Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
10-12-2016, 07:42 AM   #46
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 992
QuoteOriginally posted by Digitalis Quote

Pentax K5IIs - Aperture and shutter speeds were identical for both images, the only thing that changed was the lens aperture.
Interesting!

Yet isn't the second image much less bright than it should be? In theory, f/1.2 delivers 2.8X more light (1.5 stops) to the sensor. Yet the median intensity only rose 31% (the change in average intensity isn't as meaningful due to clipping). Unless there's some serious gamma or log mapping of the intensities, wouldn't the histogram show a much larger shift to the right for the f/1.2 image?

10-12-2016, 08:32 AM   #47
dms
Site Supporter




Join Date: Aug 2011
Location: New York, NY
Photos: Gallery
Posts: 1,607
I sampled (quick and dirty) same areas in digitalis's images w/ PS CS6 (in Bridge) and see about 1 to 1.5 stops between the two images.
10-12-2016, 09:01 AM   #48
New Member




Join Date: Feb 2012
Location: Brisbane Australia
Photos: Gallery
Posts: 21
There is only one f1.2 AF on the market at the moment it is the canon 50mm f1.2 the only other modern f1.2 lenses are all mf lenses from zeiss, samyang.
10-12-2016, 09:59 AM   #49
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,319
QuoteOriginally posted by andymuns Quote
There is only one f1.2 AF on the market at the moment it is the canon 50mm f1.2 .
Canon 85mm f1.2.





10-12-2016, 12:22 PM   #50
Site Supporter




Join Date: May 2016
Posts: 461
QuoteOriginally posted by RobA_Oz Quote
The same amount of light enters the lens regardless of aperture setting, and the same losses in transmission and reflection occur in every element and surface up to the aperture stop. For this contention to be rational, it would mean that relative losses at and beyond the aperture stop would have to be increased as it opened up. Further, there would have to be an algorithm within the camera's circuitry that detected that change, and "fudged" the ISO setting to suit. This is starting to look dangerously like magical thinking, overlaid with a conspiracy theory.

The contention is that the light from wider apertures hits the sensor at more of angle, and the "pixel pitch" of the sensor doesn't allow the full amount of light to reach the bottom of the pixel wells because they have some depth. It has not thing to do with the transmission of the lens, but the sensor. I know that Leica, for example, had a significant problem with this when they were first trying to introduce a digital camera using Leica lenses. I believe it was supposed to have an effect at even f1.4. It would make sense that manufacturers would want to reduce this effect on fast lenses so it didn't seem like they were broken; you would presumably still get an advantage from the larger aperture even if it wasn't the full amount you would expect from the aperture.

On the one hand, the measurements by DXO were supposedly objective and they observed an effect for many different cameras. Also, Pentax has introduced only one lens in the digital era with an aperture greater than 1.8. On the other hand, when I tried to find the original source for this information that I had remembered from a few years ago, the only way I could find it was through the internet archive, i.e. the article is no longer available on DXO's web site.
10-12-2016, 12:40 PM   #51
Pentaxian
RobA_Oz's Avatar

Join Date: Nov 2008
Location: Tasmania, Australia
Photos: Gallery
Posts: 4,261
QuoteOriginally posted by leekil Quote
The contention is that the light from wider apertures hits the sensor at more of angle, and the "pixel pitch" of the sensor doesn't allow the full amount of light to reach the bottom of the pixel wells because they have some depth. It has not thing to do with the transmission of the lens, but the sensor. I know that Leica, for example, had a significant problem with this when they were first trying to introduce a digital camera using Leica lenses. I believe it was supposed to have an effect at even f1.4. It would make sense that manufacturers would want to reduce this effect on fast lenses so it didn't seem like they were broken; you would presumably still get an advantage from the larger aperture even if it wasn't the full amount you would expect from the aperture.

On the one hand, the measurements by DXO were supposedly objective and they observed an effect for many different cameras. Also, Pentax has introduced only one lens in the digital era with an aperture greater than 1.8. On the other hand, when I tried to find the original source for this information that I had remembered from a few years ago, the only way I could find it was through the internet archive, i.e. the article is no longer available on DXO's web site.
There's no dispute about that, but for a given lens, the angle of incidence won't change with aperture setting, just as with transmissivity and reflectivity. The only thing that will change is the relative amount of diffraction, which, of course, decreases with increase of aperture.

We may have been talking at cross-purposes, to some extent, but essentially I was saying that there is no rational basis to the belief that faster lenses have no advantage because of light losses.
10-12-2016, 12:54 PM   #52
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 992
QuoteOriginally posted by RobA_Oz Quote
There's no dispute about that, but for a given lens, the angle of incidence won't change with aperture setting, just as with transmissivity and reflectivity. The only thing that will change is the relative amount of diffraction, which, of course, decreases with increase of aperture.

We may have been talking at cross-purposes, to some extent, but essentially I was saying that there is no rational basis to the belief that faster lenses have no advantage because of light losses.
The range of angles of incidence of light to a given pixel does change with aperture. As the aperture is opened, the cone of light arriving at the sensor subtends a larger angle. Depending on the design of the pixels and any microlenses on the sensor, light from the extreme angles in this cone may be blocked or reflected off the sensor. The overall brightness and vignetting for a given large aperture lens may be different on film versus digital because film is better at capturing light coming from more extreme angles.

This issue also affects the design of wide angle lenses on short-register mirrorless cameras.
10-12-2016, 12:55 PM   #53
Site Supporter




Join Date: May 2016
Posts: 461
QuoteOriginally posted by RobA_Oz Quote
We may have been talking at cross-purposes, to some extent, but essentially I was saying that there is no rational basis to the belief that faster lenses have no advantage because of light losses.
I agree with this. But the original DXO contention also agrees with that; they were saying you didn't get the full stop's-worth of light advantage at the faster stops, not that there was no advantage. (Though you may not be saying that same thing, depending on the specificity of your wording.)

I also personally don't believe that the DXO claim is true, given that 1) there is considerable controversy around it and 2) in the absence of outside verification of it.

10-12-2016, 01:08 PM   #54
Pentaxian
RobA_Oz's Avatar

Join Date: Nov 2008
Location: Tasmania, Australia
Photos: Gallery
Posts: 4,261
QuoteOriginally posted by photoptimist Quote
The range of angles of incidence of light to a given pixel does change with aperture. As the aperture is opened, the cone of light arriving at the sensor subtends a larger angle. Depending on the design of the pixels and any microlenses on the sensor, light from the extreme angles in this cone may be blocked or reflected off the sensor. The overall brightness and vignetting for a given large aperture lens may be different on film versus digital because film is better at capturing light coming from more extreme angles.

This issue also affects the design of wide angle lenses on short-register mirrorless cameras.
Yes, you're correct in this. I was thinking of single-ray paraxial theory and allowed myself to be misled on that, so I stand corrected on that matter.
10-12-2016, 01:10 PM   #55
Site Supporter




Join Date: May 2016
Posts: 461
QuoteOriginally posted by Digitalis Quote
The thing is the SMCP-K 50mm f/1.2 has no electronic coupling therefore: the imaging pipeline* in Pentax cameras has no flipping idea what the lens is doing. Canon DSLRs where lenses do communicate with the camera, it has been discovered the imaging processing pipeline has been found to be compensating for losses (especially at the periphery of the imaging circle) with fast f/1.2~f/1.4 lenses. With the lens coding blacked out**: the Leica Monochrom likewise does not show any significant alterations when used with the Noctiluix 50mm f/0.95 or 50mm f/1.0



I agree. Arguing with someone who has arrived to an irrational conclusion with rational arguments is fruitless.

*sensor, AD converters, PRIME processing et al
** Black and white marks on the lens which are read by a code reader on the lens mount. There is only a slight difference in the amount of vignetting with these two lenses with lens corrections enabled. With a whopping -4.8EV of vignetting in the extreme corners on the 50mm f/0.95 I suppose Leica engineers threw their hands up in exasperation and let the lens just be what it is.

Regarding "irrational conclusions", if you are talking about me... you cited two examples of the situation under discussion. In the Canon case, there is special processing going on with the larger apertures. The Leica case involves a specialized sensor which may have different characteristics than a regular sensor. I won't be so rude as to say you are irrational, but I don't think your evidence-to-conclusion process is particularly airtight in this case.

Also, regarding the Pentax case -- you overlooked the SMCP-A 1.2 lens, which would reveal the aperture to the camera.
10-12-2016, 10:12 PM   #56
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,643
QuoteOriginally posted by photoptimist Quote
Yet isn't the second image much less bright than it should be? In theory
In theory, the second image should be brighter. But due to optical losses in transmission & internal reflections the image intensity is lower than it ought to be* however there is still appreciable difference in both DOF and light transmission, which throws the whole discussion into a grey area. "Pentax claims that SMC: is a remarkable seven-layer lens coating process that cuts the reflection ratio down to just 0.2% per lens surface**" the SMCP-K 50mm f/1.2 has 7 elements in 6 groups, so that is basically 12 air-glass interfaces where 0.2% of light is being lost.

*though it is brighter than it should be considering the hypothesis presented by DxO.
** quoted from About Super-Multi Coating (SMC), original publication published. ca.1971
10-14-2016, 11:41 AM   #57
Loyal Site Supporter




Join Date: Mar 2015
Photos: Gallery | Albums
Posts: 5,485
QuoteOriginally posted by photoptimist Quote
This issue also affects the design of wide angle lenses on short-register mirrorless cameras.
Somewhat off topic, I wonder if it also makes life harder for designers of shake reduction systems.
10-14-2016, 12:47 PM   #58
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 992
QuoteOriginally posted by pathdoc Quote
Somewhat off topic, I wonder if it also makes life harder for designers of shake reduction systems.
Perhaps a little bit. The 1.5 mm typical span of an SR system is small relative to the 12 mm x 18 mm of the sensor itself. At its extremes of travel, the corners of the sensor would be intercepting slightly more oblique rays but not significantly more oblique rays.

I'd think the bigger issue is lack of control of aberrations on a larger image circle. Optical design of a bright lens often entails compensation of aberrations created by compensation of aberrations created by compensation of aberrations ...... It's all delicately massaged to work within the goals for resolution across the image circle but that's often at the expense of what happens just outside the image circle. Thus, one or two corners of the image might be horrible if the SR system happens to have pushed the sensor beyond the image circle for 35mm film.
10-14-2016, 02:14 PM   #59
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 992
QuoteOriginally posted by Digitalis Quote
In theory, the second image should be brighter. But due to optical losses in transmission & internal reflections the image intensity is lower than it ought to be* however there is still appreciable difference in both DOF and light transmission, which throws the whole discussion into a grey area. "Pentax claims that SMC: is a remarkable seven-layer lens coating process that cuts the reflection ratio down to just 0.2% per lens surface**" the SMCP-K 50mm f/1.2 has 7 elements in 6 groups, so that is basically 12 air-glass interfaces where 0.2% of light is being lost.

*though it is brighter than it should be considering the hypothesis presented by DxO.
** quoted from About Super-Multi Coating (SMC), original publication published. ca.1971
Good points! Yet those low reflection ratios are usually only for rays perpendicular to the glass. Large aperture lenses (especially wide-angle ones) often have fairly broad cones of light passing through each surface and the oblique rays may suffer greater reflection and sometimes slight color shifts.

All of these effects will modulate vignetting and the gradation of brightness across each bokeh "circle".
10-14-2016, 05:05 PM   #60
Pentaxian
Digitalis's Avatar

Join Date: Mar 2009
Location: Adelaide.
Posts: 8,643
QuoteOriginally posted by leekil Quote
The contention is that the light from wider apertures hits the sensor at more of angle, and the "pixel pitch" of the sensor doesn't allow the full amount of light to reach the bottom of the pixel wells because they have some depth. It has not thing to do with the transmission of the lens, but the sensor.
Hence the use of microlenses on sensors, these give individual photodetectors the ability to capture photons with a greater angle of incidence than what would be possible without.

QuoteOriginally posted by leekil Quote
. I know that Leica, for example, had a significant problem with this when they were first trying to introduce a digital camera using Leica lenses. I believe it was supposed to have an effect at even f1.4.
It was actually the proximity of the lenses exit pupil to the sensor that was causing excessive vignetting that was the problem. Leica solved this issue with a custom designed microlens array that reduced the severity of this issue.



Leica also pioneered the use of offset microlenses at the periphery of the sensor to compensate for light loss at the edges of the imaging circle:


Last edited by Digitalis; 10-14-2016 at 05:11 PM.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
50mm, auto, camera, f0.95, f1.2, f1.4, focus, importance, lens, lens for pentax, pentax, pentax help, photography, recompose, technique, troubleshooting
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Are there any underwater housings for Pentax still available? Paul the Sunman Pentax DSLR Discussion 4 06-27-2016 04:44 PM
Are there any 70-200mm 2.8 with silent auto focus? enyceckk Troubleshooting and Beginner Help 12 01-11-2016 05:08 PM
Are there any hoods on the market for the 50-135 that are cheaper than the Pentax? maxxxx Pentax SLR Lens Discussion 13 03-15-2015 10:44 PM
Are there any single element K-Mount lenses? TropicalMonkey Pentax SLR Lens Discussion 31 06-28-2013 06:34 PM
Is there any lens faster than F1.2 that will fit a Pentax? Christopher M.W.T Pentax SLR Lens Discussion 29 09-30-2010 12:20 AM



All times are GMT -7. The time now is 10:36 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top