f/1.2 vs. f1.4

Hard to say, but definitely not a full stop.

The scales are rounded as some have noted, but the 1.2 figure appears both on the 1/3-stop and 1/2-stop scale

1/3 stop scale
f/1, f/1.1, f/1.2, f/1.4, f/1.6, f/1.8, f/2

1/2 stop scale
f/1, f/1.2, f/1.4, f/1.8, f/2

So, it could be 1/3 stop or 1/2 stop, hard to know.

Terribilino said:

Just wondering... is f/1.2 a full stop faster than f/1.4, a half
stop, or none of the above?

Click to expand...

--
Kjeld Olesen
http://www.acapixus.dk

hunghang said:

Since the f-stops are related to the areas of a circle, there is a
squared relationship. Hence 1.4^2 / 1.2^2 is 1.36.

Therefore, the f/1.2 allows 36% more light in!!! This is approx 1/3
stop.

Click to expand...

Hmm, not quite right! Stops are exponential, not linear.

The discrepancy here (as pointed out in the above thread), is whether the f/1.2 lens is actually a f/1.26 or f/1.19 lens. The consensus seems to be that it's actually an f/1.26 lens (or thereabouts), so is about 1/3 of a stop faster.

GaborSch said:

I don't know, how to separate these two effects, except with a film
and a digital camera parallel (as the second effect is digital
specific); scanning in the film would give a good indication of the
optical vignetting, without the "well effect".

Click to expand...

You can separate them by removing the camera from the equation. f-stop is a physical property of a lens. The camera it is mounted to does not matter and has no effect on the measurement. Opening the aperture more lets more light pass through the lens barrel. There is no more to consider when talking about f-stops.

Rounding to the nearest tenth, the 1/3 stop scale would go:

1.0
1.1
1.3 (rounded from 1.26)
1.4
1.6
1.8
2.0

and the half stop scale would go:

1.0
1.2 (rounded from 1.19)
1.4
1.7
2.0

Kjeld Olesen said:

The scales are rounded as some have noted, but the 1.2 figure
appears both on the 1/3-stop and 1/2-stop scale

1/3 stop scale
f/1, f/1.1, f/1.2, f/1.4, f/1.6, f/1.8, f/2

1/2 stop scale
f/1, f/1.2, f/1.4, f/1.8, f/2

So, it could be 1/3 stop or 1/2 stop, hard to know.

Terribilino said:

Just wondering... is f/1.2 a full stop faster than f/1.4, a half
stop, or none of the above?

Click to expand...

--
Kjeld Olesen
http://www.acapixus.dk

Click to expand...

--
http://www.pbase.com/victorengel/

You are making too much sense. Take f/3.5 for instance. The "real" number is 3.563, but they don't round that up to f/3.6. I have no clue what the logic is behind the numbers they actually put on the camera/lens.

I suggest you to ignore the effect of light-loss due to the wells; only the diameter of the aperture counts. Never mind underexposure.

I suggest you to ignore the effect of optical vignetting; only the diameter of the aperture counts. Never mind the dark corners.

And, if we are at it, I suggest you to ignore, that the sensor is cropping part of the projected image. The diameter of the projected image is a physical property of the lens; why bother with the camera's properties? Count on 46° diagonal angle of view of a 50mm lens, notwithstanding the camera.

I, for myself, don't create images with a lens , but with a system and I see the lens as part of the system.

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

I know. For practical purposes, the difference between f/1.2 and f/1.3 is barely noticeable anyway. For exposure, we don't really care about the f-stop either. We care about t-stop. F-stop is relevant to depth of field issues, though. I bet the difference between f-stop and t-stop values is similar to the discrepancies we're discussing here.

I suppose someone could measure the diameters of their lenses and calculate the f-stop more precisely (an 85mm f/1.2 should have a diameter of 71mm), but then, we run into a similar issue with imprecision in the focal length of the lens.
--
http://www.pbase.com/victorengel/

I wonder if Canon has the "real" information available.

Well we know this much. The 85mm f/1.2 lens has a filter size of 72mm. In order for it to really be an f/1.2 lens, if the focal length is really 85mm, the maximum aperture must be 85/1.2 or 71mm. That leaves 1mm wiggle room to be used by housing, etc.

Now let's compare to the f/1.8 lens. It has a filter size of 58mm. At f/1.8 the maximum aperture must be 85/1.8 or 47mm. That leaves 11mm wiggle room for housing, etc. One would expect that the larger the lens, the more pressure there would be to make the housing as small as possible to keep the lens from being too cumbersome. Given that, it makes sense the f/1.8 would have relatively more housing.

How about comparing a few other lenses. Here I list focal length, maximum aperture in f-stop according to spec, resulting aperture in mm, and filter size (also maximum diameter according to http://consumer.usa.canon.com/app/pdf/lens/EFLensChart.pdf ).

50 1.4 36 58 (73.8)
50 1.8 28 52 (68.2)
85 1.2 71 72 (91.5)
85 1.8 47 58 (75)
100 2 50 58 (75)
135 2 68 72 (82.5)
135 2.8 48 52 (69.2)
200 2.8 71 72 (83.2)
300 2.8 107 * (128)

• Lens does not have front filter

Comparing all these lenses, I conclude that f/1.2 might be a bit of a stretch for the 85 f/1.2 lens. It's just as likely as the 200 f/2.8 is really an f/2.8 lens.
--
http://www.pbase.com/victorengel/

I never measured a lens cap to see if the diameter they refer to is inside or outside. If it's outside, then Canon has some 'splaining to do. Otherwise, it's ok. The actual iris doesn't need to be the size that the calculated f-stop would lead you to believe. They can put the iris further back in the lens after the light has been funneled down the tube some. If all the light from a 85mm lens get funneled to 36mm at the end, the iris could be only 29mm and still admit the same amount of light as if it were at the front of the lens and 71mm.

jbaker6953 said:

I never measured a lens cap to see if the diameter they refer to is
inside or outside. If it's outside, then Canon has some 'splaining
to do. Otherwise, it's ok. The actual iris doesn't need to be the
size that the calculated f-stop would lead you to believe. They can
put the iris further back in the lens after the light has been
funneled down the tube some. If all the light from a 85mm lens get
funneled to 36mm at the end, the iris could be only 29mm and still
admit the same amount of light as if it were at the front of the
lens and 71mm.

Click to expand...

The narrowest portion of the lens housing in front of (between lens and subject) the lens can't be smaller than 71mm (in this example). Taking a caliper to the intersection of the front element to the housing probably would be a reasonably accurate figure. I don't have an 85mm f/1.2, or I'd try this. I do have an f/1.8, though, so I think I'll try this measurement when I get time.

--
http://www.pbase.com/victorengel/

If you just want a quick method of answering that type of question then do the following:

Multiply the aperture value by itself and do the same for the comparison aperture value.

In this case, you asked about 1.2 and 1.4

1.2 x 1.2 is 1.44

1.4 x 1.4 is 1.96

Remember that one stop is photgraphic talk for the concept of letting twice as much light in. A f/1.2 lens lets in 1.44 units of light. A f/1.4 lets in 1.96 units of light in. A f/1.4 does not let in twice the amount of light in as the f/1.2. Therefore, to quickly answer your question, no, f/1.2 is not a full stop faster than f/1.4.

Summary - multiply the values by themselves and ask the question: is the first product double the second (or vice versa)?

--
30D, 17-40L, 24-105L, 70-200 4L, 300 4L, 100 Macro, 430 EX

A lens is part of the system, but the same lens can be part of many different systems. That means you need a way to quantify its properties in a way that is independent of the system on which it is being used. The aperture of a lens is one such measurement.

The simplest answer and the one you'll get from your camera is a 1/2 stop:

Set it to 1/2 stop change and you'll get the sequence

1.0 1.2 1.4

If you use 1/3 stops you get

1.0 1.1 1.3 1.4

All the numbers are rounded but who cares about 1/6 of a stop or less.

from 1.0 to 1.44 is 1 stop (square root of 2). THerefore, 1.2 to 1.4, according to the manual, would be a third of a stop.

phillip pian said:

from 1.0 to 1.44 is 1 stop (square root of 2). THerefore, 1.2 to
1.4, according to the manual, would be a third of a stop.

Click to expand...

No! It's 0.5 stops rounded to nearest value. 1.1 is 1/3 stop faster than 1.0 and 1.3 is 2/3 stop faster than 1.0.

0.5 stops faster means 50% more light is let through, so

sqrt(1.5) = 1.22 ~ 1.2. Thats it!

sqrt(1.333) = 1.15 which is rounded down. Yes I know it not entirely accurate but we are talking 1/6 stop or less. Clearly we can't round sqrt(1.5) and sqrt(1.333) to the same value.

thx1138 said:

phillip pian said:

from 1.0 to 1.44 is 1 stop (square root of 2). THerefore, 1.2 to
1.4, according to the manual, would be a third of a stop.

Click to expand...

No! It's 0.5 stops rounded to nearest value. 1.1 is 1/3 stop faster
than 1.0 and 1.3 is 2/3 stop faster than 1.0.

0.5 stops faster means 50% more light is let through, so

sqrt(1.5) = 1.22 ~ 1.2. Thats it!

Click to expand...

Oops no it's 41% more light of course since it's geometric average, so sqrt(1.41) = sqrt(2^(0.5)) = 1.189 ~ 1.2

thx1138 said:

sqrt(1.333) = 1.15 which is rounded down. Yes I know it not
entirely accurate but we are talking 1/6 stop or less. Clearly we
can't round sqrt(1.5) and sqrt(1.333) to the same value.

Click to expand...

In this case it's sqrt(2^(0.3333)) = 1.122

In general for increments of x stops it's

sqrt(2^(x)) * previous f-stop.

Sorry to nitpick, but sqrt(2) is 1.414, not 1.44. I'm an engineer, what can I say.

phillip pian said:

from 1.0 to 1.44 is 1 stop (square root of 2). THerefore, 1.2 to
1.4, according to the manual, would be a third of a stop.

Click to expand...

Colden said:

Sorry to nitpick, but sqrt(2) is 1.414, not 1.44. I'm an engineer,
what can I say.

Click to expand...

I think he was rounding-out some digits... close enuf for gummint work or landing spacecraft on mars...

Gorim said:

Colden said:

Sorry to nitpick, but sqrt(2) is 1.414, not 1.44. I'm an engineer,
what can I say.

Click to expand...

I think he was rounding-out some digits... close enuf for gummint
work or landing spacecraft on mars...

Click to expand...

Uh... 1.414 is rounded. 1.44 is just wrong. If 1.44 was gleaned by squaring 1.2 (which is a rounded version of 2^(0.25)), the correct answer is 1.4, not 1.44, because there are only 2 significant digits in 1.2.
--
http://www.pbase.com/victorengel/