The Reality of the "Crop Factor"

[HDNitehawk]

Pat, I agree with you on the 7D. I did the same type of comparisons with the same results against the 1D IV. The 1D IV could crop more and had a more pleasing image.

From what I am seeing now the 7D II has just as good of an image in respect to color and pixels when compared to the 1D IV. There has been improvement.

Of course you can not crop it as much as a FF body, but your crop is limited by resolution. Looking back at Joel's numbers the 7D II sensor is 336mm. Maybe my math and formula isn't right but the jest of it will be, if the 7D II has 20% resolution wouldn't an equal resolution FF image be 336 sq mm x1.2= 403.2 sq mm. With the FF having 854 sq mm you would be cropping 450 sq mm away.

Joel, the illusion of more reach is that your picture is already cropped by the camera. But you and I both know that if you are shooting wild birds we are cropping for framing whether it be full frame or a 7D II. Then the question is how much can we crop. The Blue Jay image is typical, I was at a distance that it was hard to get good eye detail with the 1D IV. What I look for in eye detail are the little lines around the eye ball. The 7D II could still resolve it. So for extra reach if my maximum is 20' before distance takes away resolution I can still go out another 4' and get the detail at 24'.

This resolution benefit is only realized if you are cropping, and only if you are cropping the FF close to or less than the Crop Body's sensor size.. What I see is the resolution fall off from going to a properly framed FF pictures is greater than the gain of crops resolution.

So shooting small subjects like wild birds that you almost always have to crop it could be beneficial.
It is only beneficial if you have good lighting, dawn and dusk FF bodies are going to rule.
For me the 7D II would be a good accessory to go with the 500mm to give me a x1.2 advantage when shooting small birds during mid day.
The AF system performance of the 7D II has to improve or all of this is null and void.

[Busted Knuckles]

Ok, so now I am completely lost. So Canon is coming out w/ a high(er) megapixel camera. I had read somewhere that the current lens inventory can generally run up to 50ish megapixel before running out of resolving power to the photosite.

Does this have bearing or did you guys just explode my grey matter for kicks and giggles.

[HDNitehawk]

Of course it has bearing. A FF 50 mp camera will have about the same resolving power as the 7D II.

The 7D II advantage isn't the crop it's the pixel density.

If we are getting a 20% gain with the 7D II, you would get the same with a FF. The difference is you get it across the FF pic and the FF benefit. I bet the high ISO performance will be lacking though.

As far as resolving ability, my makeshift tests were using arguably Canons sharpest lens, the 500mm II.
I wonder if 50mp will be well above any lenses ability.

[Dave Throgmartin]

50 MP will require the very best of lenses. It'll be worth it for the super user who has the best of the best lenses, but not so much for the typical user IMHO.

Dave

[neuroanatomist]

First let me say that there is no advantage in having a "Crop Factor".

I must respectfully disagree. The crop factor has significant advantage for many people – lower cost (or in the specific case of the most expensive crop 7DII vs the cheapest FF 6D, far more features and better performance of everything except the image sensor).

[HDNitehawk]

I must respectfully disagree. The crop factor has significant advantage for many people – lower cost (or in the specific case of the most expensive crop 7DII vs the cheapest FF 6D, far more features and better performance of everything except the image sensor).

I will agree with the lower cost with some bodies as an advantage, of course the value of the 7D II compared to the 1Dx makes it look very attractive. Really I wasn't comparing dollar value. If I were I would say the 7D II is the best value in a camera that I have seen from Canon since I went digital.

I will agree that the 7D II has far more features than the 6D, so far I wouldn't say it's "performance of everything except..." has been better than my 5D II. The 7D II has AF issues that in the two months since I recieved it I have been unable to resolve. It may go to Canon shortly for a check up. Some of the things I see it doing I am seeing others making the same complaints on other forums.

[HDNitehawk]

50 MP will require the very best of lenses. It'll be worth it for the super user who has the best of the best lenses, but not so much for the typical user IMHO.

Dave

Whatever a typical user is you are probably right.
The same is true of taking advantage of any resolution advantage the 7D II has. Camera shake, bad lenses, poor lighting and other things take the advantage away.

[Kayaker72]

50 MP will require the very best of lenses. It'll be worth it for the super user who has the best of the best lenses, but not so much for the typical user IMHO.

Dave

True...to get absolute optimum performance or for those cropping heavily. I am not one of those that wants a high MP sensor camera. Mostly, I am would not look forward to the file size and would absolutely be reaching the point of diminishing returns, IMO. But even for most photographers, compressing a 50+ MP image to useable ranges (web pages, 8x10, 16x20 prints, etc) would have benefits such as helping reduce noise, give better resolution, etc. Also, there are a number of affordable lenses that are among the best performers Canon offers especially when stopped down a bit.

I think the popularity of this camera, if it becomes true, will be a function of perceived need, cost, and image file size (of course, assuming features/performance are there). On the perceived need, I ran a poll about a year ago; 17 of 22 of those voted those the ideal number of MPs was in the 20-30 MP range.

Of course, that could all change if Canon blows us out of the water with IQ at a reasonable price.

[HDNitehawk]

On the perceived need, I ran a poll about a year ago; 17 of 22 of those voted those the ideal number of MPs was in the 20-30 MP range.

One of the rumors floating around is that the 50mp sensor will be the same as the 7D II, just larger. I think the key word for your poll was perceived. If the rumor is correct the resolution we are seeing with the 7D II might be close to what we can expect for a FF. A new question would be do I need that additional resolution benefit for the type of shooting I do. As well would I be able to take advantage of the increased resolution with the technique I use to shoot.

I personally can see a benefit for maximum resolution with wildlife, landscape, macro and architectural photography. Beyond that the current full frame offerings have enough resolution for any other types of photography I do.

[Kayaker72]

Rick...your post got me thinking about what the actual benefit would be. First off, IMO, IQ in general will increase with more MP. I think specific types of photographers and photography, as you mention, will benefit more from the extra resolution. If pixels were ideal, it would be a linear relationship. But they are not ideal.

If pixels were ideal, then a crop sensor would truly given 1.6x resolution bump. You came up with 1.2x. 1.2/1.6 = 0.75. If the rumored 51.1 MP sensor camera is truly based on the 7DII sensor, then the 51.1 MP sensor would be somewhat equivalent to 38.3 MP if we applied the 0.75 ratio compared to your 1.0x (5DII). So, a benefit, no doubt.

But, as I mentioned above, the reasons crop resolution isn't ideal are both sensor related and lens related. As the lens related issues would no longer be a factor, 0.75 would likely be the low end.

That made me wonder how much of the divergence from ideal is lens and how much is sensor. Which got me playing with numbers. When people report the "pixel size" the calcs I have seen are simply taking the sensor area and dividing it by the number of pixels. So, for the 5DIII, FF sensor area is 36 mm x 24 mm = 864 mm2 / 23 MP = 6.1.

But, that is another "idealized" calc. It assumes that there is no edge to the pixel. Or, using the pixel=bucket analogy, no edge to the bucket. And a bucket without edges doesn't hold water.

I suspect one of the reasons that higher pixel densities are not as efficient as lessor densities is that with higher densities more of the sensor area is lost to the "bucket edge." This may be around the actual pixel on the sensor or maybe the Bayer filter.

So, the reason for my post, I am wondering if others know, is that correct?