I am very confused by this statement post in the fourthird section by photozone. "Please note that four-thirds cameras have a different diffraction characteristic than APS-C or full format cameras. The real-world sweet spot, the point of highest resolution, is generally around f/4. Therefore it's hardly possible to improve the performance by stopping down when using (such) slow speed lenses. Remember that f/8 on a four-thirds camera is comparable to f/16 on a full format camera." (for depth of field, but also for sweet spot too?)
I thought sweet spots varied from lenses and I don't understand how a sweet spot relates to the image plane size. I know MTF resolution corresponds to printing size and smaller sensors have less information to blow up and print with. Maybe I'm on the wrong path here, don't have the foggiest, this statement just is so open ended and vague to me. I found it on the Olympus 14=42 review page for mft and it seems like a lot of info is missing, so many variables come in play when finding the lenses sweet spot, I'd hate for a image sensor to limit it too. I'd like a link or a straight of techy answer to this, really nagging me when I go on forums and ppl intend to bash the fourthird system for misconception, and I fail to correct them when I don't understand well enough myself.
Posts: 8,000
Threads: 1,850
Joined: Apr 2010
Reputation:
49
[quote name='Plochmann' timestamp='1324794477' post='14086']
I am very confused by this statement post in the fourthird section by photozone. "Please note that four-thirds cameras have a different diffraction characteristic than APS-C or full format cameras. The real-world sweet spot, the point of highest resolution, is generally around f/4. Therefore it's hardly possible to improve the performance by stopping down when using (such) slow speed lenses. Remember that f/8 on a four-thirds camera is comparable to f/16 on a full format camera." (for depth of field, but also for sweet spot too?)
[/quote]
The sentence says "generally". Your are, of course, right that there is also a lens specific aspect in there. However, when looking at the MTFs of the slower lenses you may notice that the center performance is often best at max. aperture. The border performance and contrast may still improve a little when stopping down a tiny bit.
As far as the depth-of-field aspect is concerned - f/4 on MFT is f/8 on a full format camera so you may have to ask yourself whether you really need to stop down much further than this anyway.
Posts: 6,715
Threads: 236
Joined: Apr 2010
Reputation:
22
12-25-2011, 09:50 AM
(This post was last modified: 12-25-2011, 11:58 AM by Brightcolours.)
[quote name='Plochmann' timestamp='1324794477' post='14086']
I am very confused by this statement post in the fourthird section by photozone. "Please note that four-thirds cameras have a different diffraction characteristic than APS-C or full format cameras. The real-world sweet spot, the point of highest resolution, is generally around f/4. Therefore it's hardly possible to improve the performance by stopping down when using (such) slow speed lenses. Remember that f/8 on a four-thirds camera is comparable to f/16 on a full format camera." (for depth of field, but also for sweet spot too?)
I thought sweet spots varied from lenses and I don't understand how a sweet spot relates to the image plane size. I know MTF resolution corresponds to printing size and smaller sensors have less information to blow up and print with. Maybe I'm on the wrong path here, don't have the foggiest, this statement just is so open ended and vague to me. I found it on the Olympus 14=42 review page for mft and it seems like a lot of info is missing, so many variables come in play when finding the lenses sweet spot, I'd hate for a image sensor to limit it too. I'd like a link or a straight of techy answer to this, really nagging me when I go on forums and ppl intend to bash the fourthird system for misconception, and I fail to correct them when I don't understand well enough myself.
[/quote]
There are two ways the smaller sensor influences what the f-value actually does.
First thing is softening due to diffraction. Diffraction of light happens around the edges of the aperture. Diffraction actually happens at any object at any edge, it is the reason shadow edges get less and less sharp with distance no matter how small the light source is, for instance. The smaller the aperture, the more light gets diffracted and ends up on neighbouring pixels. softening the image.
Lets say we have a 12mp full frame sensor (like a Nikon D700 or something) and a 12mp 4/3rds sensor (like an Olympus E-30 for instance). If we put a lens in front of the FF sensor, the diffracted light might reach 2 neighbouring pixels in sufficient quantity to soften the image noticeably. Now if we put the same lens in front of the 4/3rds sensor, due to its smaller pixel pitch, the same diffracted light will now reach 4 neighbouring pixels in sufficient quantity to soften the image noticeably.
Also on image level (not just on pixel level, as described above) diffraction has a bigger impact on the sharpness of the image when the sensor is smaller. With the same print size you "magnify" the diffraction caused softness from the smaller sensor image more.
The 2nd impact the smaller sensor has is with DOF. Equivalent focal lengths on FF and 4/3rds need equivalent f-values too, to get a similar DOF. So, if you use a 100mm lens on FF and 50mm lens on 4/3rds (equivalent focal length), you need to use equivalent f-vales too.
If for instance you use f8 on FF, you need to use:
100 / 8 = 12.5mm aperture
50 / 12.5 = f4 as equivalent aperture.
So, sensor size has two ways to impact the captured image, and it is good to understand them. It helps you understand how to get similar results when you use different formats. Whether you use 4/3rds, APS-C, 135 FF format, APS-H. medium format, Nikon 1, compact digital or whatever.
Notice that the two go hand in hand: When you choose a similar DOF with an equivalent lens, you get a similar diffraction softening amount on the image too. So, no need to feel "bashed", when you make a similar photo with 4/3rds and with FF (similar field of view + similar DOF) you will get similar diffraction softening in the image. Just the FF camera can go more extreme (bigger apertures). And has bigger photo sensitive diodes so can collect more light information (count more photons per sensel).
In practice the difference between 4/3 (2x crop) and APS-C sensors (say 1.6x crop) of identical resolution (pixel count) is small. The aperture thresholds at which diffraction starts to limit sharpness are probably less than half a stop apart. The other way around diffraction starts to limit sharpness on a NEX-7 (APS-C) and a 16 MP Panasonic m43 cam at virtually identical aperture settings. Of note, the effect at the aperture threshold is for pixel peepers only - 100% on screen viewing is needed.
Posts: 6,715
Threads: 236
Joined: Apr 2010
Reputation:
22
[quote name='Sammy' timestamp='1324844889' post='14113']
In practice the difference between 4/3 (2x crop) and APS-C sensors (say 1.6x crop) of identical resolution (pixel count) is small. The aperture thresholds at which diffraction starts to limit sharpness are probably less than half a stop apart. The other way around diffraction starts to limit sharpness on a NEX-7 (APS-C) and a 16 MP Panasonic m43 cam at virtually identical aperture settings. Of note, the effect at the aperture threshold is for pixel peepers only - 100% on screen viewing is needed.
[/quote]
On image level, the diffraction softening effects are 2/3rds of a stop "apart" between 4/3rds and APS-C (advantage for APS-C).
Between FF and APS-C 1 1/3rd stop (FF advantage) and between FF and 4/3rds 2 stops (FF advantage).
However, when you look at what matters (aperture size, not f-value) there are no such differences anymore. Just use equivalent apertures and the outcomes will be more or less the same, both for depth of focus and for diffraction related softening.
When you look at pixel level (not very interesting, looking at pixel level...) things are very different. Then it matters what the pixel pitch is, not the sensor size.
Really great replies guys. I understand completely now; this is a great forum- none other like it I have found. If brains could be hungry, mine is now full. I do motion picture work mostly, and I feel more and more that 4/3 is similar to super 16mm compared to super 35mm. It's good to know what are and aren't limits, there is a lot of hearsay all over the net and most gets misconstrued.
