What are you not getting?

When you're at f/11 w/M43, it's the equivalent of f/22 on FF.

IOW, there is no reason, zero, nada, none, to go past f11 since you will get diffraction.

But if you don't go past f/11 on mft then there's NO ADVANTAGE AT ALL regarding large DOF in mft cameras over 35mm. You see, ~tc~ said that there's a huge advantage that mft cameras have over 35mm for shots that require large DOF, but I don't see any. F/22 on a 35mm camera will give you ~ the same DOF as F/11 on mft, and if you still need more DOF one can assume that since m43 can be stopped down even more, past f/11, then these smaller sensor cameras have some advantage.. but when you go past f/11 you get diffraction and the image gets softer, usually very soft at f/22. What's the point in using f/22 on mft if the resolution is basically 1/3 or even less of that compared to f/5.6-f/11? You gain some extra DOF and at the same time get P&S image quality due to diffraction in return.

Heli-mech's argument about handheld shooting makes some sense, but yet again with 35mm sensors you can just easily bump ISO a stop or two and get the shutter speed you need with the similar amount of noise that you'd get from an mft sensor.

So what I'm "not getting" is where is that huge advantage that mft cameras have for large DOF shots.

The "advantage" for m4/3 is in higher shutter speed or lower iso at the same depth of field.

Fred

For any given aperture setting, completely independent of the camera trying to record the data, the diffraction (the size of the Airy Disk where two distinct points of light can be resolved) is always the same.

Your lens can make this worse, but never better that the theoretical limit.

Your camera could have really, really small sensor pixels (<1.5 nm) that are always below the theoretical limit, and diffraction will control the resolving power at all f stops. The other end of the spectrum would be a camera with very large pixels (>30 nm) where the resolving power is always limited by the sensor, and diffraction can't make it any worse because the Airy Disk is smaller than the detector cells no matter the f stop.

This is a better way of saying what I meant.

I still can't see the scenario where f/11 m43 (other than macro) is not sufficient DOF. As mentioned above, so long as the focal point is more than 7' away, you're hyperfocal.

__________________
Comments and criticism welcome! It's how we get better!

White G3:P20/1.7:P45-200:PL45/2.8:P8FE:P12-35/2.8

Pictures here: http://troyandmollyalloverthemap.shutterfly.com/

http://500px.com/TroyGorrell

If you are excluding close-ups, then the following may not be so relevant.

If the resolution, although reduced, is nonetheless sufficient, and the gain in dof is of benefit, then it is worth using a very small aperture. As it happens I use f/22 on mft for many of my close-ups, just as I used f/8 on a small sensor bridge camera, to get roughly the same dof. For me, the trade-off between resolution and dof can be, depending on the image, worthwhile up to and including f/22.

__________________
Nick

http://www.flickr.com/photos/gardenersassistant/

Here's where it can get a little more muddy. Some of the newer FF cameras (D800) have a pixel pitch fairly close almost the same as a 12MP 4/3 and just a little bigger than the E-M5. You can run various scenarios through the calculator and the bottom of this page: Digital Camera Diffraction – Resolution, Color & Micro-Contrast

Therefore even though noise wise you could bump up ISO to gain back the extra couple stops for DOF you would only have a approx 1 stop advantage in diffraction limits. (less than 1 comparing to a 12MP 4/3, more than 1 if comparing a lower resolution FF).

Now what I couldn't say is whether or not a D800 at F22 ISO800 would still out resolve a E-M5 at F11 ISO200. The D800 would be about a stop farther into diffraction and two stops higher ISO. Taking a quick look at Imaging-resource image compare tool D800 ISO800 vs E-M5 ISO 200 both at F8 I would say the E-M5 has more detail, but those are Jpegs so RAW may be different.

I think at the very least the m4/3 does not loose out to FF when it comes to Deep DOF and potentially does have a advantage of about a stop. However we have to remember that m4/3 has many advantages in terms of cost and size.

__________________

I was going to debate all this with you, but then I realized you were speaking in regards to what ~tc~ said earlier (quoted below), and after reading that over I agree that ~tc~ is wrong.

The big advantage 4/3rds has is when you want greater DOF at wider apertures so you can use lower ISO or faster shutter speeds. At narrow apertures diffraction kicks in at about the same relative point and so I wouldn't say that it is a huge advantage for large DOF shots. It is at best a limited advantage for certain special circumstances where you need large DOF, mostly involving macro work, and situations where critical sharpness is not a requirement.

This I do not agree with. It wouldn't be a stop or two, it would be a 2 stop ISO bump to get the same DOF. And the noise difference would depend heavily on the cameras involved and the starting ISO level. The noise and general high-ISO performance of the E-M5 is better than APS-C cameras from just a couple years ago, and is not even that far behind some Full Frame 35mm cameras from just a couple years ago. If you did a fair comparison (same total budget for lens and camera) then there wouldn't even be close to any 35mm option that would let you increase ISO 2 stops because it would only be old 35mm cameras in the running. Sure, a 5d3+24 TS-E would beat E-M5+12mm, but it would cost 3 times as much.

Also, one thing people don't seem to understand is that High ISO performance is not just about measurable/visible noise performance itself. In reality the main reason most Mid to High ISO shots used to look so bad was not so much the noise itself, but rather the loss in color accuracy (washed out looking), loss in tonal range, and loss in dynamic range that resulted from using higher ISOs. The most recent generations of sensors have massively improved in these areas as well as in noise itself, and this is causing high ISO images to look better than ever before (if you don't pixel peep). For instance even though a camera like the 5d3 only has a measured improvement in high-ISO noise performance of about 2/3rds stop over the 5d2, its real-life useability is 2 full stops better in many peoples opinion. If you pixel-peep it looks nearly as bad, but the image as a whole looks WAY better than it would at high ISO on the 5d2. The E-M5 is similar. it's noise performance is not what is making it so much more useable at higher ISO's, it is that the overall images are not degrading as quickly as ISO increases. So when you compare it to a camera like the 5d2 it actually compares quite favorably at higher ISO's, lagging behind less than a stop in my opinion. (For my bonafides... I own an E-M5 and a 5d3, and I owned a 5d2 for a couple years as well.)

I need to go to bed, but if I get some time tomorrow I will post a link to some sample images that show what I am trying to say here regarding slower decrease in image quality as ISO increases with newer sensors.

So, you say I am wrong, but then agree with me? LOL

I will take the time to run the numbers using a DOF ipad app ...

D3, D3X, D700 with 50mm f/22 focused at 6' -> 7' 6.7" DOF
G1, GF1, etc with 25mm f/11 focused at 6' -> 7' 8.8" DOF

Let's call it close enough in DOF, then exposure would be (using the light in my office as an example):
FF: ISO 160 f/22 1"
M43: ISO 160 f/11 0.25"
significant hand holding advantage

or, if you wanted to adjust ISO
M43: ISO 160 f/11 0.25"
FF: ISO 640 F/22 0.25"
significant ISO advantage

As mentioned above, this difference can easily cancel out the sensor advantage, and more than make up for the smaller gap between the newer sensors in the G3 and OMD. Since the real difference in todays technology is 1-1.5 stops, if you DID run up against diffraction limiting apertures, you could bump the ISO on the m43 a little and get your extra DOF with equivalent High-ISO "degradation" and avoid the diffraction issue.

__________________
Comments and criticism welcome! It's how we get better!

White G3:P20/1.7:P45-200:PL45/2.8:P8FE:P12-35/2.8

Pictures here: http://troyandmollyalloverthemap.shutterfly.com/

http://500px.com/TroyGorrell

All these numbers are irrelevant if you use a tripod for landscapes or a flash for macro. These numbers are also irrelevant in good natural lighting because then the first result would be more like:

FF: ISO 160 f/22 1/100
M43: ISO 160 f/11 1/400

Both of which would be entirely acceptable for hand-held landscape or macro use, thus giving no real advantage to 4/3rds. There would also be no need for the second numbers since ISO would not need to be adjusted.

This is a large part of why the advantage is not huge at large DOF, the advantage just doesn't apply a lot of the time. It is a definite advantage when it is needed, but it isn't a HUGE advantage even then, and it also is not actually needed most of the time for large DOF shooting. It is actually more useful at narrow to mid DOF shooting, which is where you are more often playing the balancing game to get the DOF you want without too high of ISO or too slow of shutter speed.

EXACTLY. If all the DOF advantage is doing is canceling the other sensors ISO advantage then how is it a HUGE advantage over the other format at all? Especially considering it only applies when you WANT a faster shutter speed, which you often do not (at least in landscape work). Even assuming the ISO difference thanks to the awesome E-M5 is only 1 stop, a 1 stop overall advantage is not really HUGE. Especially when it isn't always useable a/o needed.

This makes no sense. If you hit diffraction limiting apertures you would only be able to get more DOF by using a narrower aperture, which would cause more diffraction. This is a given, and no amount of ISO bumping would help to avoid it in any way.