Factors Affecting Depth of Field

Msmoto

From the thread "How to make the Nikon D3200 really perform!" this topic emerged.

Ironheart

Surely if the pixel count is higher, the resolution is better so the DoF is apparently better? :-B

Note the use of the word apparently because none of us is measuring, only looking.

Exactly, if you substitute larger/deeper for better.

So if we all agree that DoF can be affected by pixel pitch and/or the MP count of the sensor, doesn't it also stand to reason that the strength of the AA filter could also have an affect on DoF? Also that the complete lack of one would seem to increase it due to the added sharpness?

I'll ask my other question again, does anyone know the difference between the two 24MP sensors in the D3200 and the D7100? Are they the same manufacturer?

Ironheart

For the background on this topic start here:

http://forum.nikonrumors.com/discussion/1354/how-to-make-the-nikon-d3200-really-perform#Item_33

Paperman

Going the other way and giving extreme examples, then we should be able to say our 6Mp old DSLRs offered shallower depth of field.

One can then also say crappier cheap lenses with low resolution will give shallower ( more preferable ) DOFs .

It can even be concluded that higher ISOs will create shallower DoF....

Something is wrong here .... We are mistaking overall blurriness / bluntness with shallow DoF !! :-*

SquamishPhoto

Something is wrong here .... We are mistaking overall blurriness / bluntness with shallow DoF !! :-*

Ding!Ding!Ding! We have a winner! :]

Msmoto

Always interesting to attempt to discuss what IMO is a very subjective topic, and do so with scientific terms. I see DOF in a couple ways...long lenses...even stopped down, there is almost none.....like 800mm at f/8.
Medium very fast tells, wide open, almost none.
Normal to wide, dependent heavily on f/stop, and finally a tilt lens....one can have DOF from a few inches to infinity when properly used.

I think it is time to test out the 135mm f/2 wide open again.....

spraynpray

There is no mistake. DoF is the area of acceptable sharpness. If a high Mp sensor plus high quality lens and use of lowest ISO give you best sharpness, then that will allow you to see the Dof more clearly - the inverse will add it's own blur to the existing blur and narrow the apparent DoF.

Golf007sd

Some images for comparison in order to add substance to the conversation.



D4 400 2.8 1/125 ISO 100 @ F/8.0
Larger Image



D4 400 2.8 1/2500 ISO 100 @ F/2.8
Larger Imgae



D4 400 2.8 1/1600 ISO 100 @ f/2.8
Larger Image

spraynpray

I don't think they help Ali - if they had something in which ran through the DoF so we could assess the apparent DoF it would have had some merit.

Really we need a low Mp, high Mp, Low ISO, High ISO, horrible lens and excellent lens version of the same subject* - have you got any? Not that I think you would have any horrible lenses of course! :P

*All permutations of course...

Golf007sd

...DoF is the area of acceptable sharpness....high Mp sensor plus high quality lens and use of lowest ISO give you best sharpness, then that will allow you to see the Dof more clearly...

The equipment used in obtaining the images in relation to your remarks was what I was trying to add substance to. If I understood your correctly.

As for "low Mp... horrible lens" sorry to say I don't have such equipment. :P

SquamishPhoto

There is no mistake. DoF is the area of acceptable sharpness. If a high Mp sensor plus high quality lens and use of lowest ISO give you best sharpness, then that will allow you to see the Dof more clearly - the inverse will add it's own blur to the existing blur and narrow the apparent DoF.

No, it isn't. The depth of field is strictly relevant to the size of the aperture opening and to the specific focal arrangement, and also to the size of the sensor or film being use. You guys are just confusing the matter for some reason.

Correlli

The factors that affect depth of field are:
- focal length
- subject distance
- f-stop
- magnification of the image (as print or on screen)
- viewing distance

Magnification and viewing distance are covered by the so called "circle of confusion" (the acceptable sharpness). There are some values that can be found in various books or on the internet (e.g. 0,033 mm for FX format). If you look at an image from a certain distance your eye can only resolve certain details and that is the limiting factor. A out of focus detail that your eye cannot distinguish from an in focus detail will still be regarded as in focus. So in fact the circle of confusion changes with viewing distance (it will be much smaller if you pixel peep that if you look at a print from a normal distance).

Sensor size does not affect the DoF at all. You will get the same dof if you use a 85 mm lens at f/5.6 from a given distance on DX format or on 8x10'. What does change is the field of view of course. If you take the field of view as a constant (e.g. a head shot portrait) you either need to change the focal length if you want to keep the distance or you need to change the distance if you want to keep the focal length. And this will impact the DoF.

spraynpray

@squamish: I don't think we are, we are just opening our minds to the fact that our eyes perceive the DoF as the area of acceptable sharpness (which in theory is created by the sensor size, aperture, focal length, distance to subject etc. etc.) but, if any image is recorded using low resolution equipment, would appear less clear, more blurred, shallower (call it what you want) and so as far as our perception is concerned, is affected by anything which does not maximise it in terms of sharpness.

Nobody here has said anything about reproduction size either - that is another factor.

EDIT: Wait - Correlli has just posted and he mentions magnification of the image on screen.

spraynpray

Sensor size does not affect the DoF at all. You will get the same dof if you use a 85 mm lens at f/5.6 from a given distance on DX format or on 8x10'. What does change is the field of view of course. If you take the field of view as a constant (e.g. a head shot portrait) you either need to change the focal length if you want to keep the distance or you need to change the distance if you want to keep the focal length. And this will impact the DoF.

I can't agree with you there Correlli, if you go to the Dofmaster site you can demonstrate it for yourself. Enter all of the fields then just change sensor size and you will see it change. I totally accept what you said about changing the subject to camera distance to regain the same image size on sensor though.

Correlli

@spraynpray: actually you are right and I was wrong. The reason is, that for the final product the image has to be enlarged differently (a 35 mm negative needs a different magnification than a 4x5' one for the same print size). This results in a different circle of confusion.

So let me re-phrase: dof depends on focal length, subject distance, f-stop and circle of confusion. The circle of confusion depends on magnification and viewing distance of the final image. Therefore the sensor size is part of the equation.

heartyfisher

LOL good ol DOF again ?

I totally agree with what every one has said !

spraynpray

I hope that fence isn't too high hearty!

FlowtographyBerlin

No, it isn't. The depth of field is strictly relevant to the size of the aperture opening and to the specific focal arrangement, and also to the size of the sensor or film being use. You guys are just confusing the matter for some reason.

Exactly. Saying the resolution of the sensor (given the same size) influences the depth of field is like saying if you only use a very unsharp lens (i.e. the same model just with lots of scratches) this will give you shallower depth of field.

But @spraynpray and @Corelli have layed it out meanwhile.

And +1 for @heartyfisher :-)

spraynpray

What you should do is try stepping away from the safe world of calculations as they are not what you see in the real world. What you see in the real world is the result of many imperfect things and so improving things in the chain must and does improve what you see. From the beginning I have been saying 'apparent' as a hint about the direction I am coming from, but y'all just keep safely replying from behind theory.

heartyfisher

To paraphrase Neo, "There is no Fence". I believe the confusion is in the circle of confusion. I believe the theoretical CoC is undisputed by both sides of the argument so far. Yet observed data points to a difference between the D7100 and the D3200. So either this data is wrong or there is another explanation. The only major known difference between these 2 sensors is the AA filter. If you understand what the AA filter does it basically makes a single point of light big enough to cover 4 raw pixels of the sensor. i(e making a 24MP= 24/4 or 8MP). Therefore for any theoretical CoC, the sensor with an AA filter will add the width of 2-4 pixels. To put it another way, to have the same perceived CoC the theoretical CoC diameter will need to be smaller by 2-4 pixels. This will be seen/perceived at the edge of the DoF therefore making the DoF slightly thinner. I believe this is basically what @spraynpray has been saying.

There may be another factor here.. The AA filter may be much "stronger" on the D3200 as it is probably harder to make the higher resolution AA filters needed for 24MP. so Nikon may have cut some corners with the D3200 and left in a stronger "standard" AA filter thus possibly splitting the light over greater than the required 4 pixels say maybe 16 pixels ? this theory can be tested by comparing the equivalent images from the D3200 with the D5200 and the D7100. if this is correct then the D3200 should also be much less susceptible to Moire. Note that there are 3rd party Strong AA filters you can get for your DSLR if you want to use it exclusively for Video to counter moire. These AA filters are known to reduce sharpness but video is only 2MP so it does not matter if you use it only for video.. You simply made your 24MP camera into a 2MP one ;-)

PS just found some tests showing Moire being higher on the D3200 vs the D5200 .. oh well maybe the expeed on the D5200 is better ;-)

roombarobot

Wow, this discussion is very tricky and is confused by many sites on the internet. Depth of field is an optical property controlled by the lens/aperture and the distance to the subject. Depth of field does not change due to sensor size itself, DoF changes when you change the distance to your subject.

However it gets confused, as many sites presume you keep the same field of view. In that case, if you use a full-frame sensor, you would need to be much closer to your subject to keep the same field of view, therefore your DoF is smaller. But it changed because you changed your distance to your subject in order to preserve the same field of view.

This site gets it mostly right:
http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

So, if you keep the same lens, aperture, & distance to subject, your DoF will be the same with a crop or full-frame sensor. Your image may be sharper/blurrier, however, depending on the Mpix of the sensor, aggressiveness of the AA-filter, etc., but the DoF is the same.

Paperman

So, if you keep the same lens, aperture, & distance to subject, your DoF will be the same with a crop or full-frame sensor.

I am sure you know it right but you are expressing it wrong somehow.

Same lens ( 50mm ), same aperture ( f8 ) same distance (3 meters ) ....

FF DoF = 1.85m

APS-C DoF = 1.17 m :O :O :O

The APS-C has shallower DoF ( may surprise some ) but it is simply because the same lens acts as a "wider" angle lens on FF than APS-C.


Uploaded with ImageShack.us />

Rephrasing by replacing " distance" with "size of object in frame " may get it right.

Ade

We already exhaustively discussed the main factors affecting perceived DoF -- including the effect of sensor size and CoC -- in this thread:

http://forum.nikonrumors.com/discussion/510/dof-fx-vs-dx/p1

What's new here is the claim that there is a noticeable perceived difference of DoF between a D3200 and D7100 in otherwise identical situation.

Without example pictures from a D3200 and a D7100 demonstrating this difference so we can all see what's actually going on, I feel we're just rehashing the old thread again.

roombarobot

@Paperman, I am not sure that the calculator that is quoted above is actually producing the right results. I don't know the assumptions that they have programmed into it. I have seen similar calculators that are programmed incorrectly.

Don't believe everything you read on the Internet.
--Abraham Lincoln

Paperman

You'd better believe this one mate ! It is the mostly used one - comes as Apps as well.

Feel free to check with others you know. The figures may vary slightly depending on CoC but the conclusion won't !

heartyfisher

Paperman is right.. Most people think FF has the DOF advantage but for the same Focal length and same subject dist same aperture, the DOF is thinner on a DX camera vs FF. But this has been discussed in the other thread already. Just go there to discuss/read about it.

As Ade says we need test images from D3200 and D7100 to progress this discussion.