question about aperture

[BlahBlahYouToo]

is it a standard value or is it relative to the lens itself (so different for each lens)?

so the opening on a 2.8 50mm prime is the same as the opening on a on 18-200mm 2.8 fixed aperture @ 50mm (fictional lens)?

 

[GoSharks]

Relative to focal length. So yes to your question.

 

[BlahBlahYouToo]

awesome, thanks.
so it's both really then, isn't it? standard across different brands/types of lens, and also relative to focal length.

1. T/F?
the opening of a canon f2.8 85mm prime is equal to the opening of a fixed aperture f2.8 nikon 55-300mm @ 85mm.

2. lets assume that the fictional zoom nikon lens in the above example is zoomed all the way out to 300mm f2.8.
the opening is different now when compared to the canon prime, T/F?

 

[AkumaX]

1) T
2) T? you're basically saying you're comparing an 85mm f2.8 to a 300mm f2.8, of course its different?

gimme your fictional lenses plz

(although, the 70-200mm f/2.8 is real haha)

==

hey i'll throw in a curve ball at you

50mm f/1.4 on a full frame == 85mm f/1.8 on APS-C

 

[maniac5999]

1) T
2) T? you're basically saying you're comparing an 85mm f2.8 to a 300mm f2.8, of course its different?

gimme your fictional lenses plz

(although, the 70-200mm f/2.8 is real haha)

==

hey i'll throw in a curve ball at you

50mm f/1.4 on a full frame == 85mm f/1.8 on APS-C

To expand on this line of thought, is the light on the sensor the same, at the same apeture regardless of focal length, or does it change?

i.e. if I have a 17-200mm f/2.8, and I point it at a graycard at 17mm apeture, and proper exposure is 1/100th, then I zoom it in to 200mm, is the proper exposure for 18% gray still going to be 1/100th, or will it change with focal length?

 

[ElFenix]

the f-numbers are actually relative apertures, not absolute apertures. in the same light, the same relative aperture will give the same light intensity at the body end of the lens, be it a 24 f/2, 50 f/2, or 100 f/2. because they give the same intensity, for the same shutter value they'll give the same exposure (well, close enough--due to internal reflection they'll all be a little off, the t-stop was developed to account for that).

absolute apertures are the aperture measured in millimeters. so, for your T/F questions, 1. is T and 2. is T. a 300 at f/2.8 is a 107mm aperture.

hey i'll throw in a curve ball at you

50mm f/1.4 on a full frame == 85mm f/1.8 on APS-C

no, you've got that wrong. for equivalence purposes, a 50 f/1.4 on full frame would be roughly equivalent to a 31 f/0.9 on APS-C. for equivalence you need the same angle of view (so divide 50 by the 1.6 crop factor to get 31.25) and the same absolute aperture (50/1.4 = ~35.7 = 31.25/0.875).


equivalence may be very difficult to pin down with certainty due to the nature of digital sensors. short dof is created by the rays coming from the edge of the lens. because they are coming in relatively flat to the sensor, they are very difficult for the sensor to catch (and at certain relative apertures are impossible for the sensor to catch). basically, the sensor can't capture additional light that the lens throws after the lens is opened up past a certain point. so, super fast lenses may be pointless on digital sensors.

 

[slashbinslashbash]

BlahBlah, I think I know what you're getting at here, and I have kind of an answer, but not a very good one because I don't really understand it either.

Let's take a 70-200mm f/2.8 constant max aperture lens. Obviously there is a single diaphragm mechanism within this lens; it is a fixed maximum size, no matter what the zoom level. But an f/2.8 aperture at 70mm would be 25mm in diameter, while an f/2.8 aperture at 200mm would be 71mm in diameter. So if the physical size of the diaphragm mechanism is all that matters, this lens should be a variable-aperture lens, approximately f/1.0 at 70mm, and f/2.8 at 200mm. But we know this is not the case. Now what is the reason for this?

It is because the aperture that matters is the "apparent aperture", not the real, physical aperture. How the apparent aperture is calculated is beyond my knowledge and probably requires a lot of arcane optical equations. There is part of an answer here:

http://photo.net/bboard/q-and-a-fetch-msg?topic_id=23&msg_id=0000Wf

along with a recommendation for a detailed optics book.

To expand on this line of thought, is the light on the sensor the same, at the same apeture regardless of focal length, or does it change?

i.e. if I have a 17-200mm f/2.8, and I point it at a graycard at 17mm apeture, and proper exposure is 1/100th, then I zoom it in to 200mm, is the proper exposure for 18% gray still going to be 1/100th, or will it change with focal length?

The aperture value is supposed to be constant for level of light, across all focal lengths (although the physical aperture size changes with the focal length).

50mm f/2.0 = 25mm aperture
50mm f/4.0 = 12.5mm aperture (1/4 the light of 50/2.0 => 4x shutter speed to keep same exposure)
100mm f/4.0 = 25mm aperture (same size as 50/2.0, but same light as 50/4.0)
50mm f/2.0 = 50mm aperture (same light as 50/1.0)

So, to answer your second question, yes, your exposure will stay the same at 17mm f/2.8 as at 200mm f/2.8. This is the great thing about constant maximum aperture lenses. Variable maximum aperture lenses are hard to use in manual mode because you constantly have to adjust when you zoom (if you are at or near maximum apertures, that is).

no, you've got that wrong. for equivalence purposes, a 50 f/1.4 on full frame would be roughly equivalent to a 31 f/0.9 on APS-C. for equivalence you need the same angle of view (so divide 50 by the 1.6 crop factor to get 31.25) and the same absolute aperture (50/1.4 = ~35.7 = 31.25/0.875).

I think he meant a 50/1.4 lens becomes the equivalent of an 85/1.8 lens on a crop body. I know this is true from a DOF point of view, but when it comes to exposure, I'm not so sure (and would welcome a lesson on this). It seems to me that, when using a 50/1.4 on a crop body, you are only using the middle part of the image circle, which will be, on average, brighter than the entire image circle, i.e. what is used on full-frame. Also, it seems that I have never noticed a difference between my crop an full-frame bodies when it came time to set exposure. I can set my 5D and my 20D to the same exposure settings (ISO, aperture, shutter speed) and get a photo with roughly the same apparent brightness. Of course this could easily be something that the camera makers compensate for by making the ISO equivalents different for crop and full-frame bodies, but I have no idea whether this is the case. I would love to hear more info on this.

 

[ElFenix]

I think he meant a 50/1.4 lens becomes the equivalent of an 85/1.8 lens on a crop body. I know this is true from a DOF point of view, but when it comes to exposure, I'm not so sure (and would welcome a lesson on this). It seems to me that, when using a 50/1.4 on a crop body, you are only using the middle part of the image circle, which will be, on average, brighter than the entire image circle, i.e. what is used on full-frame. Also, it seems that I have never noticed a difference between my crop an full-frame bodies when it came time to set exposure. I can set my 5D and my 20D to the same exposure settings (ISO, aperture, shutter speed) and get a photo with roughly the same apparent brightness. Of course this could easily be something that the camera makers compensate for by making the ISO equivalents different for crop and full-frame bodies, but I have no idea whether this is the case. I would love to hear more info on this.

equivalence is the matching of several things that make up the image but can result in differing image quality.

• perspective
• field of view
• dof (and focal point)
• shutter speed

dof is dependent on absolute aperture rather than relative aperture. so the exposure is different with different sensor sizes. to compensate for that, you need to change the sensitivity so that apparent exposure remains the same.

when you put a 50f/1.4 on a canon crop body it is equivalent to a 80f/2.24 on a full frame. to lose 1 stop in equivalency your crop factor would have to be the square root of 2, usually denoted 1.4 in photography. that's because it's using a linear function to express an area. that's why you lose 2 stops in speed going to 4/3 from 35 mm for equivalent images (i.e. the 4/3 35-100f/2 lens is equivalent to a 70-200f/4 lens on a 5D).

under the ISO standards for apparent exposure, the camera is allowed to do basically whatever it wants to get the apparent brightness it thinks you are looking for. it's basically the same as when kodak or fuji would sell the same emulsion film in both 400 and 800 packaging, the difference being the processing instructions. the cameras seem to do this more with faster lenses. that tells you not as much light is being captured by the sensor as one would expect moving to a faster lens (iow, going from f/2 to f/1.4 doesn't get you a full stop reduction in sensitivity like you would expect).

 

[AkumaX]

no, you've got that wrong. for equivalence purposes, a 50 f/1.4 on full frame would be roughly equivalent to a 31 f/0.9 on APS-C. for equivalence you need the same angle of view (so divide 50 by the 1.6 crop factor to get 31.25) and the same absolute aperture (50/1.4 = ~35.7 = 31.25/0.875).

Yeah, I totally reversed it :whiste:

 

[BlahBlahYouToo]

thanks all.

just heard the iphone 4s has a f2.4 max aperture.
pretty much clear proof that it's relative.

 

[AkumaX]

thanks all.

just heard the iphone 4s has a f2.4 max aperture.
pretty much clear proof that it's relative.

yeap! relative to sensor-size

compared to 35mm, think about aps-c (1.6x), m4/3 (2.0x) and nikon's new cx sensor (2.7x). it just sounds "funny" that there's a 10mm f/2.8 for the cx sensor, but translates to 27mm when scaled to 35mm range

 

[ElFenix]

yeap! relative to sensor-size

compared to 35mm, think about aps-c (1.6x), m4/3 (2.0x) and nikon's new cx sensor (2.7x). it just sounds "funny" that there's a 10mm f/2.8 for the cx sensor, but translates to 27mm when scaled to 35mm range

no, it's relative to focal length. the f in f-stop is a reference to focal length.

 

[slashbinslashbash]

no, it's relative to focal length. the f in f-stop is a reference to focal length.

This is true, but the focal length required for a given angle of view is directly correlated to the sensor size. The lens in the iPhone has a 4.28mm focal length.

 

[ElFenix]

This is true, but the focal length required for a given angle of view is directly correlated to the sensor size. The lens in the iPhone has a 4.28mm focal length.

well, sort of. i could crop. though getting a 4.28 mm focal length lens on my canon would be next to impossible. but i have a 10 mm lens and if i crop it down to nikon cx size it'd have the same fov as the 10 mm on the nikon.

a 10 mm lens is a 10 mm lens is a 10 mm lens regardless of the sensor behind it.

 

[slashbinslashbash]

well, sort of. i could crop. though getting a 4.28 mm focal length lens on my canon would be next to impossible. but i have a 10 mm lens and if i crop it down to nikon cx size it'd have the same fov as the 10 mm on the nikon.

a 10 mm lens is a 10 mm lens is a 10 mm lens regardless of the sensor behind it.

Of course. But a 10mm lens on a 35mm film sized sensor would be massively wide-angle, while on an iPhone sized sensor it would be a telephoto. (Not to mention the required diameter of the lens elements to cover the sensor.) To get something in the traditional "moderately wide to normal" (30mm to 50mm) angle of view that we all expect from a fixed-lens camera such as you'd find on a smartphone, the lens' focal length needs to scale with the sensor.

 

[ElFenix]

well, technically it'd only be a telephoto if the lens was physically shorter than it's focal length

 

[RobDickinson]

Its easy to make 'fast' lenses for very small sensors because of the very small focal lengths involved.

Remember though the small sensor captures a lot less light due to much smaller surface area (amongst other reasons), so will need to push the amplification (ISO/Gain) more and also wont be able to isolate subjects even with a 'fast' lens