I came to know that a 50mm prime lens on a "cropped" sensor behaves like an 85mm prime lens.So, does the 35mm prime lens behave like a 50mm prime lens on a "cropped" sensor?If yes, then will it show the same bokeh as the 50mm lens on a "cropped" sensor?

no, the 35mm behaves like a 35mm, you just dont see as much whit a crop- sensor. but bokeh, perspective and everything stays the same! 

Depends what you're looking at specifically ;-)

but frederic, change of dof, perspective and bokeh results from the changing viewpoint and is not a result of the crop- factor.... 

Depends what you keep as a constant in the comparison. In my comparison, the constant was that you wanted to get the same framing, i.e. fill the frame with the same motif instead of having the image from the DX as a cropped version of that of the FX.

Didn't understand this. Please write that statement once again with only cropped sensor's respect.

Hi Frederic
I'm confused by you sentence here
I can easily understand the cropping factor due to the size of the sensor

Where I'm lost is about the aperture
When I'm using an external light meter, the result (speed, aperture) will be the same for any kind of camera (FX, DX, Full Frame format)


If I'm looking to the mathematical formula
need to scratch my head a bit but if my memory is good
F-Number = Focal / Diameter of Diaphragm
So for me the (real) Focal of a lense is not depending of the size of the sensor, but it's a physical data

as mentionned, I'm confused here

Just look at the second image in the link above. You'll see that the DX-crop is a ... crop of the full-frame image. That is as Remo rightfully states when using the same POV and the same lens, but the only thing that was changed was that we have used a DX-cropped sensor instead of a full frame one. 

I am not completely sure here, but I think it has to do with the working distance being different, as we want to keep the same framing.

I use "field of view" rather than angle.  The angle is constant.I'm with Remo.  You are are speaking of the the relative amount of sensor real estate taken up by the area of interest (motif).  Moving closer to the motif *changes* the perspective.

Yes, you're right, King. "field of view" is a better term.

lets say: a 50mm lens on a camera with crop sensor will produce the same photo (perspective, dof, bokeh) than the same lens on a full frame.

but:

if you move 5 steps away from your motive to have the same framing it will change perspective, dof and bokeh because the distance to the motiv has changed.

My 5 cent wrt bokeh/DOF: I could not believe the relationship Frederic highlighted myself, so I had to prove it (some time ago already): I shot with a P&S with crop factor around 5 and check aperture (no control, so I take what I get). Then I did the same picture (same perspective) and applied different apertures at my DSLR (crop 1.5). You will see that you get similar DOF/bokeh only with much higher DX aperture - almost proportional with crop difference (slightly less, because of the larger magnification involved). So my opinion would be DX aperture 1.4 is closer to FX at 1.9 (1.4 aperture * 1.5 = 2.1 minus a little bit), but that difference to Frederics 2.2 is probably irrelevant for most applications. Maybe I did some mistake (it was no scientific experiment) and Frederic is right, does not matter. In the end the relationship may differ by lens, by object/background distance, ... but generally: you get nicer bokeh with larger sensors, longer lenses and larger apertures. All clear? Wolfgang

Yes, normally would moving closer change perspective. By definition, I would say.

This is simply not true unless some variables that should remain constant for a valid comparison have changed, e.g., camera-subject distance.

Yes, you did.  See my post just above.  How did you judge between the two images?...by observing them on your monitor?  Of course you did.  You did not hold image enlargement constant.

Yes, exactly, and the distance has to be changed to get the same framing...

Thats what I call same perspective - same magnification... actually my results confirm mostly what Thom also calculated, so it should not be too far off :-) But in some way we can discuss this forever, because you cannot construct 2 completely identical scenarios to compare. So every comparison is weak.

Simply not true and this is very easy to demonstrate...easier to demonstrate than to explain, although I'm going to try in a minute.

Not at all the same thing.  This is a matter of physics, not perception. 

I agree completely. I am also trying to approach this using my background in applied physics, and sources that I trust...

Mostly true if you are comparing FX to DX (I'm sure it could be done in the lab).  Not true if you are working with film.

Hi King
your experiment was my first understanding about what Frederic was talking about
Your experiment is mostly change the size of the "Sensor"
and your experiment is really clear for me and that's were I was confused

After some reading on the provided web site, I have understand a bit more
my understanding is to have identical picture
- one taken with a DX and a 35mm lens
- the other one taken with a FX and a 50mm lens

To achieve that, you will have
- to move between the two shots (keep the same framing)
- adjust the aperture to keep the same DOF

Am I good this time Frédéric ?

That is the same as saying, "My constant is variable."

Bokeh varies from lens to lens depending on the number of lens elements and the number of blades in the aperture.
So yes the bokeh will vary.

King, I think thats really a topic we should discuss over a glass (or more) of beer... Without being deep in physical or optical laws, I try to understand it from a systemic perspective. It is a topic with many (infinite?) facets. Probably I already stated smy issue omewhere - I love these discussions, but it keeps me from getting out shooting :-) Wolfgang

beer changes my perspective too!

Wolfgang, have you ever noticed that the DOF and Bokeh when viewed through the bottom of a beer glass varies with how much beer has been drunk from the glass?

Yes, I completely agree... (Quality of) bokeh varies with the actual build of the diaphragm; i.e. the number and shape of the blades. I am not completely sure what the "number of lens elements" would have as an effect, though. Could you explain that part, Ken?

It has also to do with what number the glass is... Even with identical glasses, this becomes a time-variant system ;-)

Anisha, is it all clear now?

if 'f' is a focal length of a lens in mm then its 'equivalent' focal length 'fe' on a sensor or film of a certain size will be 'fe = f * k', where 'k' is so called crop factor and is 1 for 35mm film or full-frame sensor (which is of the same size as 35mm film), 'k' is greater than one for smaller sensors, say in case of a four-thirds sensor (which is twice as small liner-wise, four times area-wise, than 35mm film or full-frame sensor) 'k' is 2, for some APS-like size sensors 'k' is around 1.5 or 1.6, etc., the reason for this 'confusion' is the fact that how much of a scene will be recorded on a sensor or film depends on both focal length 'f' and the sensor or film size, but as 35mm film was de facto a standard for years photographers kept mentioning only focal length 'f' in mm (implicitly assuming the 35mm film), so they would say that a 50mm lens gives you a 'normal' view (similar as a human eye), but if a 50mm lens is put on a camera with a sensor smaller than 35mm film it will be 'equivalent' (as far as which part of the scene will be recorded on the sensor or film) to a longer lens put on 35mm sensor camera, say a 50mm lens put on a camera with four-thirds sensor is 'equivalent' to a 100mm lens put on 35mm sensor camera, the larger the photo site the better the image quality, that is why larger sensors give better image than smaller (assuming thy have the same number of pixels), note that if you shoot the same scene with full-frame sensor camera with 50mm lens and four-thirds sensor camera with 25mm lens you will get more DoF blur on full-frame sensor camera as DoF blur depends on focal length 'f' in mm (absolute length measure), that is why our eye has very little DoF blur (short focal length and a small 'sensor'), the 'k' factor can be approximately calculated as 'k = d35/ d' where 'd35' is the length of the diagonal of 35mm film and 'd' the length of the diagonal of the sensor or film in question, I hope that this was 'crystal clear' as is a fresh spring water :)

I see we have a stand in for Wolfgang if we ever need it!!


OK just to make things just a bit simpler: how much of all of the above including these two quotes is made moot if the 35mm lens is this one:
Nikon AF-S Nikkor 35mm 1:1.8G DX
You see this lens and many others like it are made specifically for APS sensors and doesn't throw a big enough circle of light to cover a full sized sensor. So the notion that capturing only a portion of the circle of light give you better edge sharpness is gone, same with vignetting, etc.
Sorry Anisha...
 

The same applies, really ;-) I would say this lens is having very severe vignetting on a FX-sensor.

You simply can't say that all of the above applies to a lens like this. Maybe some of it does, but certainly not all of it. And yes you would not only see vignetting, you would see a big circle of black all around the edges.

Wiki says ut better tha me "Differences in lens aberrations and aperture shape cause some lens designs to blur the image in a way that is pleasing to the eye, while others produce blurring that is unpleasant or distracting,"good" and "bad" bokeh, respectively."

I think I'll take that as a compliment, Clyde.

You may, but I was referring to Robert...based on his redefining "f" to refer to focal length instead of aperture...which he did to clarify things even more for dear Anisha.

In some way, you can say that a DX-lens is a lens that is vignetting so heavily that it is considered useless on a FX-body. For some this would mean just a plain black circle, for others this would be a more gradual transition that is so heavy that it is considered useless.OK... thank you for clarifying. 

Some times ago, I have played with a friend with these DX lenses on a FX camera

The Camera was a D3
and
The lens was a AF-G 10,5mm Fisheye DX

here is the result
http://www.flickr.com/photos/ganymede2009/4398101607

Who was talking about a small vignetting????
;-))))

I have the Tokina 12-24mm f4 which is a DX ultra wide zoom. It's actually quite usable on a film camera between @16-24mm but if you zoom all the out to 12mm you get a fully round circle of black showing.

Anisha, the simple answer to your OP question is yes. In many ways a 35mm would behave like a 50mm. Would it show the same bokeh as the 50, well no because no two lenses will show "the same" bokeh. If you mean the same degree of bokeh, then yes pretty much the same if you keep ALL things constant and just change the lens. What varies lens to lens is the nature and perceived quality of the bokeh.

Thanks everyone for the explanations. I have tried to absorb some of it (though I can't say that I have understood the technical talks 100%). :):).

I actually meant to ask about the comparison of shallow depth of fields, not the bokeh. I thought shallow depth of field and bokeh are the same things, but now I realize that they are not!!
Bokeh is just the shape of the light sources in the background! :doh:

Anyway, Frederic explained quite well the differences, so thanks to him (and of course to others too who raised questions on his post for further clarifications).
And thanks also to Clyde for referring to me as "dear" Anisha. That was a sweet gesture.

Hey!!!

the term "circle of confusion" pops up in my mind ... :-)

The term "over sensitive" pops up in my mind...