[Digitalis]

lens coatings actually increase the transmittance of a lens, not reduce it.

uncoated glass lenses from the 1930s lost nearly 60% of the light that passed through them and they only had 4~6 pieces of glass in them, and these lenses were quite slow; f/5.6 was as fast as it got. In contrast FA31mm f/1.8 has nine elements in it and it has 96% transmittance at the focal plane, no I suspect the aperture mechanism is to blame.

---

Last edited by Digitalis; 05-10-2010 at 06:16 AM.

[Lowell Goudge]

Regardless of cause, and I doubt coatings because the SMC coatings in the late 70's upped the transmittance from about 98% to 99.9% therefore there is not a lot of room left, and the 2% gain in the 70's was already much much less than one stop,

I have found some lenses tend to drift towards overexposure (almost linearlly with aperture) as you stop down. My Tamron 28-75 F2.8 is like this, with exposure rising about 3/4 of a stop from wide open to F11 and then remaining stable at +3/4 stop from F11 onward.

You should as a practice, map out the exposure of your lens across the entire range of apertures and for each body you have. I do this by shooting a uniform and uniformly lit surface at each aperture the lens supports and plotting average greyscale value for the center 10% of the frame against aperture. It is part of understandig your equipment.

Also note that in the digital world, we do not have things that are continuously variable, but which operate in steps. Depending on the exact light level which is analog, you have only limited exposure steps of either 1/3 or 1/2 a stop in your control, and as a result you could be below a threshold at F8 but due to very minor errors in aperture control or calculation be above the threshold at F11 and see a bigger than expected swing in exposure.

[pcarfan]

Thank you for all the responses. I wanted to gather all the variables involved in this discrepancy and the responses have more than answered them for me. Thanks!

P.S: The lens didn't have a filter and the light was consistent

[Lowell Goudge]

Thank you for all the responses. I wanted to gather all the variables involved in this discrepancy and the responses have more than answered them for me. Thanks!

P.S: The lens didn't have a filter and the light was consistent

so now your next task is to calibrate/document the exposure error by shooting a block wall and adding it to a lens review

[Blue]

just a thought:

light will pass through ~100% through a pipe's hole.

put a clear glass to that hole and light maybe will pass ~98%.

then let's put some coatings to that glass, myabe only ~96% will pass.

light passes through 14 coated glass elements (assuming they're all coated?) with the FA*28-70/2.8, and only 9 through the FA*31.


Although i'm just trying to guess here hehe

Take a look at this RolloR. It discusses this very thing, only SMC helps increase light transmission to the film or sensor plane.

About Super-Multi Coating (SMC)

Pentax has added SP (Super Protect) and Aero Bright to SMC and Ghostless Coating line up.

[Doanh]

Hi pcarfan,

I know this is a 10 months old post but I want to let you know that I notice the same with my 31 ltd. It is brighter than most of the lenses that I have (about 25 of them) but not to the point of overexposure. I was surprised at first because I thought the lens had sticky aperture blades but that's not the case. Other lens that I found having similar "character" (not issue) is the Tamron 28-75 f2.8 and Voigtlander 58 f1.4. Pictures taken with them are brighter than normal. Photozone confirms the Voigtlander overexposes by approx 1/3 to 2/3 stop. Anyway I have no issue with any of them.

Regards,

Peter

[Nick Siebers]

Ha! pcarfan's 31 is now in my hands... I'll have to test it out. :-)

[Doanh]

I love the 31 ltd. It stays on my camera 50% of the time now. It's crazy sharp! There's a photo that I cropped almost 100% and printed 8x10 and it still look great.

[sewebster]

This is probably irrelevant, but the f/stop is a number based only on the size of the aperture. If you want to take into account how much light actually goes through you want to use the t stop. So, for the same camera settings (iso, shutter speed, aperture) different lenses with different light transmission will expose differently. Though perhaps 1 stop is a bit much.

Apparently some cameras "fudge" the results at large (small f/#) apertures though.

[Digitalis]

Apparently some cameras "fudge" the results at large (small f/#) apertures though.

The funny thing is that exposure meters that are incapable of taking such optical foibles into account are still quite accurate when the exposure is made. For critical work I always look up my lens test data especially where vignetting is a concern, I have witnessed a photographer using a studio white background who forgot to take her wide angle lens's vignetting into account. So she ended up with images with dark corners because she didn't boost the lights on the background enough to eliminate it.