gdanmitchell
Offline
Upload & Sell: Off
|
 Re: Fuji's marketing concept? | |
zhangyue wrote:
gdanmitchell wrote:
ruthenium wrote:
Steve Spencer wrote:
SGinNorcal wrote:
gdanmitchell wrote:
EB-1 wrote:
gdanmitchell wrote:
EB-1 wrote:
What about the 40x53mm MF Phase One sensors like with Hasselbald? I suppose we need a format name for the 33x44mm frame but it is not the largest. EBH
That one is pretty easy. It is nominally the same as the 41.5mm x 56mm area of the 645 film MF, which is probably why Phase One chose to make it that size. If a true digital equivalent to film medium format exists, it would be about that size. (Note that its short dimension is close to the long dimension of the 44x44 format.)
(Film medium formats were defined by their use of film with a 6cm width, which allowed an image area of up to about 56mm in the dimension using the width of the film.)
Exactly. Some gates were 54 or 55 mm and even so they didn't go through all that straight. 
The main reason I avoided the "miniMF" was that it is isn't enough more than single frames from 24x36 to replace the panning/stitching.
EBH
Another frame of reference is to note that the difference between FF and minMF is about half the size of the difference between APS-C and FF.
But that's only true if you view it as a percentage increase like you must be in order to make this claim. If you look at sensor area, its not true at all.
It isn't even true as a percentage increase, but that doesn't stop Dan from saying it over and over again. Fuji APS-C has an area of 367 square mm. Full frame 35mm has an area of 864 square mm. FF 35mm is 2.35 times larger in area. Half of that would be 1.17 times as large. The 33 X 44 sensor has an area of 1452 square mm, which is 1.68 times as large as FF 35mm, so much larger than 1.17 times half the the size difference between APS-C and FF 35mm.
Another way of looking at the differences between the three sensors is to compare the largest amounts of light they can collect (at base ISO, assuming no bias in how the different system define ISO).
The GFX100 sensor collects 1 stop of light more than a FF sensor, while a FF sensor collects 1.5 stops more light than the APS-C.
The “larger sensor collects more light” business always intrigues me.
On one level, it is true. More photons fall on a larger area.
On the other hand, consider an analogy. Let’s say my garden occupies a half acre and my next door neighbor’s garden occupies a full acre. It is true that twice as much rain falls on my neighbors garden, but if my neighbor gets one inch of rain, so do I. A plant in my garden that needs and inch or rain does just as well as a plant in his that needs an inch or rain, and plants that require two incenses are rain fail in both.
Unless…
… each plant collects water from an area twice as large in his garden.
So the potential benefit of the larger garden accrues in two says:
1. You can fit twice as many plants requiring an inch of water into the larger garden, OR
2. You can can reduce the number of plants in the larger garden and grow plants that require more water.
The analogy here, as far as I can tell, is that a larger sensor can:
1. Allow more photo sites and those record more detail, OR…
2. Allow images to be made in somewhat lower light.
(Or some balance of the two.)
So, the business about “more light on the larger sensor” providing better low light performance would not see to work if we compare sensors with the same pixel pitch. (Which I think is the case when we compare the high resolution Sony sensors of different sizes used in APS-C, full-frame, and mniMF cameras these days.) Each “plant” (e.g. photo sites) gets the same number of photons in all three cases.
There are a couple of “complexifying” issues:
1. The “noise grain” will be finer with the larger sensor, given that each photo site is a smaller percentage of the frame size. So if the size of the grain is an issue, the larger sensor with the same pixel pitch will produce a slightly “smoother” (subjectively) image.
2. There are obviously different DOF and aperture choice (relative to diffraction) imperatives with different size sensors.
There are theory beneath it.
Let's say we have two systems all take "exact image" using "proper focal length". 40X30, 20X15.
Case A: If they all share the same resolution with sensor area difference: means larger sensor has "larger" pixel that can collect more light (in this case 4X) "for the same scene" before photon saturation, you could exposure "more" to improve Signal level to improve signal to noise ratio. (noise is square root of 4, and signal is 4, SNR improvement will be 2X)
Case B: If they all share the same pixel pitch, larger sensor has more resolution. Obviously, a advantage for print but this is separate topic. let's focus on SNR signal to noise ratio here. Per pixel SNR will be same as your post wrote as well as a few others. However, we need bring target display resolution or target print size to this discussion. As ultimately, for artwork to show, it has to be displayed in its target form, either screen or print. Say: 20X15, a 10M resolution camera, 40X30 will be 40M. Once down size to the same target resolution, there is a concept call over sampling. Basically signal is sampled 4 times, but noise is random so it will be again squre root of 4. so the SNR is again 2X.
Talk about per pixel SNR is OK but it is not very helpful in term of sensor size/area technical discussion here. To think this in extreme term, considering you have a single pixel camera, you can't say you have the same performance compare to 100M medium format sensor just because per pixel performance is the same.
That’s why I wrote this in that post:
gdanmitchell wrote:
The analogy here, as far as I can tell, is that a larger sensor can:
1. Allow more photo sites and those record more detail, OR…
2. Allow images to be made in somewhat lower light.
(Or some balance of the two.)
- - -
ruthenium wrote:
Dan, by the way, thinking along the lines "Let’s say my garden occupies a half acre and my next door neighbor’s garden occupies a full acre. It is true that twice as much rain falls on my neighbors garden" - this is not how things work in photography. The nearest analogy is not rain, but watering a garden through a hose (the hose being the lens in front of a sensor). The amount of water that falls on the ground is decided by the diameter of the hose and the time between opening and closing of the water tap. The size of the garden is irrelevant.
Let’s try a thought experiment: Think of two cameras whose sensors have the same pixel pitch. I believe that this is the case with the Sony sensor used in Sony’s 60MP FF cameras and in Fujifilm’s 100MP miniMF cameras.
There’s no question that a greater quantity of light falls on the larger area of the miniMF sensor than on a FF sensor. But the intensity of the light is the same. In either case, a photosite getting light from the same point in the subject image will record the same number of photons, and each will fill (be saturated) at the same point. So the “more light” business doesn’t accomplish anything other than greater resolution and smaller noise grain.
Now if you compare (theoretical*) sensors in the two formats that have the same number of megapixels, the larger sensor’s individual photosites will be exposed to more photons due to their larger area. In low light situations (such as the darkest tones in a very wide DR subject) this can extend the dynamic range of the sensor with the larger area.
* I don’t think there are currently any otherwise identical sensors in the two formats that both use the same pixel dimensions and MP.
Another thought experiment: If you take a wide dynamic range photograph from the 100MP miniMF sensor and crop it to 60MP FF dimensions, do you lose dynamic range? How so?
|
