Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: fredjeang on May 06, 2010, 07:28:54 pm
-
Dear all.
I'm asking this technical question and would be gratefull if you could bring some light on that matter:
Do your lenses match your sensors?
In other words, is it reasonably truth to consider that your lenses are not limiting the sensor quality?
But what about tomorrow, when they will produce more demanding sensors to, let's say 150MP for example.
Will your lenses then be able to handle more and more resolution?
I've always liked the Phase philosophy: big heavy body with small and light lenses. But these lenses are really capable to handle
the pixel race?
Some sources are talking that to reproduce digitally the resolution of the biggest view cameras, that would be the equivalent
to a 200MP sensor. To be frank, I find these numbers a little exagerated.
What are your thoughts on this topic?
-
I wonder if photographers really need 150-200MP backs?
I think If you put a large extension ring you can see how sharp a lens is. I might actually be equivalent to 4x the resolution.
-
less megapixels, more accurate algorithms.
-
Dear all.
I'm asking this technical question and would be gratefull if you could bring some light on that matter:
Do your lenses match your sensors?
As such, the question cannot be answered until one defines what's supposed to be matched.
In other words, is it reasonably truth to consider that your lenses are not limiting the sensor quality?
But what about tomorrow, when they will produce more demanding sensors to, let's say 150MP for example.
Will your lenses then be able to handle more and more resolution?
Usually the question is narrowed down to resolution, but that's too simplistic because resolution varies with contrast, and contrast varies with spatial frequency. Enters the MTF curve.
One needs to understand that all optical components (assuming an otherwise rigid setup) have their individual MTF response, and thus contribute to the combined system MTF by multiplication of the individual component MTFs. Each component that doesn't have a 100% MTF response at a given spatial frequency will pull down system MTF.
The sensor MTF can at best approximate a Sinc curve (assuming square sensels and theoretical 100% fill factor), thus gradually declining from 100% response at low spatial frequencies, to 0% response at 2x Nyquist frequency. Lenses are not perfect, so all lenses will deteriorate the combined system MTF, even if the sensor would be 'perfect'. How much deterioration depends on the specific lens MTF and the spatial frequencies that are considered important.
The important spatial frequencies for the human visual system in turn have to do with output magnification and viewing distance, with a peak contrast sensitivity (http://www.cg.tuwien.ac.at/research/theses/matkovic/node20.html) around 8 cycles/degree.
So, we can only come to a general conclusion for 'resolution'. A higher 'resolution' sensor (assuming a theoretically perfect MTF) will help to get a system MTF thats closer to the inherent lens' MTF than a lower 'resolution' sensor, when 'resolution' is interpreted as sampling density.
Unfortunately, higher sampling densities come with other trade-offs, such as generally lower dynamic range and sensitivity, lower fill-factors, and higher cost (at large sensor array sizes).
Cheers,
Bart
-
Dear all.
Do your lenses match your sensors?
If they did not, multi-shot would be no better than single shot.
If you have an anti-aliasing filter on the sensor, a large proportion of the res of the sensor and the lens is wasted anyway.
If you want 200 Mpx, try a hex-stitch back ¿Kapture group? with a 60Mpx back on a P3.
-
If they did not, multi-shot would be no better than single shot.
If you have an anti-aliasing filter on the sensor, a large proportion of the res of the sensor and the lens is wasted anyway.
If you want 200 Mpx, try a hex-stitch back ¿Kapture group? with a 60Mpx back on a P3.
When is your H4D60 due in?
Cheers,
Bernard
-
When is your H4D60 due in?
Cheers,
Bernard
The last ETA was last month.
-
My questions came from several observations from the field and PP.
On my street camera, wich has a 12MP aps sensor, if I mount a 35mm top prime, it is like the sensor is not the same and these lenses reveal
the real nature of the equipment. In fact in PP, just working with top lenses gives me more flexibility. IQ is way better, not little, but a lot better.
The overall detail is much much present, so logically, any lens that is not top class is limiting a 12MP sensor resolution power.
Parting from that simple observation, I asked myself if there is a limitation in the other side.
Are these lenses capable of rendering properly the sensor's increased resolution that we see year after year? What's the limit?
Let's take a canon 7D, 18MP. Then the last cheap entry level 550D, 18MP, same lens on both, top quality. I guess IQ will be exactly the same.
What's not clear for me is that one day, in the On line Photographer website, he did test an EP1 with Leica M lenses. To his surprise, he did
not found any increment in IQ between the "cheap" Oly lens and the top Leica M. (Can't find the link any more but it's there).
So the logical conclusion is that the MFT sensor is limiting the real power of the lens. (or maybe an even more exotic conclusion would be that the Oly lens is as good as the Leica M?!?...won't come in that terrain)
So, what's worth investing in a very expensive lens if the
sensor's size is not capable to deserve it to it's full potencial.
And on the contrary, what's worth investing in a top MFD camera if the lenses don't follow and are not capable to deserve the sensor's capability. Do they deserve well your huge resolution sensors?
When sensors will reach more resolution in the future, do you think a new lens design will be necessary?
So my conclusion (but not sure about it) is this: Bigger possible sensor with better possible lens, thinking that the lens is maybe more important than the sensor itself.
-
As such, the question cannot be answered until one defines what's supposed to be matched.
Usually the question is narrowed down to resolution, but that's too simplistic because resolution varies with contrast, and contrast varies with spatial frequency. Enters the MTF curve.
One needs to understand that all optical components (assuming an otherwise rigid setup) have their individual MTF response, and thus contribute to the combined system MTF by multiplication of the individual component MTFs. Each component that doesn't have a 100% MTF response at a given spatial frequency will pull down system MTF.
The sensor MTF can at best approximate a Sinc curve (assuming square sensels and theoretical 100% fill factor), thus gradually declining from 100% response at low spatial frequencies, to 0% response at 2x Nyquist frequency. Lenses are not perfect, so all lenses will deteriorate the combined system MTF, even if the sensor would be 'perfect'. How much deterioration depends on the specific lens MTF and the spatial frequencies that are considered important.
The important spatial frequencies for the human visual system in turn have to do with output magnification and viewing distance, with a peak contrast sensitivity (http://www.cg.tuwien.ac.at/research/theses/matkovic/node20.html) around 8 cycles/mm.
So, we can only come to a general conclusion for 'resolution'. A higher 'resolution' sensor (assuming a theoretically perfect MTF) will help to get a system MTF thats closer to the inherent lens' MTF than a lower 'resolution' sensor, when 'resolution' is interpreted as sampling density.
Unfortunately, higher sampling densities come with other trade-offs, such as generally lower dynamic range and sensitivity, lower fill-factors, and higher cost (at large sensor array sizes).
Cheers,
Bart
Excellent summary Bart. Resolution is easy to define but difficult to explain. Someone once told me that an optical system is like a group of people pulling on a rope. You can always exert more pull by substituting a stronger person (you can gain system resolution from a sharper lens). However, that may or may not move what you are trying to pull (you may or may not see the effect).
-
I've always liked the Phase philosophy: big heavy body with small and light lenses. But these lenses are really capable to handle
the pixel race?
Just a quick technical point...
If your comparison is to lenses from Hasselblad then you should be aware that adding a leaf shutter to a lens design requires a physically larger lens. So while lenses can be larger or smaller, heavier or lighter, for various reasons (minimum focus distance, largest aperture, number of elements, flange distance / mirror-box size, front lens element design, number of elements) that a line of lenses including a leaf shutter will be larger on average than another line of lenses without leaf shutters. An excellent apples-to-apples example would the Schneider 80mm LS lens and the Phase One 80mm D lens. Both are f/2.8, both are 6 elements in 5 groups, and both have a minimum focus distance of 2.3 ft, yet the LS version is 8% longer and 52% heavier.
Now as your main question... my two cents:
My real-world results are that most of the older Mamiya manual focus and non-D lenses do fine on a P25, P21, or P20 (9 micron) but are often a slight to moderate disappointment on a pixel-sharpness basis for a P65+. Of course there are exceptions such as the 120mm MF macro lens which is a tremendous lens even it its earliest generation. There are also plenty of applications where the per-pixel sharpness is not as important as the "look" of the lens where a very cheap (few hundred dollars) lens from this series can be a great addition to your kit. For instance the 80mm f/1.9 (http://forum.getdpi.com/forum/showthread.php?t=16330) lens is getting a lot of love over on getdpi.
The Phase One D series lenses and Schneider LS series and the Schneider/Rodenstock digital series large format lenses are almost without exception fantastically sharp even on a P65+ and I strongly suspect will continue to show absolutely excellent real-world results on whatever smaller micron size comes next. In particular the Schneider 47mm XL, and the Phase One 80mmD, 120mmD, 150mmD are amongst the best lenses I've used on any platform from any manufacturer and seem like lenses that will still be considered excellent in several generations.
Doug Peterson ()
__________________
Head of Technical Services, Capture Integration
Phase One, Leaf, Cambo, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter: Read Latest or Sign Up (http://www.captureintegration.com/our-company/newsletters/)
RSS Feed: Subscribe (http://www.captureintegration.com/2008/08/11/rss-feeds/)
Buy Capture One at 10% off (http://www.captureintegration.com/phase-one/buy-capture-one/)
Personal Work (http://www.doug-peterson.com/)
-
So, what's worth investing in a very expensive lens if the
sensor's size is not capable to deserve it to it's full potencial.
And on the contrary, what's worth investing in a MFD if the lenses are not capable to deserve the sensor's capability.
When sensors will reach very high resolution, will a new lens design be necessary?
Lens MTF x Sensor MTF = Sytem MTF at a given spatial frequency.
A 'better' lens on a given sensor will improve total system MTF. So a lens improvement of 90% to 100% on a sensor response of 50% e.g. at the absolute limiting resolution (Nyquist) will boost system performance from 45% to 50%.
A 'better' sensor (denser sampling) with a given lens will improve total system MTF at a given spatial frequency. It will only get closer to the worst contributor to system performance at a given spatial frequency.
If we compare a non-existing perfect lens on a perfect sensor array between normal and double sampling density, we will increase the MTF response from 56% to 87% at the same spatial frequency, as illustrated in the following example of a 6.4 micron sensel pitch array versus a 3.2 micron sensel pitch array:
[attachment=21867:PerfectS...6_4vs3_2.png]
Anything in that perfect scenario that's less than perfect, will reduce the combined response to below the worst contributor at a given spatial frequency (78 cy/mm taken as an example, the Nyquist frequency of a 6.4 micron sensel pitch array). The denser sampling of the smaller sensel pitch array will -as a bonus- allow to also record higher spatial frequencies, if they are present (but that becomes increasingly unlikely for most lenses near the edges/corners).
Cheers,
Bart
-
Did some tests with my ixpress 528c in 16 shot mode which gives effective 88mpix with 4.5 micron spacing.... Definitely the lenses I tested (rollei 90mm apo, 150mm apo ) support that resolution with no probs but only at F/8 and wider. At smaller apertures you still see more detail than in the 22mpix images. What you loose is the fine detail but the other facets of high quality lenses are still there. Color, macro contrast, bokeh, etc.
-
Did some tests with my ixpress 528c in 16 shot mode which gives effective 88mpix with 4.5 micron spacing.... Definitely the lenses I tested (rollei 90mm apo, 150mm apo ) support that resolution with no probs but only at F/8 and wider. At smaller apertures you still see more detail than in the 22mpix images. What you loose is the fine detail but the other facets of high quality lenses are still there. Color, macro contrast, bokeh, etc.
Hi Eric,
Yes, that's what's to be expected. The denser sampling of the same lens will pull more accurate detail out of the lens, but not necessarily more detail (e.g. when it's diffraction, or residual aberration, limited).
The more accurate detail though, even if not sharper on a per pixel level, will allow to do a better (deconvolution) sharpening job.
Cheers,
Bart
-
Thank you very much for these technical repplies.
But could somebody translate this chinese language for the useless tech guys like me, into something easier?
What does all that means in practise?
I heard that the Leica S2 is allowing some R lenses. It's gona change the focal lenght but what about resolution.
Is it possible that lenses built for 35mm work well on MFD? strictly talking about IQ results, not the focal.
Pentax allows on the contrary the vintages 6x7 lenses on their cropped dslr. Never tried but ask myself the same question.
-
Just a quick technical point...
If your comparison is to lenses from Hasselblad then you should be aware that adding a leaf shutter to a lens design requires a physically larger lens. So while lenses can be larger or smaller, heavier or lighter, for various reasons (minimum focus distance, largest aperture, number of elements, flange distance / mirror-box size, front lens element design, number of elements) that a line of lenses including a leaf shutter will be larger on average than another line of lenses without leaf shutters. An excellent apples-to-apples example would the Schneider 80mm LS lens and the Phase One 80mm D lens. Both are f/2.8, both are 6 elements in 5 groups, and both have a minimum focus distance of 2.3 ft, yet the LS version is 8% longer and 52% heavier.
Now as your main question... my two cents:
My real-world results are that most of the older Mamiya manual focus and non-D lenses do fine on a P25, P21, or P20 (9 micron) but are often a slight to moderate disappointment on a pixel-sharpness basis for a P65+. Of course there are exceptions such as the 120mm MF macro lens which is a tremendous lens even it its earliest generation. There are also plenty of applications where the per-pixel sharpness is not as important as the "look" of the lens where a very cheap (few hundred dollars) lens from this series can be a great addition to your kit. For instance the 80mm f/1.9 (http://forum.getdpi.com/forum/showthread.php?t=16330) lens is getting a lot of love over on getdpi.
The Phase One D series lenses and Schneider LS series and the Schneider/Rodenstock digital series large format lenses are almost without exception fantastically sharp even on a P65+ and I strongly suspect will continue to show absolutely excellent real-world results on whatever smaller micron size comes next. In particular the Schneider 47mm XL, and the Phase One 80mmD, 120mmD, 150mmD are amongst the best lenses I've used on any platform from any manufacturer and seem like lenses that will still be considered excellent in several generations.
Doug Peterson ()
__________________
Head of Technical Services, Capture Integration
Phase One, Leaf, Cambo, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter: Read Latest or Sign Up (http://www.captureintegration.com/our-company/newsletters/)
RSS Feed: Subscribe (http://www.captureintegration.com/2008/08/11/rss-feeds/)
Buy Capture One at 10% off (http://www.captureintegration.com/phase-one/buy-capture-one/)
Personal Work (http://www.doug-peterson.com/)
Hi Doug,
Tell me if I'm wrong but I saw in Capture integration's website the Fuji GX680 III Kit, prism, 3 lenses, KG adpator and Plate at $2,990 ??
-
Hi Eric,
Yes, that's what's to be expected. The denser sampling of the same lens will pull more accurate detail out of the lens, but not necessarily more detail (e.g. when it's diffraction, or residual aberration, limited).
The more accurate detail though, even if not sharper on a per pixel level, will allow to do a better (deconvolution) sharpening job.
Cheers,
Bart
I shot a chart I downloaded from Norman Koren's imatest site with a gradient of fine line pairs. I found that at f/5.6 I could count exactly double the number of line pairs when I went from 4 shot to 16 shot modes and concluded that the lens was not the limiting factor but rather the sensor was in both cases. In other words, I am also saying that the the Rollei 90mm apo and 150mm apo lenses can do better 110 lp/mm at f/5.6 since the pixel pitch is 9micron or half that for 16 shot mode.
-
Hi Doug,
Tell me if I'm wrong but I saw in Capture integration's website the Fuji GX680 III Kit, prism, 3 lenses, KG adpator and Plate at $2,990 ??
That's correct. You can call the Atlanta office for more details (404-522-7662). I'd do it soon as the kit is purposefully priced to move off the shelf as quickly as possible so it may not be around if you wait a week or two.
http://www.captureintegration.com/2010/04/...ing-cleaning-2/ (http://www.captureintegration.com/2010/04/22/spring-cleaning-2/)
Doug Peterson ()
__________________
Head of Technical Services, Capture Integration
Phase One, Leaf, Cambo, Canon, Apple, Profoto, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter: Read Latest or Sign Up (http://www.captureintegration.com/our-company/newsletters/)
RSS Feed: Subscribe (http://www.captureintegration.com/2008/08/11/rss-feeds/)
Buy Capture One at 10% off (http://www.captureintegration.com/phase-one/buy-capture-one/)
Personal Work (http://www.doug-peterson.com/)
-
What does all that means in practise?
The sensor sets a hard limit. How close you can get to that is determined by the lens quality.
Cheers,
Bart
-
The sensor sets a hard limit. How close you can get to that is determined by the lens quality.
Cheers,
Bart
Thanks Bart.
-
The sensor sets a hard limit. How close you can get to that is determined by the lens quality.
Cheers,
Bart
But then, if so, the pixel race is totally fake.
I mean, there are pressures for more megapixels in a XY sensor area, but if I understand, we rarelly use our older sensor to their full potencial till we
can get the best possible lenses.
In that sense, the first thing we should upgrade is the lens equipment and not that much the sensor.
So instead of pixel race, what should be more appropriate is a lens quality race.
-
But then, if so, the pixel race is totally fake.
I mean, there are pressures for more megapixels in a XY sensor area, but if I understand, we rarelly use our older sensor to their full potencial till we can get the best possible lenses.
In that sense, the first thing we should upgrade is the lens equipment and not that much the sensor.
So instead of pixel race, what should be more appropriate is a lens quality race.
Yes, and No. When the sensor has higher resolution, it also gets a better sampling of whatever the lens can provide. That helps with sharpening as well. When the lens has higher resolution, especially across the image including corners, then the sensor can record a better image quality.
So it works both ways. That's why I mentioned the MTF performance in an earlier post, it's the combined performance that matters, both the sensor and the lens are a factor in the equation and the combined performance cannot exceed (it doesn't even come very close to) the worst contributor. So improving the worst component will allways help most, but even improving the better components will help some.
Lenses can have a very high resolving power (although usually less in the corners than in the center). The capability to accurately sample the blur with a higher resolution sensor array also allows to do a better job of deconvolution sharpening afterwards, effectively removing the blur (although there are limits).
Cheers,
Bart
-
Yes, and No. When the sensor has higher resolution, it also gets a better sampling of whatever the lens can provide. That helps with sharpening as well. When the lens has higher resolution, especially across the image including corners, then the sensor can record a better image quality.
So it works both ways. That's why I mentioned the MTF performance in an earlier post, it's the combined performance that matters, both the sensor and the lens are a factor in the equation and the combined performance cannot exceed (it doesn't even come very close to) the worst contributor. So improving the worst component will allways help most, but even improving the better components will help some.
Lenses can have a very high resolving power (although usually less in the corners than in the center). The capability to accurately sample the blur with a higher resolution sensor array also allows to do a better job of deconvolution sharpening afterwards, effectively removing the blur (although there are limits).
Cheers,
Bart
Bart, your points are very clear indeed.
But let's take for example the P65, wich is truth MF FF sensor. So this is a standard physical dimensions, exactly like we have the 35mm standard.
If Phase is keeping the pixel race in this standard format, the backs will have more and more pixels for a same physical area.
This is where I can't understand the limits. I thought that pixel density was relevant in IQ.
To my mind, more pixels for a same size means less IQ. So I see that the lenses tech could be the right response and not so much the sensors.
But I'm far from being a tech guru.
Cheers.
-
Hi,
The IQ issue is quite complex. The human vision is a limiting factor. So if the pixels are much smaller than what you would see in print they may not matter. So if you print small enough, increasing the number of pixels gives diminishing returns.
Now, IQ is often seen as noise related issues. We may call it noise or DR. The main source of noise in a picture is shot noise, random variation of the number of photons. Shot noise has a Poisson distribution, so it is proportional to the square root of the detected photons. Lets now assume that we replace a big pixel with a smaller one. Each pixel will count a quarter of the number of photons so noise will increase by a factor of two (the square root of 4 is 2). But the four pixels together will see the same number of photons as a single, so assuming the same enlargement the shot noise will be the same.
If we look at read noise as the limiting factor it will be the same weather the pixels are large or small. But the small pixels will be able to hold only a forth of the photons, so DR (which is defined as full well capacity / read noise) will be a fourth on a per pixel basis. To sum up:
- If you print large enough, resolution will be better with more pixels, God, lenses and diffraction permitting.
- Normally it's no big deal
- At the pixel level (like actual pixels in Photoshop) the pixels will be noisier
- If read nose dominates, you will be worse of with more pixels
On the other hand:
- There is no good reason to make pixels very small. If the pixels are smaller than what the lens can resolve or the diffraction limit, we will get diminishing returns.
- A bigger sensor will always see more photons
- Some sensors may be more efficient than others
Finally, obtaining maximum performance from a sensor is not a trivial task. Stopping down to f/16 may reduce a 50 MPixel sensor to a 12 MP one, because of diffraction.
Best regards
Erik
Bart, your points are very clear indeed.
But let's take for example the P65, wich is truth MF FF sensor. So this is a standard physical dimensions, exactly like we have the 35mm standard.
If Phase is keeping the pixel race in this standard format, the backs will have more and more pixels for a same physical area.
This is where I can't understand the limits. I thought that pixel density was relevant in IQ.
To my mind, more pixels for a same size means less IQ. So I see that the lenses tech could be the right response and not so much the sensors.
But I'm far from being a tech guru.
Cheers.
-
The important spatial frequencies for the human visual system in turn have to do with output magnification and viewing distance, with a peak contrast sensitivity (http://www.cg.tuwien.ac.at/research/theses/matkovic/node20.html) around 8 cycles/mm.
Bart,
An excellent post (as always), but I think the Contrast Sensitivity should be 8 cycles/degree, not 8 cycles/mm.
Regards,
Bill
-
An excellent post (as always), but I think the Contrast Sensitivity should be 8 cycles/degree, not 8 cycles/mm.
Hi Bill,
Thanks. OOPS, you're right, 8 cycles/degree is what I meant to say. Thanks for catching that (I've corrected that post to avoid confusion).
Cheers,
Bart