Does format size reduce lens brightness???

[DSG]

I was reading a review for the Tamron SP 300mm f2.8 LD IF (60B) (Which I already have BTW) when I spotted this unsettling sentence:

"The big question is, how does it compare to a comparable ZD 4/3rds lens? Remember that all versions of this Tamron 300mm lens are designed for film cameras so its image circle is sized for a 35mm film frame. With a 4/3rds camera, only half that diameter is used and, as a result, you can expect the image to be slightly less bright compared to a lens that focuses all of its light onto a 4/3rds image sensor at the same f-stop."

Is this true???..If I use a full frame lens on a crop format sensor camera do I actually lose f-stops of brightness at the sensor for the same f-stop on the lens?
I hope not as it would mean that when my 60B is wide open on my 1.7x crop camera it would in fact only be as bright as it is at about f4.5-f5 on a full frame sensor or film camera!
I also have a M645 80mm f2.8 N, made for much larger format sensors/film...If the above is true, how bright would that actually be when wide open on my crop format camera?
Before I read the above quote I thought that the brightness of the light in the lenses image circle stays constant per lens f-stop no matter how small the sensor format is, but now I'm confused...What do you guys think?

[fermy]

No, this is incorrect, you don't lose f-stops if you go to a crop body. What matters is the amount of light per unit of area and this stays the same. Actually, I'd expect to slightly gain brightness on a crop body as for most lenses there is light fall-off towards the corners (at least wide open and couple of stops down) and these darker corners get cropped.

[mdarnton]

I read sentences like that, and I despair for the future (or lack of same) of the human race. I hope that was some little public self-posted thing on a buying site like Amazon, not a real review that someone got paid real money for.

[buerokratiehasser]

You do lose light, but 5.6 stays 5.6 if the ISO stays the same.

You lose light - compared to what the lens is capable of catching

f5.6 (actually T5.6) stays the same - these things are following standards .. same ISO is same ISO. Of course, since the sensor has to deal with less light per pixel (assuming pixels are the same), it may be more difficult for the sensor to deliver that ISO in the first place. But that's another calculation.

If you are doing these gedankenexperiments, always start with a biig film plate where the sensor is. It doesn't matter how you crop that film plate, you will always have t5.6 (subject to falloff effects). (The larger formats and lenses are actually often "not that good" as 35mm with respect to lpmm and are usually stopped down way more. F32 makes same fuzziness as F32 on 35 mm. Their massive detail is just because the film is so freaking large, and tilt/shift helps a bit)

There are "equivalent" aperture calculations from Frank Klemm, who says you can model your 1/2.3" compact sensor and 6mm f3.3 lens as equivalent full frame 35mm f18 or something (not checked for the exact numbers). But these calculations involve comparisons of whole systems of sensor - lens - aperture. (I.e. you would need f18 to get the same diffraction and DOF relative to pixel pitch on full frame)

Again, you get (macro bellows pain excluded) the same light as with a crop lens with same T. Their advantage from the wasted light is just that they can be made smaller and cheaper because they don't have to light up full frame.

[Opticus]

Smaller pixels tend to catch less light (photons)?

[aspen]

The reviewer is misleading, and inaccurate. The light gathering property of a lens remains consistent across formats. A f2.8 will have the capacity to gather, relatively, the same amount of light. What isn't mentioned in the synopsis that you provided is the register distance of any particular lens. One would intuit that the closer distance to the plane of film (sensor) then the MORE light would reach it without fall-off (and taking into consideration the image circle). But this is a red herring. The writer uses linguistic hedges, "slightly", "less", and "bright", as if this conveyed some level of objective, and measurable, accuracy, which it does not.

[DSG]

[buerokratiehasser]

Well, f-stops technically only give the f/d ratio, and t-stops are not given so... it may not violate the laws of physics too much to concentrate light almost exactly into the circle you need, and nothing more. In practice however I think this is not done and the circle you see mounting it on ff/falloff effects worse than cos^4 come from artificial obstacles. Also 0.7 at the border is regarded as bad already, there is not much room for artificial "peaking".. but I don't think lens review check for t-stops or whether you actually get your 1/500 you are supposed to have at f4.
(NOTE: the preceding is for any lens - FF/crop doesn't come into play)

I only read that review hastily, but I think the brightness issue is nonexistent compared to the following points
- no autofocus. DUH!
- possibly other things not working (aperture, metering, exif)
- can at least use focus trap with that body?
- no autofocus. DUH!
- why purchase this when you can get M42 300/4.5, 300/5.6 or 500/8 for less. They are manual too and possibly cost less than that adaptor you need for OM or whatever.

[DSG]

[Lightshow]

Yeah, the properties of a lens don't change, the amount of light that reaches a sensor will not change if you measure the same area(mm^2).

[DSG]

[martinsmith99]