mu43 vs APS-C aperture

[DanS]

There are also two ways to look at light gathering (per area which determines noise/total light and per unit (ie mm²) which determines exposure), but people tend to get quite upset when this is brought up so I will say no more.

To be blunt, I think that's mainly because way to many people believe the sudo-science espoused on the internet, or have been taught by "educators" who still act like we are talking about film.

to be fair the math and physics need to explain it in detail is beyond most people abilities. Not to mention various sensor technologies and generations further muddy the waters.

 

[EarthQuake]

Earthquake..
"In no uncertain terms, the difference in DOF for the same AOV and same distance to subject between M43 and APS-C is about 2/3 stops, it's ~0.6 vs Canon and ~0.7 vs Nikon/Sony/everyone else who use slightly larger APS-C sensors than Canon."

Can you clarify how can you have the same Angle Of View shot at the same distance to subject with M43 and APS-C when the crop factor is different?

Certainly, by using lenses of different focal lengths for the differenct sized sensors, so that the AOV is consistent (again, AOV is a ratio of FL to sensor size) as should be clear from the examples following the quoted statement.

And how do you get 2/3 stops difference in DOF when all lenses have the same DOF?

I don't quite understand this question, can you rephrase it?

 

[bye]

Certainly, by using lenses of different focal lengths for the difference sized sensors, so that the AOV is consistent (again, AOV is a ratio of FL to sensor size) as should be clear from the examples following the quoted statement.



I don't quite understand this question, can you rephrase it?

All lenses can focus perfectly only on 1 focal plane. 1 area of the image plane is technically sharp, which is 1% and 99% is blur of acceptable sharpness which is what DOF is. All lenses can focus sharp only on 1 focal plane.

 

[Amin]

There are also two ways to look at light gathering (per area which determines noise/total light and per unit (ie mm²) which determines exposure), but people tend to get quite upset when this is brought up so I will say no more.

From a standpoint of avoiding moderator time being wasted, I have come close to installing software that automatically closes threads when the phrase "total light" finds its way into a post. I understand the concept perfectly, but it's hard for a lot of folks to get past the fact that some folks on some sites only participate on certain subforums to patronize and pontificate on this topic.

 

[Drdave944]

I am really getting dizzy. in practice,here is the difference. When I got my first M-43 camera the first thing I noticed was it was easier to take sharp pictures. This is because you have a 50% smaller frame. This inherently will give you a larger depth of field and less problem with camera shake. If you use an even smaller sensor and lens ,like a point and shoot camera your sharpness and DOF are even better,but your image size and amount of information are impoverished. M-43 holds a middle ground which is rich in possibilities.
Now some people are worshipers of narrow DOF and bokeh. I am not. If i want bokeh,the best way for me to get it is to use full frame cameras with huge apertures. You can get truly freakish bokeh with an 85mm f1.2 Canon lens,but it has a lot of CA. I don't have one,since this not a good general photography lens and is really heavy.
My bokeh kings are a fast telephotos on a full frame. But I really don't care about bokeh,for the most part. I no longer bother with APS-C, a compromise having big heavy lenses ,even if the camera is getting smaller. Except that Canon is now coming out with the M-6 with M series lenses that may a comer. An APC-C sensor with reasonable size lenses.

 

[bye]

Sometimes the easiest way to learn DOF is just to go out and shoot more photos and experiment.

 

[EarthQuake]

What I meant that all lenses have the same DOF is because, all lenses can focus only on 1 focal plane. 1 area of the image plane is technically sharp, which is 1% and 99% is blur of acceptable sharpness which is what DOF is, so how can you have 1 lens of 1 AOV is 2/3 less of 1% than another lens of a different AOV?

Still not entirely following you, as I specifically stated both lenses have the same AOV in my comparisons. But I'll try to give you a response:

For lenses with the same AOV, but different sensor sizes, the focal length of the lens must by necessity different. When you change the FL, but keep the f-ratio constant, you're changing the physical aperture opening of the lens which results in a change in DOF. One could argue that the physical opening is all that really matters, and the sensor size is meaningless, but you need to know sensor size to know FL for a given AOV, and you need FL and f-ratio to calculate physical opening, so you can't really remove the various parameters and look at them in isolation.

For instance, the 42/1.4 that I spoke of earlier has a 25/1.4 = ~30mm physical opening, and a 56/1.8 lens = ~31mm physical opening, thus, the resulting DOF is roughly the same.

Now, if we look at the same f-ratio, our 42/1.4 still has the same 30mm opening, but a 56/1.4 has a 40mm opening, which means it's letting in more light and will give us slightly narrower DOF.

@tkbslc explained essentially the same concept earlier in this thread.

 

[EarthQuake]

From a standpoint of avoiding moderator time being wasted, I have come close to installing software that automatically closes threads when the phrase "total light" finds its way into a post. I understand the concept perfectly, but it's hard for a lot of folks to get past the fact that some folks on some sites only participate on certain subforums to patronize and pontificate on this topic.

Yes, it's rather unfortunate, the topic is often automatically viewed as some sort of attack or inflammatory statement. Personally, I simply find it interesting to talk about. But I understand your position and I don't envy your role as a moderator when these discussions pop up. So don't worry, I don't have any intention on dragging this point out.

 

[Amin]

Yes, it's rather unfortunate, the topic is often automatically viewed as some sort of attack or inflammatory statement.

I agree. But it's also unfortunate that the topic has been used so often for that purpose.

 

[Amin]

I understand the difference in crop factor (2x vs 1.5x) but did not realize there was also a difference in DOF at the same aperture. This leads me to ask...

Is this purely a DOF phenomenon, or is the light gather of a mu43 at 2.8 similar to a APS-C at 1.8 and therefore as "Fast"??

First, you have to clarify whether by "aperture" you are talking about the physical aperture of the lens or the relative aperture (f-number). From your question, it sounds like you are talking about the latter.

It also depends on which specific APS-C format you choose. In this case you specified 1.5X crop.

So let's say we have the a MFT camera and a 1.5X crop APS-C camera both taking a photo of a subject 10 feet away and that the MFT camera is using a 25mm lens set to f/2.4.

In order to get the same diagonal angle of view, the APS-C camera will need to use a 33mm lens. To match depth of field for a given diagonal print (or display) dimension, the APS-C lens will need to be stopped down to f/3.2 (roughly).

Now if you are calculating exposure, those two f-numbers are what they say they are, f/2.4 and f/3.2. Therefore for a matched resulting brightness of final image displayed, assuming that both images use the same shutter speed, then the APS-C camera would need a higher ISO value than the MFT camera. For example if the MFT camera were to use ISO 200, the APS-C camera would need ISO 339 (roughly).

None of that is controversial. The part that gets people riled up is whether ISO 339 on the 1.5X APS-C sensor is just as good as ISO 200 on the MFT sensor. And the answer is, it depends.

In general, under conditions of matched exposure (shutter speed, f-number, and ambient light at the scene) and image output size, larger sensors deliver greater signal to noise than smaller sensors for a given ISO. And in theory, if they two cameras have equally efficient sensors, ISO 339 on the 1.5X APS-C camera will be just as good as ISO 200 on the MFT camera under the proposed conditions (that's where the part about the matched physical apertures delivering the same total amount of light to the total sensors comes into play).

But in reality, there are a lot of different sensor efficiencies out there, so trying to work out exactly what ISOs would match signal/noise based on format is not an exact calculation in the way that working out focal lengths for angle of view or f-numbers for depth of field are.

 

[bye]

Still not entirely following you, as I specifically stated both lenses have the same AOV in my comparisons. But I'll try to give you a response:

For lenses with the same AOV, but different sensor sizes, the focal length of the lens must by necessity different. When you change the FL, but keep the f-ratio constant, you're changing the physical aperture opening of the lens which results in a change in DOF. One could argue that the physical opening is all that really matters, and the sensor size is meaningless, but you need to know sensor size to know FL for a given AOV, and you need FL and f-ratio to calculate physical opening, so you can't really remove the various parameters and look at them in isolation.

For instance, the 42/1.4 that I spoke of earlier has a 25/1.4 = ~30mm physical opening, and a 56/1.8 lens = ~31mm physical opening, thus, the resulting DOF is roughly the same.

Now, if we look at the same f-ratio, our 42/1.4 still has the same 30mm opening, but a 56/1.4 has a 40mm opening, which means it's letting in more light and will give us slightly narrower DOF.

@tkbslc explained essentially the same concept earlier in this thread.

Yes, I understand the concept of AOV. As you know, photography is 2 dimensional, so only 1 plane is sharp and therefore any other planes of focus in front or behind this 1 focus plane will never be as sharp as this 1 focus plane. So I was confused why you say about light theories and about stops of DOF, because it seems you are trying to explain this in a 3 dimensional plane when DOF is only 2 dimensional. Hope this is not so?

 

[EarthQuake]

Yes, I understand the concept of AOV. As you know, photography is 2 dimensional, so only 1 plane is sharp and therefore any other planes of focus in front or behind this 1 focus plane will never be as sharp as this 1 focus plane. So I was confused why you say about light theories and about stops of DOF, because it seems you are trying to explain this in a 3 dimensional plane when DOF is only 2 dimensional. Hope this is not so?

I think we're talking about two different concepts interchangeably here.

Firstly, we have the plane of focus, you're correct in that is extremely thin, basically only a two dimensional plane that is technically in focus.

The second concept is depth of field. DOF defines the range in front and behind the focal plane that is perceived as "acceptably sharp" enough to be considered as in focus for practical purposes. This is most certainly a three dimensional space.

How this 3D plane of acceptably sharp focus is determined is by measuring the physical size circle of confusion, or the size of the "bokeh ball", if it's under a certain size (this is not a fixed variable, it depends on output size and viewing distance, though most DOF calculators don't let you adjust these parameters) that means that it should be perceptively similarly sharp as the sharpest plane of focus, and it's deemed within the three dimension range of acceptable sharpness/focus.

Essentially this:

Subscribe to see EXIF info for this image (if available)

The reason to use f-stops in this context when comparing is that it's relatable. Generally, most people have an idea of what the difference in DOF/background blur looks like when they stop down a lens 1 stop, 2 stops, etc, so it's an easy way to phrase it. The conclusion could be stated as "The APS-C system has X% larger circles of confusion" and would be just as accurate but probably harder to visualize.

You probably understand most of this so forgive me if this comes off as condescending, but we've gotten quite detailed so I just want to make sure I am being thorough.

 

[tkbslc]

Yes.. For convenience, because of the crop factor. People can associate focal length because it's printed on the lens. Angle of view is not printed on the lens. But focal length is NOT what determines DOF. Angle of View is. 2 different things. 2 lenses made by 2 different makers that have the same focal lengths can have differing angle of view. This is illustrated between the Panasonic 25mm f/1.4 and the Olympus 25mm f/1.8 where one is actually slighter wider in angle of view as opposed to the other. Another example also would be the Nikkor 70-200 VR2 and the Nikkor 70-300 VR where at 200mm, one is also slightly wider than the other as well.

While true, the AOV differences between two lenses of the same specified focal length will be a rounding error in the resulting DOF. It's just like there's never been a perfect f1.80 lens. But photography isn't that precise, so it doesn't really matter.

 

[bye]

I think we're talking about two different concepts interchangeably here.

Firstly, we have the plane of focus, you're correct in that is extremely thin, basically only a two dimensional plane that is technically in focus.

The second concept is depth of field. DOF defines the range in front and behind the focal plane that is perceived as "acceptably sharp" enough to be considered as in focus for practical purposes. This is most certainly a three dimensional space.

How this 3D plane of acceptably sharp focus is determined is by measuring the physical size circle of confusion, or the size of the "bokeh ball", if it's under a certain size (this is not a fixed variable, it depends on output size and viewing distance, though most DOF calculators don't let you adjust these parameters) that means that it should be perceptively similarly sharp as the sharpest plane of focus, and it's deemed within the three dimension range of acceptable sharpness/focus.

Essentially this:
View attachment 511474

The reason to use f-stops in this context when comparing is that it's relatable. Generally, most people have an idea of what the difference in DOF/background blur looks like when they stop down a lens 1 stop, 2 stops, etc, so it's an easy way to phrase it. The conclusion could be stated as "The APS-C system has X% larger circles of confusion" and would be just as accurate but probably harder to visualize.

You probably understand most of this so forgive me if this comes off as condescending, but we've gotten quite detailed so I just want to make sure I am being thorough.

Thank you for clarifying it..

 

[Robert Watcher]

Subscribe to see EXIF info for this image (if available)

2 pages worth. These threads are so funny. All I ever knew when I started taking pictures, was that 80mm was the lens on my camera (medium format). It was all I had and I loved the look that it produced - and have never been able to duplicate the exact feel with any small 35mm film and smaller digital sensors lens combinations. But that has never really mattered to me. I find it more productive to take whatever I have access to and try to get an image that pleases me (and my clients).

Technically I suppose you shouldn't get any DOF from a puny phone camera sensor with a wide angle lens - like my wife's Google Nexus - but she took this shot while out for supper a couple of nights ago, and it looks so pretty and soft in behind. She knows nothing about DOF or even how to use a camera - but she knows that she took a pretty picture. I know about DOF and how to use a camera, and I think she took a pretty picture.

I may be a little simplistic with my methods, but I really see the value of trying to get a photo that satisfies, by trial and error and honing my skills - than spending time worrying about what expensive lens or camera system I probably will never afford or desire to buy. Acceptable DOF can be had with any camera or lens through learned camera technique and control of lighting. But that is just my old fashioned viewpoint I suppose.

 

[Clint]

I understand the difference in crop factor (2x vs 1.5x) but did not realize there was also a difference in DOF at the same aperture. This leads me to ask...

Is this purely a DOF phenomenon, or is the light gather of a mu43 at 2.8 similar to a APS-C at 1.8 and therefore as "Fast"?? I'm guessing not, but was not sure as I'm new to the system and coming from APS-C. I imagine it is just a DOF change, but why?

The supposed equivalence is based on the premise that you want to photograph an area that is identical in size with the same Depth of Field (DoF) with different format cameras, or at least the same DoF to prove the equivalence theory. In this example that is 7’ 2.4” wide and 4’ 9.6” high for a landscape format photo. Under this premise, these are the two constants that must remain the same.

To photograph an area that is identical in size as a 35mm camera with a m4/3s you either need to change your focal length (FL) to half that of the 35mm and keep the same distance to subject, or use the same FL as 35mm and double you distance from the subject. This works to keep this example simple as there are many other alternatives – which often seem not be considered.

And here are the results

Subscribe to see EXIF info for this image (if available)

I stated “supposed equivalence” because

• I’m not thinking with FF I’d shoot with this XXmm lens at f/x.x so with m4/3s I need to …
  
• Perspective that I want is typically my first consideration – something equivalence mostly disregards
  
• I know that with a 2:3 aspect ratio and 4:3 aspect ratio I frame up my subjects very differently making the steps above very hard to do, even if I wanted to
  
• My DoF choice is going to be based on my choice within the limits of my gear, not dictated by some equivalence factor. This is a major factor for typically buying fast lenses as the faster lenses provide more DoF options.
  
• When I need out of focus areas I know how to achieve what I need, and if I wanted the FF wide open look – I just move back to FF or actually to medium format so I’d have even more control.
  
• I shoot wide open a lot, in many situation most won’t – that can cause DoF issues with FF normally resolved by closing down somewhat. So with m4/3s I’m already there!
  
• The m4/3s for gives me a DoF advantage for close in shots as well as at telephoto lengths

 

[SojiOkita]

The gist of my argument is that a 20/1.7 is a 20/1.7 on both FF and m43, regardless of physical crop (m43) or crop in PP (FF). Even if you don't crop the FF picture, the DOF will still be the same

I think we don't have the same definition of depth of field.
Depth of field is what "appear acceptably sharp in an image".
There is only one plane of focus. Then, the perception of what is acceptably sharp will depend on the size of the picture.
That's why it's easier to see if a picture is in focus when zooming at 100% on a computer.

And that's why all depth of field formulas take into account the "circle of confusion", which is taken twice lower on m43 because the image will be magnified twice more.
(nice comparison table in the post just above by Clint, for example)

If taking the same image, with the same camera, and making two prints (of the same size):
- first print of the full image
- second print of a 2x crop section of the center of the image
The perceived depth of field will not be the same on the 2 prints.

Frankly, I don't understand why crop factor is relevant. People have been using medium, large and half format for a long time without worrying about equivalence. Just a few years ago, when APS-C was almost standard among hobby photographers, hardly anyone talked about crop equivalence and even less about DOF equivalence.

Crop factor is only relevant when comparing two different sensor formats.
24x36 is not a reference for me, there are bigger sensors. I hate the "full frame" term and never use it.
However, when using 2 formats, or hesitating between 2 formats, it can be useful to get some comparison point.
When I came from APS, relative crop factor was an important data to choose what lenses I needed to buy (in terms of field of view & depth of field)

 

[SojiOkita]

I will just add that the value taken for the circle of confusion is of course an assumption.
Considering, for example, 0.03 for 24x36 and 0.015 on m43 is based on assessing depth of field on prints.
Some depth of field calculators will take values depending on sensor size and megapixels, and that will be relevant for assessing depth of field at 100% on a computer, but will not be relevant for prints or images watched on screen at the same size.

I found that interesting page that gives an explanation about the circle of confusion value :
Depth of field and diffraction

The standard criterion for choosing C (...) is that on an 8x10 inch print viewed at a distance of 10 inches, the smallest distinguishable feature is (allegedly) 0.01 inch. That was the assumption in the 1930's when film was much softer than it is today. At 8x magnification this corresponds to 0.00125 inches = 0.032 mm on 35mm film, close to the standard 0.03 mm used by 35mm lens manufacturers to calculate their DOF scales. If you've ever had a close look at a fine contact print from 4x5 or 8x10 film, you'll doubt that 0.01 inch feature size is a good criterion. Studies on human visual acuity indicate that the smallest feature an eye with 20:20 vision can distinguish is about one minute of an arc: 0.003 inches at a distance of 10 inches. But inertial prevails: 0.01 inch is universally used to specify DOF.

That's probably why "hyperfocals" values have to be taken with care...

 

[Costas_Gr]

First, you have to clarify whether by "aperture" you are talking about the physical aperture of the lens or the relative aperture (f-number). From your question, it sounds like you are talking about the latter.

It also depends on which specific APS-C format you choose. In this case you specified 1.5X crop.

So let's say we have the a MFT camera and a 1.5X crop APS-C camera both taking a photo of a subject 10 feet away and that the MFT camera is using a 25mm lens set to f/2.4.

In order to get the same diagonal angle of view, the APS-C camera will need to use a 33mm lens. To match depth of field for a given diagonal print (or display) dimension, the APS-C lens will need to be stopped down to f/3.2 (roughly).

Now if you are calculating exposure, those two f-numbers are what they say they are, f/2.4 and f/3.2. Therefore for a matched resulting brightness of final image displayed, assuming that both images use the same shutter speed, then the APS-C camera would need a higher ISO value than the MFT camera. For example if the MFT camera were to use ISO 200, the APS-C camera would need ISO 339 (roughly).

None of that is controversial. The part that gets people riled up is whether ISO 339 on the 1.5X APS-C sensor is just as good as ISO 200 on the MFT sensor. And the answer is, it depends.

In general, under conditions of matched exposure (shutter speed, f-number, and ambient light at the scene) and image output size, larger sensors deliver greater signal to noise than smaller sensors for a given ISO. And in theory, if they two cameras have equally efficient sensors, ISO 339 on the 1.5X APS-C camera will be just as good as ISO 200 on the MFT camera under the proposed conditions (that's where the part about the matched physical apertures delivering the same total amount of light to the total sensors comes into play).

But in reality, there are a lot of different sensor efficiencies out there, so trying to work out exactly what ISOs would match signal/noise based on format is not an exact calculation in the way that working out focal lengths for angle of view or f-numbers for depth of field are.

You was perfectly understandable. You explained with very simple words and examples the question.
At the low light situations there are 2 ways. If you are using FF you play with ISO because you are not afraid from that. At mft you are not afraid from shallow DOF, so you can play with aperture and shooter speed to keep ISO low.
Before one year when I was ready to move from aps-c to foul frame, a professional photographer advised me to try first a cheap Olympus mft and then if I am not happy can move to FF and keep the small mft as second camera. Now, I know what I need. It is an EM1 mk2.

 

[EarthQuake]

The supposed equivalence is based on the premise that you want to photograph an area that is identical in size with the same Depth of Field (DoF) with different format cameras, or at least the same DoF to prove the equivalence theory. In this example that is 7’ 2.4” wide and 4’ 9.6” high for a landscape format photo. Under this premise, these are the two constants that must remain the same.

To photograph an area that is identical in size as a 35mm camera with a m4/3s you either need to change your focal length (FL) to half that of the 35mm and keep the same distance to subject, or use the same FL as 35mm and double you distance from the subject. This works to keep this example simple as there are many other alternatives – which often seem not be considered.

And here are the results
View attachment 511486
I stated “supposed equivalence” because

• I’m not thinking with FF I’d shoot with this XXmm lens at f/x.x so with m4/3s I need to …
  
• Perspective that I want is typically my first consideration – something equivalence mostly disregards
  
• I know that with a 2:3 aspect ratio and 4:3 aspect ratio I frame up my subjects very differently making the steps above very hard to do, even if I wanted to
  
• My DoF choice is going to be based on my choice within the limits of my gear, not dictated by some equivalence factor. This is a major factor for typically buying fast lenses as the faster lenses provide more DoF options.
  
• When I need out of focus areas I know how to achieve what I need, and if I wanted the FF wide open look – I just move back to FF or actually to medium format so I’d have even more control.
  
• I shoot wide open a lot, in many situation most won’t – that can cause DoF issues with FF normally resolved by closing down somewhat. So with m4/3s I’m already there!
  
• The m4/3s for gives me a DoF advantage for close in shots as well as at telephoto lengths

I think the confusion with equivalence comes from people thinking they need to apply it to every day shooting, which of course is not necessary. Equivalence theory is mostly useful when comparing two camera systems with different sensor sizes, generally when trying to figure out which system will meet your needs when it comes time to purchase something. It's also useful when switching from one system to another so that you have a solid reference for which settings do what.

At the end of the day, it's important to know the gear you have and know how to work with it. But still, understanding equivalence can be helpful. Let's say you switch from a FF system with a fast 35/1.4 (as I did when I came to M43 land). Because I understand equivalence, I knew I wouldn't be able to get quite the same look as as with that lens, and to get similar subject isolation I would need to use a longer lens, and take more care to place my subject further from the background. Sure, with enough trial and error I probably would have figured this out on my own, or developed my own unique style with the new camera system, but equivalence makes it easier to understand. Equivalence certainly isn't the end all and be all though, it's simply a useful tool. Nor are bigger sensors necessarily better for everyone. At the end of the day, I think the wider DOF forcing me to think more carefully about my composition can be a good thing as well, even if larger systems offer more flexibility in this regard. Personally I've gotten rid all of my APS-C and FF gear and shoot exclusively with M43 now, when I need extensive DOF control my 42.5/1.2 provides it.

As to FF vs medium format, the idea that MF will provide narrower DOF is somewhat of a myth. With most digital MF cameras (44x33mm sensor, 0.8x crop), the difference is about 2/3rd stop, in other words the same difference between APS-C and M43. The kicker is that MF lenses are rarely faster than F2.8, while there are many 1.2, 1.4 and 1.8 lenses for FF, meaning that in practical use, FF/35mm provides more DOF control than medium format. FF is typically better in low light for the same reason.

There are some digital MF cameras that are about 0.65x (~54x40mm sensor), and that's more like a 1 1/3rd difference, which is still generally negated by the faster 35mm lenses.

Now, if we go all the way to 6x7 film, then we're looking at a 0.5x crop factor which makes those F2.8 MF lenses behave like F1.4 35mm lenses as far as DOF is concerned, and that's quite a significant difference, but even still, there are plenty of F1.4 35mm lenses. To get significantly more DOF control than 35mm + a fast prime provides, one has to go up to large format film. The good news is more DOF control than 35mm provides is not often necessary, so this point is mostly academic and generally moot for practical purposes.

There are of course other advantages to MF, generally dynamic range, color depth, and resolution, though with the 50MP 44x33mm sensors most commonly seen in MF cameras these days, these advantages are relatively slim over sensors like the Sony 36MP, 42MP, and Canon 50MP sensors used in various cameras.