[bobmaxja]

Need to be untouch image , no PP at all in camera or out of camera

[Just1MoreDave]

The underlying problem is that a lens does not create noise nor does it affect noise in the light that passes through it.

Yes, this is true. But with different lenses, different stuff happens to the light.

I have two hypothetical situations:

Say you have a lens that's not very efficient at transmitting light, so its T-stop rating is a couple of stops below its f-stop. If you compare this lens to a lens with much better transmittance, at the same focal length, the second lens will produce a brighter image. So if you're shooting in the real world, in a situation where you want a certain aperture setting to control depth of field, and a certain shutter speed for motion or shake reduction, the only way to get a similar histogram for both lenses is to raise ISO and thus noise.

Or say a lens has a lot of vignetting in the corners. Compared to a second lens with no vignetting, again there's a brightness difference.

I don't know if the signal to noise ratio has actually changed in either example, or the noise simply becomes more apparent when any attempt is made to increase brightness. But when someone is using the two lenses, they are going to see either brightness or noise differences.

[newarts]

Yes, this is true. But with different lenses, different stuff happens to the light.

I have two hypothetical situations:

Say you have a lens that's not very efficient at transmitting light, so its T-stop rating is a couple of stops below its f-stop. If you compare this lens to a lens with much better transmittance, at the same focal length, the second lens will produce a brighter image. So if you're shooting in the real world, in a situation where you want a certain aperture setting to control depth of field, and a certain shutter speed for motion or shake reduction, the only way to get a similar histogram for both lenses is to raise ISO and thus noise.

Or say a lens has a lot of vignetting in the corners. Compared to a second lens with no vignetting, again there's a brightness difference.

I don't know if the signal to noise ratio has actually changed in either example, or the noise simply becomes more apparent when any attempt is made to increase brightness. But when someone is using the two lenses, they are going to see either brightness or noise differences.

The first example makes sense but, I think what people have been saying here is that better lenses make better images (from a noise standpoint) at constant ISO.

I'm interpreting this to mean that for the same iso, scene, and exposure, one lens gives an image where the noise is less objectionable than for the other lens.

The more I think about it the less surprised I'll be when I find an image with more noise that looks better than an image of the same scene with less noise (and fuzzy edges and low contrast.)

[loco]

Well, I had a chance to do a quick test tonight. I actually took shots with all four of my lenses but am only posting the results with the DA 35mm/f2.4 and the DA L 18-55 kit lens because they are the only two I could compare at 35mm.

Personally, I have come to the conclusion that the phenomenon of less noise with certain lenses was all in my head! Now, it may be that shooting in bright sunlight would produce a different result. However, photographing my spice/cereal cabinet inside under tungsten light seems to indicate that there really isn't any difference in noise level at either ISO 1600 or ISO 2500. If anything, the kit lens might actually be slightly better in that regard to my eyes.

I used live view so I could autofocus reliably. The camera was on a tripod. The first photo is just a small resize of the full scene. Focus point was on the Hellman's mayo.


The next two are 100% crops from the lower left area of the capture. I did not do any post processing, shot RAW, turned sharpening off completely in Lightroom and absolutely no noise reduction used. I am only posting the ISO2500 because the ISO1600 didn't show any real difference either.

DA 35mm/2.4 ISO 2500 F2.5 1/100


DA L 18-55 at 35mm ISO 2500 F4.5 1/25



The other two lenses (A 50/1.7 and DA L 55-300 at 50mm and 55mm respectively) appeared to show a bit more noise. However, I attributed this to the longer focal length. Maybe that's incorrect? If anyone wants to see those crops I can post them, but I didn't know if they were relevant to this discussion.

Anyway, if there are any differences in noise level in these photos, they appear very minor to me. Not sure what to make of this. Perhaps it was some placebo effect or the thrill of getting a new lens? Or maybe certain other conditions bring out what I was seeing more than this particular one.

But it was an interesting discussion and if you have any more thoughts, I'd love to read them! Thanks all!

[newarts]

Thank you SO much for doing the test. Your results don't put the question to rest because there's a reasonable chance that different scenes might give different results. But your test shows that the influence of lens on noise isn't a "slam dunk"!

I suspect that the "lens quality effect on noise" is small.

Focal length shouldn't matter either - iff exposure is the same.

[Digitalis]

I attributed this to the longer focal length. Maybe that's incorrect?

definitely incorrect, the only way to test this for certain is to stop both lenses down to f/8 and use a longer exposure - but many of us have already done this and can tell you that with longer exposures the biggest impact on noise levels is heat generated by the sensor - with really long exposures you can also see amplifier glow.

[loco]

Thank you SO much for doing the test. Your results don't put the question to rest because there's a reasonable chance that different scenes might give different results. But your test shows that the influence of lens on noise isn't a "slam dunk"!

I suspect that the "lens quality effect on noise" is small.

Focal length shouldn't matter either - iff exposure is the same.

Yeah, exposure wasn't the same with the 50-ish mm lenses. I used Sv mode and with the A50/1.7 exposure was F2.8 1/100 and the DA L 55-300 was F4.0 1/40. I thought perhaps the fact that the shots were "closer in" and a bit darker produced a bit more noise? (I did not move the tripod back when I switched to those lenses.)

definitely incorrect, the only way to test this for certain is to stop both lenses down to f/8 and use a longer exposure - but many of us have already done this and can tell you that with longer exposures the biggest impact on noise levels is heat generated by the sensor - with really long exposures you can also see amplifier glow.

I've seen that problem with the longer shutter introducing more noise. So, I was trying to keep it at reasonable shutter speeds.

I know it's not very scientific, but I did think the results showed that shooting the same scene with the same lighting at the same ISO resulted in very similar noise levels between these two lenses.

[Designosophy]

This is interesting. I'd like to try the same thing with my M 50/1.4 vs. DA L 18-55. And maybe again with my A 135/2.8 vs. Tamron 28-200 zoom and Sigma 70-300 zoom.

[newarts]

This is interesting. I'd like to try the same thing with my M 50/1.4 vs. DA L 18-55. And maybe again with my A 135/2.8 vs. Tamron 28-200 zoom and Sigma 70-300 zoom.

Please share results if you can. It is a difficult test to do as it is not strictly objective.

[loco]

Yep, I'd love to see the results too, Designosophy!

[Designosophy]

Well, looks like I have some work to do this weekend...