey at pretty much the same point in all algorithms. I then did a content aware scale. This was, uh, unique. Text was quite readable. Lines remained visible all the way in. It was not, in any way, a usable result though.

[Kayaker72]

I was surprised how close the 5DIV and M6II 10ft shots were. The 5DIV looked to be a bit better resolved, using the 600 marker on the right as my guide. In the 5DIV image I could see the white between the lines continuing, while on the M6 that was the point where it became more of a general grey tone.

So, imagine my surprise when you followed up with the M6II full size, and you can see detail all the way to the 800 mark.

My takeaways are:

1) modern crop sensors are quite an improvement from even 5 years ago.
2) Current resizing algorithms can hide detail that should be visible. I tried every resize option in PS CC. It appears you used Bicubic. While I could get slightly clearer text, the lines blurred to grey at pretty much the same point in all algorithms. I then did a content aware scale. This was, uh, unique. Text was quite readable. Lines remained visible all the way in. It was not, in any way, a usable result though.

Ha...you took the resizing to a whole other level. I just used LR and set it to resize to the dimensions of the 5DIV image. No clue what algorithm it used.