50L still worth the premium?

snake0ape wrote in post #16966761
Yeah you are right about photons. Even I knew this in high school from the seventies. But you can laugh, that is easy. It's more productive if you can offer a hypothesize on what you think might cause the drop in efficiency when the aperture becomes wide.

Under the caveat that I haven't researched this further, a guess is that there is a logarithmic relationship between the iris of the lens and the light that passes through it.

I just took a quick look at some of the formula used to calculate aperture sizes and that would seem to discredit this hypothesis. But I do see a legitimate explanation being tied to rounding errors in the calculation of an f-stop geometric series.

Is f/1.4 really f/1.4? No. It's actually f/(1/rad(2)). And at elements of a geometric series where the tenths significant digit represents such a substantial difference in precision between elements of the geometric series (f/1 and f/1.4 versus f/22 and f/32), I would hypothesize that this is a more reasonable explanation (significant digit difference of 4 on the one hand, and a significant digit difference of 10 on the other hand) for why there seems to be a variation in the amount of light you are seeing collected at these apertures.

CRCchemist wrote in post #16966857
Under the caveat that I haven't researched this further, a guess is that there is a logarithmic relationship between the iris of the lens and the light that passes through it.

I just took a quick look at some of the formula used to calculate aperture sizes and that would seem to discredit this hypothesis. But I do see a legitimate explanation being tied to rounding errors in the calculation of an f-stop geometric series.

I have not re-read the article, but I think the article was about empirical data showing a loss of light capture when compared to non-digital camera (aka film) and bigger pixel sensors.. I don't think it was compared to mathematical formulas.

snake0ape wrote in post #16966866
I have not re-read the article, but I think the article was about empirical data showing a loss of light capture when compared to non-digital camera (aka film) and bigger pixel sensors.. I don't think it was compared to mathematical formulas.

Since there was so much speculation and assumption on the part of the open letter author, to the point that they actually had clear erroneous assumptions, I am going to venture a guess that he didn't consider the possibility of mathematical rounding errors contributing to his observations. Like I said above: there is a rather notable difference between early elements of the geometric series (f/1.0 to f/1.4) than there is at later elements of the geometric series (f/22 to f/32) that a rounding error of f/1.4142 to f/1.4 could be significant to explain why you have to vary the exposure a bit in your images.

That being said, I don't design lens optics. I'm in a different line of work. We need an optical engineer to weigh in on this subject, because they'll be far more familiar with the engineering principles and physics at work on this matter than I am.

CRCchemist wrote in post #16966870
Since there was so much speculation and assumption on the part of the open letter author, to the point that they actually had clear erroneous assumptions, I am going to venture a guess that he didn't consider the possibility of mathematical rounding errors contributing to his observations. Like I said above: there is a rather notable difference between early elements of the geometric series (f/1.0 to f/1.4) than there is at later elements of the geometric series (f/22 to f/32) that a rounding error of f/1.4142 to f/1.4 could be significant to explain why you have to vary the exposure a bit in your images.

That being said, I don't design lens optics. I'm in a different line of work. We need an optical engineer to weigh in on this subject, because they'll be far more familiar with the engineering principles and physics at work on this matter than I am.

the speculation the author makes is trying to explain the relationship between F stop and sensor gain, not whether it happens or not, so you can skip your rounding errors nonsense. The question is not IF there is an ISO push, there definitely is, it's why. This whole time, you've been arguing about the IF there is a sensor gain or not

^^^ Holy smoke! What are you talking abovut? ^^^

hiketheplanet wrote in post #16949278
I'm loving the 50L. The more I use it, the more I become enamored with it. You can take dreadfully dull images with it (like any lens for that matter), but give this thing the right light and it just steps up to the plate (I figure it's gotta be the lens because I'm not great at this photography thing).

It really likes contrast in lighting, like street lights against the cool dim sky just after sundown. It loves the golden hours, and it loves making giant bokeh balls in the background. It's just so damn cool what you can do with it. For my style, which is like, tinker around with the snapping the dumbest little things, it works swell.

This isn't typical gearhead talk Well described from real photog perspective!

Charlie wrote in post #16967361
the speculation the author makes is trying to explain the relationship between F stop and sensor gain, not whether it happens or not, so you can skip your rounding errors nonsense. The question is not IF there is an ISO push, there definitely is, it's why. This whole time, you've been arguing about the IF there is a sensor gain or not

Oh, then we need someone who works at Canon in the software department to explain that. I'm still not satisfied with someone trying to write an open letter that is filled with accusations based on assumptions and speculations, several of which I have identified as erroneous.

If we can't get someone from Canon Inc., how about the guys who work at Magic Lantern? Anybody around the forums here that wrote software for that project? Did they implement the same thing (ISO gain) that you're talking about the stock Canon OS doing?

That will really cinch this whole debate up. If someone who has Magic Lantern OS running on their camera observes the ISO gain just like observed on the stock Canon OS, then we have something to work with and people who are accessible to us. Then we can ask them what they did and why.

If the Magic Lantern dev team didn't do anything with ISO at large apertures, then we are in a bit of a conundrum, because that means there is another cause. But either way, the people who programmed Magic Lantern will be able to shed more light on this than anyone else... especially these people who are writing articles on the internet full of guesswork.

Nobody on here was on the Magic Lantern dev team that can weigh in on the "sensor gain" issue?

Maybe a user that is running Magic Lantern OS can tell us if they see the "ISO push" that Charlie mentioned?

I think you guys officially killed this thread.

mystik610 wrote in post #16969414
I think you guys officially killed this thread.

Maybe someone can start a new thread for this iso push topic.

Back on topic, the used market price of the 50L seems to be not too much more than the art lens. So on a resale standpoint, the 50L will probally hold more of its value than a used 50art by next year.

I got the 50mm f/1.2L some 8 years ago after having the 85mm f/1.2L for a while. What I had wanted was a lens similar to the 85mm, but with a wider view on my 1Ds Mk II. The 50mm did not quite do that on a full-frame camera in that the color rendering at low apertures on the 85mm seemed "painterly," whereas the 50mm not as much, but some. Now, on a FoVCF = 1.6 camera like the 7D or a 70D, the 50mm renders colors very nicely at wide aperture, "neo-painterly," I would describe it as, but with approximately the same view as the 85mm on a full-frame camera.

I agree with some others in this thread about the 50mm f/1.4. I did not like it's rendition at all, "flat" and lifeless. I actually prefer the 50mm f/1.8 II over the 50mm f/1.4. Also my f/1.4 lens fell apart rather easily, was not happy with the build quality.

The 50mm f/1.2L is in a different class, and of course is much more expensive. For really dim natural lighting in restaurants or clubs for fotos of belly dance, that lens has color and patina that is unrivaled to my eyes. It's certainly not sharp at low aperture as others have described the 50mm Art; but it does render dark scenes in a bright velvety color rendition. On a 7D/70D camera, it's the perfect thing for low light club fotos of fast-moving dancers at medium working distances, say 2 meters and farther.

CRCchemist wrote in post #16951664
It isn't! It's because these people are talking about stuff that doesn't exist. LOL. I can't believe anyone is buying all this.

Mounting your lens and then dismounting it slightly so all the contacts on the lens are touching the contact on the body that is directly next to correct contact that the lens should be making an electrical connection with. Then claiming that because you've screwed with the electronic communication between the body and the lens that you've stumbled upon a conspiracy by Canon to secretly boost the ISO in your image to hide a manufacturing defect. You people are hilarious.

Why don't you mount a wide aperture FE lens with an adapter and let's see if you can get the same light collecting ability as the EF lens, and what the noise looks like then. You silly people.

CRCchemist wrote in post #16951703
As a scientist, I can't get into methodological problems with what they do, because I'm not a part of the study these photographers are doing. But I suspect that if they actually had a peer-reviewed researcher over their shoulder watching variables and closely scrutinizing their testing setup, they would probably react the same way that 'ole Lens Pirate did:

Missing punctuation and all. Haha

CRCchemist wrote in post #16966523
I finally got around to reading about these silly claims all of you have been touting, based on this open letter someone wrote:

http://www.luminous-landscape.com …amera_manufactu​rers.shtml

I ran it by a colleague and we both agree. Totally bogus conclusion. That's probably why nobody has bothered to answer him... because there's nothing to answer. Right there, in the 4th paragraph of The Data section, this guy makes two huge erroneous assumptions that the rest of his argument is predicated on: the fundamental structure of a photodiode, and the particle-like properties of a photon. Both are wrong. And both gave us a good laugh reading a layman write like he knew this subject. I already explained to all of you why photodiode sites aren't deposited and etched the way some of you think they are. And as to the particle and wave properties of photons. They aren't like little specks flying through space. They are both waves and particles, and how they propagate, absorb, reflect, change energy state, etc. is as complicated as characterizing electrons. This guy bases his whole argument on "a light ray [haha again, photons are both particles and waves] that is not parallel [did we just forget they can characterize as waves] to the tube hits the photo site, chances are the light ray will not get to the bottom [hey dude! it acts like a wave] of the tube and will not hit the sensing element." Someone needs to buy this guy a double-slit interfometer so he can get a refresh of his high school physics and mathematics classes and stop promulgating confusing science to people who read his open letters on the internet.

CRCchemist wrote in post #16966857
Under the caveat that I haven't researched this further, a guess is that there is a logarithmic relationship between the iris of the lens and the light that passes through it.

I just took a quick look at some of the formula used to calculate aperture sizes and that would seem to discredit this hypothesis. But I do see a legitimate explanation being tied to rounding errors in the calculation of an f-stop geometric series.

Is f/1.4 really f/1.4? No. It's actually f/(1/rad(2)). And at elements of a geometric series where the tenths significant digit represents such a substantial difference in precision between elements of the geometric series (f/1 and f/1.4 versus f/22 and f/32), I would hypothesize that this is a more reasonable explanation (significant digit difference of 4 on the one hand, and a significant digit difference of 10 on the other hand) for why there seems to be a variation in the amount of light you are seeing collected at these apertures.

Charlie wrote in post #16967361
the speculation the author makes is trying to explain the relationship between F stop and sensor gain, not whether it happens or not, so you can skip your rounding errors nonsense. The question is not IF there is an ISO push, there definitely is, it's why. This whole time, you've been arguing about the IF there is a sensor gain or not

Certain posts in this thread read like an episode of Big Bang Theory crossed with Beavis and Butt-head....