HerbChong wrote:
yet some people really do buy a fast lens specifically to get higher shutter speeds wide open and that is the most important thing with everything else except center sharpness secondary. i shoot under those conditions several months of the year. when i am shooting that way, i don't care about light falloff or bokeh because if i can't get my shutter speeds up while holding ISO to limit noise, i don't get the shot.
Really? So what did you do three years ago when you couldn't shoot above 1600 cleanly? What did you do when you shot film, which wasn't always properly labeled in terms of film speed anyway?
HerbChong wrote:
some sports shooters have similar requirements. they shoot indoors at the high ISO edge and will choose the highest ISO possible commensurate with their tolerance for sensor noise to stop fast action. having the camera increase the ISO on them as they open up increases the noise. good bokeh and low corner falloff are nice to haves.
Herb...
Come on. No one shooting sports shoots wider than f2, and that's a rarity. If you showed most successful sports shooters this thread they'd probably just laugh.
Michael Gordon wrote:
(...)
See this link---they referred to "ghosting" mounting OM lenses on the EP-1 that I don't think I see on my OM's on film or 5DII, though I don't have a MFT body yet to directly compare. I'll defer to the formidible optics experts on this list as to the precise explanation. http://www.biofos.com/cornucop/omz_ep1.html
Hi Mike,
I checked the link and biofos is quite vague about the problem he run into: This occurs with all E-System cameras I've used when using most fast OM Zuiko lenses wide open and is to do with light being reflected back and forth between the camera sensor and the rear element of the lens. The larger the rear element, its curvature and distance from the sensor, the greater can be the problem. Here is an example...
(He also mentions he is speculating on the reason.)
Anyway, the rear lens element size, its curvature and the distance from the sensor should all be the same whatever camera you mount the lens to. Reflections bouncing around... shouldn't the risk for bouncing be bigger with a bigger sensor? The glass surface of the sensor should reflect light better than the matte black areas in the mirror box of the smaller cameras.
I don't recall it as the very fast lenses flared more on 4/3 sensors than on FF sensors. Or maybe they did - but I think the differences are down to the pixel pitch. The smaller sensors are more demanding. My G1 for example has a pixel density corresponding to a 45MP FF sensor.
Maybe we never will get the complete truth on these problems. But for now I'm not sure there is a "mirror box problem". Maybe there is. If you some day get the chance to ask Ken about the details it would be very interesting to hear about his thoughts, more in depth.
In my experience, both the Oly 50/1.2 and 55/1.2 exhibit noticeable ghosting, or Mandel-like 'glow', when shot wide open in good light, on FF EOS cameras like the 1DsIII and 5D. On the 50/1.2, it's not a big issue, but it's very significant on the 55/1.2. I like my fast lenses to be sharp wide open (within the many caveats that are appropriate), and so I'm selling the Oly 50/1.2, which I bought from proceeds of selling the Oly 55/1.2. I'll stick with my trusty old SMCP-A 50/1.4 for 'alternative' applications for now, until I score an -A 50/1.2. I look forward to using the EF 50/1.2L that I ordered yesterday.
Jonas B wrote:
I don't recall it as the very fast lenses flared more on 4/3 sensors than on FF sensors. Or maybe they did - but I think the differences are down to the pixel pitch. The smaller sensors are more demanding. My G1 for example has a pixel density corresponding to a 45MP FF sensor.
Maybe we never will get the complete truth on these problems. But for now I'm not sure there is a "mirror box problem". Maybe there is. If you some day get the chance to ask Ken about the details it would be very interesting to hear about his thoughts, more in depth.
i believe you are correct here. the high pixel densities of 4/3 sensors magnify the spherical aberration inherent in those old lens designs. people see that and don't notice it on film or FF frame due to the lower resolution, so they've manufactured a few different explanations many of which are a bit ridiculous.
as far as the mirror box calculation goes, i believe that is assuming a symmetric 50mm lens on 4/3 (not µ4/3). obviously different designs regarding telecentricity and rear element size/location would change that calculation. on a related note: i have noticed a number of different cameras have mirror boxes that will clip the oof highlights of f/1.2 lenses though (including the 5D). it seems to me that this implies that those cameras must not be getting the full dof blurring effect of those lenses (or light gathering) at f/1.2 since some off angle light is obviously blocked by the mirror box?
jcolwell wrote:
In my experience, both the Oly 50/1.2 and 55/1.2 exhibit noticeable ghosting, or Mandel-like 'glow', when shot wide open in good light, on FF EOS cameras like the 1DsIII and 5D. On the 50/1.2, it's not a big issue, but it's very significant on the 55/1.2. I like my fast lenses to be sharp wide open (within the many caveats that are appropriate), and so I'm selling the Oly 50/1.2, which I bought from proceeds of selling the Oly 55/1.2. I'll stick with my trusty old SMCP-A 50/1.4 for 'alternative' applications for now, until I score an -A 50/1.2. I look forward to using the EF 50/1.2L that I ordered yesterday. ...Show more →
Of the 10 or 12 normals I have, and have tested, that are 1.4 or 1.2, they all do this to some degree. It's only the width and intensity of the "glow" that varies. But under less intense lighting conditions I can get all of them to clean up their acts. You may have seen me document a recent case here: https://www.fredmiranda.com/forum/topic/962872 . And because every lens I've owned and tested have done this (to some degree) I've always attributed photos that suffer from the affect to be the error of the photographer in not knowing how to use the lens(s).
I'm on a m4/3 so there may be some other differences (especially in light of what's been said in this thread) but I'm of the opinion that all super bright lenses do this. I can remember seeing this happen a lot on film too BTW so I'm not even sure it's only a digital thing either. Long super bright lenses do it too BTW. There's more of it in a 135mm f/1.8 for example than there is in a 135/3.5 - no matter about the other attributes of the lenses in question.
Bifurcator wrote:
Of the 10 or 12 normals I have, and have tested, that are 1.4 or 1.2, they all do this to some degree. It's only the width and intensity of the "glow" that varies. But under less intense lighting conditions I can get all of them to clean up their acts. You may have seen me document a recent case here: https://www.fredmiranda.com/forum/topic/962872 . And because every lens I've owned and tested have done this (to some degree) I've always attributed photos that suffer from the affect to be the error of the photographer in not knowing how to use the lens(s).
I'm on a m4/3 so there may be some other differences (especially in light of what's been said in this thread) but I'm of the opinion that all super bright lenses do this. I can remember seeing this happen a lot on film too BTW so I'm not even sure it's only a digital thing either. Long super bright lenses do it too BTW. There's more of it in a 135mm f/1.8 for example than there is in a 135/3.5 - no matter about the other attributes of the lenses in question. ...Show more →
yes, they all do it to some degree, however modern aspherical designs will minimize it to the point where it isn't really noticeable (especially on FF). if you use the new sigma 50/1.4, the panaleica 25/1.4 on 4/3, or (if you want to go all out) the leica 50/1.4 asph you probably would not be able to detect it without purposely shooting to bring it out (i'm basing this on sample crops i've seen from these lenses - i haven't shot with any of them).
of all the lenses i do own i have to say that, like jcolwell, i find that my smc 50/1.4 performs the best in this regard. i have to really try to bring out any ghosting shooting on my NEX. i haven't tried it on my 4/3 camera, but i expect it would be more noticeable there. the pentax may also be the sharpest lens i have at f/1.4 (but it would be a close fight with my pen f 42/1.2), shame the focus ring and aperture ring turn the wrong way and i like the rokkor 50/1.4's bokeh better. otherwise it might be the perfect f/1.4 lens.
It's slightly down to the sensor toppings also. You have an angle-dependent amount of internal "scatter & bounce" in both the birefringents and the IR-cut glass. At 90º incident angle (0º optical angle) the amount of internal reflection is low. At 5º it's still very low - but at 15º it skyrockets (well.... relatively).
So that haze might also be the sensor's fault to some degree, at least the spread diameter.
One of the damning facts of optics is that light energy is linear, and our vision system is (fairly) logarithmic. So if you dampen a defect or aberration from original 1:100 to 1:1000 you might feel like a hero, but just 2Ev more of scene DR and 1Ev of shadow "pull" in PP and you're back where you started... You just increased the demands on the optics by 2^3 = from 100:1 to 800:1...
But there does exist a few very fast lenses where the SA line beds back by a lot, and this usually happens in the last 1/3 Ev before "full open". This is often older lenses, where absolute brightness was more important in many cases than absolute contrast. No-one expected an ISO800 film pushed to ISO1600 with a F/1.2 lens at 1/40s to give a sharp 15MP image back then. If you could get a 5x8 print you were reasonably pleased.
David Baldwin wrote:
Anyone know if the Canon 5D2 performs such compensation, if so by how much? I am using mine with a 24L mk2, usually at infinity and f2.
I don't have the same equipment. But, you should be able to find out by yourself pretty easy.
Tripod, your settings, take a shot. Then press the DOF button and keep it pressed while un-mounting the lens. Check the aperture blades are stil closed to f/2. Tape the connectors. Mount the lens and take a second image. Contemplate the difference, if any. Post about your findings.
.....and if you don't want to get your hands dirty with specialized measuring softwares, LR or aperture or Photoshop works just fine.
Take the testshots in manual focus, lens set at infinity. Make sure you have some very large portion of the picture in a uniform gray. Easiest way (and with the least risk for casualties) :-) is to place camera on a tabletop, facing a blank wall. Manual shutterspeed.
Shot 1: with electrical contacts, maximum aperture.
Shot 2: just very slightly loosen the lens by releasing the lock and turning it just over a mm or so. It will lose electrical contact. Same shutterspeed.
Compare the two shots in your raw-converter, and try to get the histograms to match. Your used +Ev amount on "shot 2" is the amount of digital after-correction applied in the camera.
Bif:
SA is short for spherical aberration, aka "focus shift". The rays from the aperture rim @F/8 will in most cases not focus at the same distance as the rays from the aperture rim @F/2 and F/1.4. The sum of all light passing through the aperture has some kind of "mean focus distance" where you have your plane of best focus. A very sharp lens (at maximum aperture) will have very similar focus F/8-F/5.6-F/4-F/2.8-F/2-F/1.4. Which in itself is self-explanatory, if all light is concentrated to the same point, you have a sharp lens. If F/4 and F/2 light isn't focused at the same distance, then you get blur. Only one can be critically sharp at any time, but both contributes light to the projection at all times (unless you stop down, thereby cutting off the F/2 light...)
Older lenses usually use second order correction, meaning that you get a curve that gets more and more undercorrected the closer you get to maximum aperture. Then the correction kicks in and often overcompensates at maximum aperture. This gives bright background bokeh outlines
Bifurcator wrote:
Of the 10 or 12 normals I have, and have tested, that are 1.4 or 1.2, they all do this to some degree. It's only the width and intensity of the "glow" that varies. But under less intense lighting conditions I can get all of them to clean up their acts.
It's noticeable on most of the fast 50's that I've owned, including SMC Pentax 50/1.2, EF 50/1.4 USM, Super-Takumar 50/1.4, SMC Pentax-M 50/1.4, and SMC Pentax-A 50/1.4, but not as much as the Oly 55/1.2.
Edit: change "it's not noticeable" to "It's noticeable", although it is fair to say that the Oly 55/1.2 has the most 'glow'. I guess that's what I remembered when I wrote this...
theSuede wrote:
Bif:
SA is short for spherical aberration, aka "focus shift". The rays from the aperture rim @F/8 will in most cases not focus at the same distance as the rays from the aperture rim @F/2 and F/1.4. The sum of all light passing through the aperture has some kind of "mean focus distance" where you have your plane of best focus. A very sharp lens (at maximum aperture) will have very similar focus F/8-F/5.6-F/4-F/2.8-F/2-F/1.4. Which in itself is self-explanatory, if all light is concentrated to the same point, you have a sharp lens. If F/4 and F/2 light isn't focused at the same distance, then you get blur. Only one can be critically sharp at any time, but both contributes light to the projection at all times (unless you stop down, thereby cutting off the F/2 light...)
Older lenses usually use second order correction, meaning that you get a curve that gets more and more undercorrected the closer you get to maximum aperture. Then the correction kicks in and often overcompensates at maximum aperture. This gives bright background bokeh outlines...Show more →
Ah, thanks man! Much appreciated! Nice explanation too!
--
jcolwell,
You're probably just a better photographer is all.
jcolwell wrote:
It's not noticeable on most of the fast 50's that I've owned, including SMC Pentax 50/1.2, EF 50/1.4 USM, Super-Takumar 50/1.4, SMC Pentax-M 50/1.4, and SMC Pentax-A 50/1.4, nor on any of my fast short teles: EF 85/1.2 L II, Zeiss Planar T* 85/1.4 MM, Rokinon 85/1.4, Summicron-R 90/2, and Nikkor 105/1.8 AIS.
Hmmm.... Of the ones you listed, I have noticed it on EF 50/1.4, Super-Tak 50/1.4, and Planar T* 85/1.4. The Super-Tak is almost famous for its wide-open "glow"...
jcolwell wrote:
It's not noticeable on most of the fast 50's that I've owned, including SMC Pentax 50/1.2, EF 50/1.4 USM, Super-Takumar 50/1.4, SMC Pentax-M 50/1.4, and SMC Pentax-A 50/1.4, nor on any of my fast short teles: EF 85/1.2 L II, Zeiss Planar T* 85/1.4 MM, Rokinon 85/1.4, Summicron-R 90/2, and Nikkor 105/1.8 AIS.
I have owned all the 50mmlenses you list. They all had some "glow" or veiling flare or whatever one may call it.
It is very easy to see something happening with the overall contrast and around small high contrast details when one compare an image taken wide opne with, say an f/2 image taken at the same time of the same target.
Are you thinking of something else?
i didn't try for those shots. they didn't exist as far as i was concerned. there are sports where a 200/2 is the right lens, and so is the 85/1.4. if you can't imagine that, i can't help you.
Herb...
cgiff wrote:
Really? So what did you do three years ago when you couldn't shoot above 1600 cleanly? What did you do when you shot film, which wasn't always properly labeled in terms of film speed anyway?
Come on. No one shooting sports shoots wider than f2, and that's a rarity. If you showed most successful sports shooters this thread they'd probably just laugh.
