Luminous Landscape Forum
Equipment & Techniques => Digital Cameras & Shooting Techniques => Topic started by: dwswager on October 28, 2014, 12:59:50 pm
-
With Film, image quality depended mostly on your lens, film choice and expertise. The camera merely was a tool that brought functionality, reliability and durability to the table. With digital bodies, image quality has always been limited by the camera (sensor sub systems, DSP, electronics, etc.).
With the 36MP D810, we have already hit the plateau in terms of raw megapixels. A significant portion of DSLR photographers deem 36MP too much, at least compared to trade-offs in other attributes like FPS, total frames buffered, file size storage and bandwidth issues. Hence, they have determined that image quality attributes gained by more pixels already are good enough at a lesser figure than what is currently available. Now the 4.9μm pixels size of the D810 is still well above that of the DX D7100 at 3.9μm. A full Frame sensor with 3.9μm pixels would be approximately 56MP! While obviously, technically feasible, I doubt from a practical standpoint we would ever see such a sensor, or if it would actually yield better image quality. I suspect the D5 (Edited) will have 24MPs or less because of this, though there may be a higher resolution variant released.
There are, of course, other attributes to image quality beyond resolution as proxied by MPs. Dynamic range and higher ISO performance being two biggies. Any body have ideas as to which attribute might be the next to plateau?
The Point
Not sure this has a whole lot of significance to a professional or the wealthy as they will purchase incremental upgrades to quality based on value or status. But for the amateur, somewhere in the plateau is where investment in new camera bodies make sense. The plateau can be caused by technical or practical limitations.
Prior to digital I shot a Nikon N90/N90s (forever) while playing with the Coolpix 9x0 series (900, 950, 990). I started in DSLRs with the DX D300 and currently shoot the D7100. I determined that I would jump back to FF only when the technology progressed (started to plateau) to the point that I wouldn't mind getting stuck with that level of image quality for an extended period of time. (Example: I have a 18 month old Samsung Galaxy Note II. The Note 4 is better in lots of ways, but I don't really mind using the Note 2. I'd hate to be stuck with a DroidX!) I personally think the D810 represents a seminal mark in DSLR evolution. In fact, other than resale value, If I was offered a free camera, I'd rather have a D810 (or to lesser extent due to 24MP and idiot dial interface, the D750) than any other current Nikon or Canon body! The D810 is that camera with which I wouldn't mind getting stuck! Glad I resisted biting on the D600, D610, D800/800e. While there will be better DSLRs down the road, I suspect they will be only small incremental advances to what the D810 offers. (Not that I won't lust after them!)
The larger point is that once the DSLR platform stabilizes, similar to how PCs have, the manufacturers can go back to doing what they should be doing, making them function better and consumers can be confident that their purchase isn't going to be 'outdated' in a few years time. I like that because I hate switching cameras. Once I spend the time an effort to learn a camera, I much prefer making images than spending time relearning a replacement tool.
-
if it would actually yield better image quality.
oversampling is good
-
I personally would hate to be stuck with a camera the physical size of a D810. After shooting a Fuji X-E2 for close to a year I'm waiting for something like the Sony A7* series with a bit more refinement.
Later,
Johnny
-
Maybe we are at the edge of the plateau, as nothing better than D800/810 has appeared in a few years. Of maybe camera sensor R&D departments are just catching breath and give us something even more amazing next year or year after that. In any case getting to next level of performance will be more expensive even if the sensors were cheap: designing and manufacturing lenses which are good enough for 50+ MPix FF sensors can not be cheap (from the average amateur viewpoint). I have D4 and D800e, most of the time I use D4 as it is faster and has better ergonomics. So I could say 36 MPix is too much, but sometimes I like to go "hi-fi"...
-
Hi,
I would expect that pixels shrink with improving manufacturing technology. The Sony RX100 I have has 2.4 micron pitch, that would correspond 150MP on full frame. So I would say 150 MP technology is feasible, I would also suggest that some oversampling is good. The images will be soft but will have better detail and less artefacts.
Many lenses are good enough for such a sensor, probably not over the entire frame, but over a significant central part. Large aperture lenses with maximum performance fuly open can be very expensive, see the Otus but it seems that the Sigma Art 50/1.4 is quite close, and it is of course possible to build f/2 or f/2.8 lenses that are equally sharp at a much lower price.
I would guess the main limitation may not be pixel size, but processing capability.
Best regards
Erik
Maybe we are at the edge of the plateau, as nothing better than D800/810 has appeared in a few years. Of maybe camera sensor R&D departments are just catching breath and give us something even more amazing next year or year after that. In any case getting to next level of performance will be more expensive even if the sensors were cheap: designing and manufacturing lenses which are good enough for 50+ MPix FF sensors can not be cheap (from the average amateur viewpoint). I have D4 and D800e, most of the time I use D4 as it is faster and has better ergonomics. So I could say 36 MPix is too much, but sometimes I like to go "hi-fi"...
-
Maybe we are at the edge of the plateau, as nothing better than D800/810 has appeared in a few years.
talking about cameras (and not sensors) why 'd you combine D800 with D810 ? let us consider a primitive absence of EFCS in D800 or primitive abilitity to use EFCS only w/ MUP in D810... here we go... and that affects image quality
-
There is no plateau in sight yet. Mobile tech will make it happen. Read this (somewhat overhyped maybe) article about the new Samsung NX1: http://www.imaging-resource.com/news/2014/09/27/photokina-interview-samsung-nx1-redefine-pro-performance-quantum-leap-tech
That's 28 megapixels readout at 240 fps from the sensor and 15 fps continous shooting. Nikon could do a 10+ fps D820 with those kinds of specs. We already see Nikon moving their processors to basically mobile tech with ARM chips in newer Expeed processors.
It's no stretch to imaginge "APS-C" sized pixels in full frame sensors. You are already using the same telephoto lenses so no problem there.
-
Imagine that around 100 yrs or so ago it was pronounced that all Physics had been illuminated and it was just a matter of solving a few equations to complete the entire knowledge.
It's that problem of the unknown unknowns again...
-
With the 36MP D810, we have already hit the plateau in terms of raw megapixels.
I have four [oops, five; I forgot the H2D-39] cameras that can capture images with greater than 30 MP resolution: D800E, D810, a7R, Betterlight Super 6K. The last one is the only one which hardly ever could use more pixels as I use it, because with the BL, there's no Bayer CFA. WRT the others, I'd like more than 200MP. Many of the lenses that I have could take advantage of the greater resolution. Even If I'm not going to print at 360 captured pixels per inch, Bayer-sourced images can be res'd down to give per-pixel quality of that of multi-shot cameras.
A significant portion of DSLR photographers deem 36MP too much, at least compared to trade-offs in other attributes like FPS, total frames buffered, file size storage and bandwidth issues.
I'm not arguing that cameras with hundreds of MP will be needed for every purpose. I use the D4 more than any camera I own. IT's great for situations where it can get the job done. I am arguing that High MP cameras will be in demand for the jobs that only they can do well. I am also arguing that the storage and transmission speed issues we have today will seem quaint in the future.
There are, of course, other attributes to image quality beyond resolution as proxied by MPs. Dynamic range and higher ISO performance being two biggies.
Dynamic range and resolution can be traded off against each other. You can lower resolution and get more dynamic range after the exposure, but you can't increase resolution after the exposure.
The larger point is that once the DSLR platform stabilizes, similar to how PCs have, the manufacturers can go back to doing what they should be doing, making them function better and consumers can be confident that their purchase isn't going to be 'outdated' in a few years time. I like that because I hate switching cameras. Once I spend the time an effort to learn a camera, I much prefer making images than spending time relearning a replacement tool.
If your camera does what you need it to do to make the images that you want to make, it is not outdated, no matter what the capability of newer cameras. There is no reason to switch cameras unless your present camera can't do what you want it to do, at least for the first 250,000 or 300,000 exposures.
Jim
-
About dynamic range: now we have about 14 stops of dynamic range from the sensors (D8xx series from Nikon et all), but that seems to be about the maximum possible, not that sensors could not be better maybe, but the optics can not deliver more due to internal reflections. They could, using simple designs with few lenses, but then the resolution suffers. So I suspect we have already reached the DR plateau what comes to total system performance.
-
About dynamic range: now we have about 14 stops of dynamic range from the sensors (D8xx series from Nikon et all), but that seems to be about the maximum possible...
Since you mentioned lens effects, I believe you're talking about intra-scene engineering dynamic range, the ratio of full scale to the read noise floor. In my opinion, that vastly overstated the dynamic range that is usable for normal photographic purposes. I'd -- and I'm not alone on this -- define photographic dynamic range as the ratio of the largest signal that doesn't clip due to photon noise to the smallest signal that produces a signal-to-noise ratio of 10. If the read noise is around an electron or two, that would be the photon noise from a signal with a mean count of 100 electrons. Using the Sony a7S as an example, that dynamic range would be about 150000/100 = 1500 = 10.5 stops.
Note that reducing read noise below the level I assumed can improve inter-scene DR, but not the DR of a single exposure, unless multiple pixels are averaged.
Jim
Jim
-
Nikon could do a 10+ fps D820 with those kinds of specs. We already see Nikon moving their processors to basically mobile tech with ARM chips in newer Expeed processors.
I agree, that processing capability is the one area not yet exploited in a significant way. As long as you can stay withing reasonable onboard power constraints, there is a lot that can be done with extra horsepower.
But if 10+ fps is the 'big update' in a D820, that it is making my point. Now, 10fps is a significant increase from 5fps, but other than very specific instances, it isn't a requirement. At worst, it is a crutch for those that don't actually know their subject or environment. Anyone remember manual focus/manual winding? Auto winders and 4fps motor drives were Da Bomb!
-
I have four cameras that can capture images with greater than 30 MP resolution: D800E, D810, a7R, Betterlight Super 6K. The last one is the only one which hardly ever could use more pixels as I use it, because with the BL, there's no Bayer CFA. WRT the others, I'd like more than 200MP. Many of the lenses that I have could take advantage of the greater resolution. Even If I'm not going to print at 360 captured pixels per inch, Bayer-sourced images can be res'd down to give per-pixel quality of that of multi-shot cameras.
The discussion is DSLR platform. I personally believe that the practical solution for those that need more than say 50 MPs is gong to be a system with larger sensors and/or different technologies.
If your camera does what you need it to do to make the images that you want to make, it is not outdated, no matter what the capability of newer cameras. There is no reason to switch cameras unless your present camera can't do what you want it to do, at least for the first 250,000 or 300,000 exposures.
That was my point. The D810 is really the 1st camera to provide the performance necessary to make the images I want to make. I have made great images with the D300 and D7100, but when I look back, I see the trade-off I was forced to make that the D810 minimizes, if not eliminates. The D810 doesn't make these 2 cameras worse, only shows that they weren't good enough...they were a stepping stone. Upgrading from and F2 to an F6 gets you basically zero image quality. That isn't true going from a D1 to a D4 or a D300 to a D810.
-
Do any of you photographic engineers think that Bayer technology is part of the "speed limit" in terms of useful DR, fps, buffers, noise etc....?
-
Do you remember the almost infinite enlargements of the point&shoot camera with amazing resolution and software enhacement in the classic movie Blade Runner?
I believe one day during the course of this century, we will be there and beyond. In technology plateaus are only temporary.
Eduardo
-
I don't think that there need ever be a "plateau" of image quality as far as dSLRs are concerned.
Perhaps a more meaningful plateau is fairly close now in terms of the ability to view those very high quality images - either as prints or on screens/projectors.
Or maybe the limitations of our own eyes will be the critical factor.
-
That was my point. The D810 is really the 1st camera to provide the performance necessary to make the images I want to make. I have made great images with the D300 and D7100, but when I look back, I see the trade-off I was forced to make that the D810 minimizes, if not eliminates. The D810 doesn't make these 2 cameras worse, only shows that they weren't good enough...they were a stepping stone. Upgrading from and F2 to an F6 gets you basically zero image quality. That isn't true going from a D1 to a D4 or a D300 to a D810.
Shooting at 10+ fps was not really my point. Only that processing power is already there. What I would like is electronic shutter and a near instant multishot/hdr/bracketing/whatever mode. Don't need 10 seconds worth of buffer. But I would like a full bracket in a second or less.
-
The D810 is really the 1st camera to provide the performance necessary to make the images I want to make. I have made great images with the D300 and D7100, but when I look back, I see the trade-off I was forced to make that the D810 minimizes, if not eliminates. The D810 doesn't make these 2 cameras worse, only shows that they weren't good enough...they were a stepping stone.
I'm happy for you, but I'd be careful about generalizing your personal experience. I'm different from you, and it's likely that others are as well, since we all are doing different things with our cameras, and have different quality criteria.
For example, I make more images with my D4 than any other camera, in spite of its relatively puny pixel count (although I remember when 16 MP was a high enough count to get many photographers to jump the MF ship). I find the ergonomics superior to the D810, never have to think about the battery life, hardly ever fill the buffer, etc. On the other hand, for some work I want the increased pixel count of the D810, and I'm happy to use it. By the way, I find the incremental improvements of the D810 over the D800E (AF, EFCS, LV, are my biggies) to make a big difference in the usability of the camera, but they are simply a series of incremental improvements. I can see another set of incremental improvements (global shutter, hybrid finder, 54 MP, variable conversion gain) making me want the next version.
I don't see the progress of technology stopping any time soon or my desire for better tools to slake. As I said before, I'd like more than 200 MP (a pixel pitch of 2 um wouldn't be too fine). I'd like fully electronic shutter with 1000+fps 8K readout. I'd like to choose between an optical finder and an EVF with a press of a button, and I'd like EVF information overlayed on the optical finder. I'd like a half-silvered, un moving mirror for both phase and contrast AF. Eventually, I'd like to ditch the mirror altogether. By the time I can get all those things, I'll probably want more.
And I'll bet all those improvements will let me take my photography to places that I can't imagine today.
Jim
-
Progress will continue because you have thousands of talented engineers pressured by greedy stock holders needing them to innovate to generate the cash they need to buy their next Ferrari, itself designed by talented engineers pressured by greedy stockholders needing them to innovate to generate the cash they need to buy their next camera. ;)
Now, the D810 is superior to MF digital cameras that were themselves described not long ago as having convinced top landscape photographers to abandon 4x5. So that tells us something about the level of performance relative to most people's real needs.
Cheers,
Bernard
-
I think a square sensor is a likely evolutionary step. Take the D800/810, nikon has already built in different aspect ratios like 5x4. A square sensor with dimensions of 36x36 would allow use of existing lens for the necessary image circle would not change.
-
I think a square sensor is a likely evolutionary step. Take the D800/810, nikon has already built in different aspect ratios like 5x4. A square sensor with dimensions of 36x36 would allow use of existing lens for the necessary image circle would not change.
If you draw an image circle and fit 36x24 just inside it, and then replace it with 36x36, you will notice the above does not hold water...
-
The D810 is an excellent camera, but concluding that it marks the plateau of digital camera and sensor development is akin to arguing about how many angels dancing on pin heads a camel passing through the eye of a needle actually sees.
-
The D810 is an excellent camera, but concluding that it marks the plateau of digital camera and sensor development is akin to arguing about how many angels dancing on pin heads a camel passing through the eye of a needle actually sees.
We are not at the plateau yet, but the incline is flattening out. Like I mentioned previously it is next to impossible to improve dynamic range, because of the internal reflections in lenses, not because sensors could not be improved at least slightly. Getting more resolution is possible, but only with best lenses, and the improvement is not going to be a quantum leap, like what we have seen from 4 to 36 MPix in ten years. Making a hypothetical 300 MPix FF sensor is not going to make the same impact going from 4 to 36 MPix did, it would go mostly unnoticed by the general public.
-
Maybe we _can_ make a quantum leap? (pun intended)
http://www.invisage.com/
-
Maybe we _can_ make a quantum leap? (pun intended)
http://www.invisage.com/
looks like something that may find its way into non-phones too....!
-
The D810 is an excellent camera, but concluding that it marks the plateau of digital camera and sensor development is akin to arguing about how many angels dancing on pin heads a camel passing through the eye of a needle actually sees.
;D
-
I personally would hate to be stuck with a camera the physical size of a D810. After shooting a Fuji X-E2 for close to a year I'm waiting for something like the Sony A7* series with a bit more refinement.
Indeed they are lots of ways a camera can be improved beyond just megapixels.
An EVF that will please those who currently hate them for example.
-
If you draw an image circle and fit 36x24 just inside it, and then replace it with 36x36, you will notice the above does not hold water...
circle has shortest perimeter (circumference) for any real figure.
-
If you draw an image circle and fit 36x24 just inside it, and then replace it with 36x36, you will notice the above does not hold water...
Petrus is correct: the diagonal of a 36x36 square is 50.9117 while that of a 24x36 rectangle is 43.2666 . That 15% difference is considerable in terms of lens' image circle Also even with fabrication and manufacturing costs falling the fabrication costs and rejection rate to make a perfect or near perfect CMOS wafer with a 51mm diameter that isremain substantially higher than to make one with a 44mm diameter.
On top of that larger and heavier camera bodies and lenses would have to made raising costs again, including R&D and tooling, and a larger and heavier overall package would also raise packaging, shipping, warehousing and delivery costs per unit; and finally a larger, heavier camera and lens combination would be harder to handhold and carry around -and push more people into choosing smaller devices for making photographs. In short a larger camera format would be a no-win situation for both manufacturers and photographers.
What would make more sense is a camera system built around a 24x24mm or 30x30mm sensor.
-
Also even with fabrication and manufacturing costs falling the fabrication costs and rejection rate to make a perfect or near perfect CMOS wafer with a 51mm diameter that isremain substantially higher than to make one with a 44mm diameter.
The two most common sizes for wafers in a CMOS fab are 200 and 300 mm. Perhaps you're not talking about wafer size at all, but of chip yield? You don't need a perfect wafer to get good chips from it, but yield is affected by chip size.
Jim
-
"Perhaps you're not talking about wafer size at all, but of chip yield?"
Yes! Thank you.
-
Indeed they are lots of ways a camera can be improved beyond just megapixels.
An EVF that will please those who currently hate them for example.
If you want that improved you have to say what is wrong with it, then give them [manufacturer employees reading the site] an idea of what would make it better.
-
If you want that improved you have to say what is wrong with it, then give them [manufacturer employees reading the site] an idea of what would make it better.
I'm not an EVF hater, but I'd like to see EVF's that are brighter in bright light, have a larger (and user-selectable) visual field, have less latency, have a higher refresh rate, have focus point selection by eye tracking (maybe some do that one already, but not on the cameras that I use), have enough dynamic range to show you what the image will look like with the a light source in the frame.
My guess is that the product managers at the camera companies know customers want all those things, and are beating on the poor engineers daily to design them in -- at no increase in COGS, of course.
Jim
-
I'm not an EVF hater, but I'd like to see EVF's that are brighter in bright light, have a larger (and user-selectable) visual field, have less latency, have a higher refresh rate, have focus point selection by eye tracking (maybe some do that one already, but not on the cameras that I use), have enough dynamic range to show you what the image will look like with the a light source in the frame.
My guess is that the product managers at the camera companies know customers want all those things, and are beating on the poor engineers daily to design them in -- at no increase in COGS, of course.
Jim
On the first point would a moth's eye screen protector work?
-
Found a screen using it.
http://www.trustedreviews.com/sharp-70in-4k-moth-eye-review
If they can put them in big tvs a tiny camera screen should be cheap.
-
If you draw an image circle and fit 36x24 just inside it, and then replace it with 36x36, you will notice the above does not hold water...
Yup, geometry was my strong suit in high school, and as a structural engineer, it's still pretty good.
-
Found a screen using it.
http://www.trustedreviews.com/sharp-70in-4k-moth-eye-review
If they can put them in big tvs a tiny camera screen should be cheap.
As I understand the technology that's in the display you linked to, it reduces reflections from a screen. That's not usually a problem with EVFs, since your head pressed against the eyepiece blocks most ambient light (although the camera manufacturers could provide better cups for eyeglass wearers. What I'm talking about is how dim the display is compared to ambient light in bright sun. The displays in SLRs aren't as bright as sunny 16 light, but they're serviceable in light that bright, at least with f/2.8 or f/4 lenses. EVFs are dimmer, making it harder to adjust to the change in light level, and making any leaking around the eye cup more significant. A hat helps, but that makes it harder when you're using the rear LCD, since the EVF light sensor detects the shadow of your hat and turns off the rear LCD, at least on the cameras I have, with the exception of the M240.
Jim
-
Right, I was thinking of the back screen which I always have to cup my hand over.
-
How does how we display or print the images figure into this? There is a 24x60" panorama lying atop my Stylus Pro 7900 right now - a single shot from an A7r, printed on Epson Exhibition Fiber. That's a center crop (not even - most of the crop was off the top, so it wasn't truly using the sweet spot of the lens) from a 40x60" image! It's a high-detail landscape, and it looks as good as a very well-made 11x14" print from most 35mm films. A 200 lb 7900 is about as big a printer as most people are going to tolerate sharing their house with (actually, most people won't tolerate anything bigger than a 3880) - a 9900 (the next size up) is the size of a piano! The very best FF cameras can fully utilize even a 44" 9900 or iPF 8400, which is what it would have taken to print this image full frame.
In film terms, this sort of print was solidly in large-format territory. A 40x60" enlargement from 6x9 cm film at the top end of medium format was equivalent to printing 16x24" from 35mm. Yes, it was possible with the right film and the right subject, but for a high-detail landscape, you REALLY wanted to have a 4x5 negative (or, for a panoramic image, 6x17 cm).
We don't really need more pixels if we're already printing 44"! No screen is close to using what we have - even the lab-curiosity 8k displays are only around 33 mp, and 4k displays that are more available are only 8 mp. Any recent Micro43 camera can exceed not only the resolution, but the color and dynamic range capabilities of the best 4K TV. In my part of the country, it is nearly impossible to sell a print with a short dimension exceeding 24", because of wall space restrictions (occasionally, a gallery wrap can go bigger). The one exception is institutions - our local hospital displays some VERY large prints, but there aren't that many hospitals and such around. Places where architecture tends towards bigger walls than New England may increase opportunities in the 30x45" and 40x60" range, but such a print will nearly always be viewed from a distance, and is well within the capabilities of current cameras. No, I wouldn't push the A7r (or, presumably the D810) as far as 60x90" for close inspection, but I'd probably try it on canvas for an installation with a reasonable viewing distance, and locations that can hold a 60x90" print (or even a 36x90" crop panorama) are very rare indeed.
There is certainly more to an image than resolution, but the best modern digital cameras spoil us there, too. We have enough dynamic range that there is REAL detail in Zone 0 and Zone X, and Zone -I (?) and Zone XI both hold the hints of tonality that film photographers worked to find in Zones I and IX. Color has improved tremendously, to the point where no film ever gave us what we have today.
Although serious photographers have the best equipment we've ever had (Ansel Adams would have killed for an A7r and a Cambo Actus to give him movements, although his back surgeon would have cried), most pictures today are taken with cameras that are actually historically BAD. Cell phones' 4-5 stops of dynamic range, diffraction-limited resolution and terrible lenses are worse than almost any film camera ever sold. The most sophisticated iPhone is somewhere in the quality range of 110 Instamatic and Kodak Disc film formats, due to the relentless drive to make phones ever thinner - we will never beat the laws of physics!
-
It really depends on the subject.
There are subjects for which a 60 inch print from a 36mp file feels very rough.
It may stand by itself, but when compared to a 150 megapixel files it does feel pretty low fi.
Cheers,
Bernard
-
If the Nikon D810 is the reference you suggest to get stuck on, I'd definitively say not for me at all. As an (advanced) hobbyist I'd never carry such a heavy and bulky camera with me (despite its sensor being a beast) in my street walks or vacation time. For me it was really liberating to forget the nightmare of carrying my DSLR's, and enjoy such a nice compact high IQ system like this:
(http://www.guillermoluijk.com/misc/pen1.jpg)
I can fit all the stuff I need inside the bag beside my girlfriend below, and enjoy it without having any neckache at the end of the day nor feeling observed when doing street photography. Not to mention that today I wouldn't like to live without the EVF, something out of the reach of DSLR systems.
(http://www.guillermoluijk.com/formentera/sanfrancesc.jpg)
With today's sensor technology, unless you are printing huge and need the highest quality (which I find reasonable for a professional photographer), I consider compactness more limiting than any technological parameter. We are perhaps demanding too much in the endless IQ race.
-
I will second Guillermo's point, and go one step farther - it's really LENS compactness that matters! An A7r body is compact enough for any shooting situation, but a Zeiss 24-70 f4 is a bit big for some! The Sony system would undoubtedly offer the best balance of size and image quality for almost any situation where its autofocus wasn't a problem - except that its lenses are big!
We have a higher-quality version of the physics problem that says that no normal-thickness cell phone will ever have a better than low-end Instamatic camera. In both cases, the lens you'd need simply can't exist. A compact 24-70 that covers 24x36 mm with decent speed is just as impossible to make using conventional techniques as a lens less than 5mm thick that covers any but the tiniest of sensors (the reason cell phones use the tiny sensors isn't space for the sensor itself, it's lens coverage).
There are a couple of ways forward while maintaining compactness. One is to improve the image quality of somewhat smaller sensors - APS-C lenses ARE smaller (and some of the Fujis are beautiful), and Micro43 lenses are smaller still (again, there are quite a few really nice ones). A second would be to improve the ability of sensors to deal with off-axis light (Leica M lenses are small and beautiful, but many don't translate well to digital because of their extreme ray angles). A third would be to use unconventional optics (radically shaped aspheric elements, exotic materials, diffractive optics or possibly even mirrors, for example) to decrease lens size. Cell phones already use some pretty radical aspherics, but every time a lens designer figures out how to bend light in less space, the phone designers use it to shave another mm off the thickness, rather than allowing the lens designer to improve image quality.
Dan
-
Hi,
One reason I am not interested in the Sony A7/A7r is that camera size matters little to me, as I always carry a 70-400/4-5.6 lens with me.
There are very few really good lenses for APS-C. Most good lenses are really full frame and larger than they need to be.
With 4/3 it is a bit different. The system is built around the sensor, and the lenses that are good are really good, at least regarding MTF data. So, I would say that if 16 MP or so is good enough than 4/3 is a very reasonable choice. I would also say that 16 MP is good enough for A2 size prints. That is my experience, but more significantly, this is also what Ctein said in a pretty recent interview with Michael Reichmann.
But of course, I want some more…
Best regards
ERik
I will second Guillermo's point, and go one step farther - it's really LENS compactness that matters! An A7r body is compact enough for any shooting situation, but a Zeiss 24-70 f4 is a bit big for some! The Sony system would undoubtedly offer the best balance of size and image quality for almost any situation where its autofocus wasn't a problem - except that its lenses are big!
We have a higher-quality version of the physics problem that says that no normal-thickness cell phone will ever have a better than low-end Instamatic camera. In both cases, the lens you'd need simply can't exist. A compact 24-70 that covers 24x36 mm with decent speed is just as impossible to make using conventional techniques as a lens less than 5mm thick that covers any but the tiniest of sensors (the reason cell phones use the tiny sensors isn't space for the sensor itself, it's lens coverage).
There are a couple of ways forward while maintaining compactness. One is to improve the image quality of somewhat smaller sensors - APS-C lenses ARE smaller (and some of the Fujis are beautiful), and Micro43 lenses are smaller still (again, there are quite a few really nice ones). A second would be to improve the ability of sensors to deal with off-axis light (Leica M lenses are small and beautiful, but many don't translate well to digital because of their extreme ray angles). A third would be to use unconventional optics (radically shaped aspheric elements, exotic materials, diffractive optics or possibly even mirrors, for example) to decrease lens size. Cell phones already use some pretty radical aspherics, but every time a lens designer figures out how to bend light in less space, the phone designers use it to shave another mm off the thickness, rather than allowing the lens designer to improve image quality.
Dan
-
The one disagreement I'd have with Erik's statement that "there are few really good lenses for APS-C" has to do with the Fuji system. APS-C lenses outside of Fuji tend to be low-end lenses made for cheap DSLRs (where the upgrade path is full-frame, and the better lenses are developed for full-frame) and kit and travel zooms predominate (there is the occasional better lens, predominantly wide-angle primes and zooms). Fuji, on the other hand, has a whole lineup of APS-C lenses, made for a system without full-frame bodies, and including a number of exquisite lenses.
-
Hi,
Sorry, I forgot about Fuji! Mostly, I was thinking Nikon, Canon and Sony. Fuji is like 4/3 optimized for a smaller sensor.
Best regards
Erik
The one disagreement I'd have with Erik's statement that "there are few really good lenses for APS-C" has to do with the Fuji system. APS-C lenses outside of Fuji tend to be low-end lenses made for cheap DSLRs (where the upgrade path is full-frame, and the better lenses are developed for full-frame) and kit and travel zooms predominate (there is the occasional better lens, predominantly wide-angle primes and zooms). Fuji, on the other hand, has a whole lineup of APS-C lenses, made for a system without full-frame bodies, and including a number of exquisite lenses.
-
Apart from Fuji, the world of APS-C lenses is a sad desert of 18-something zooms with elements made from the bottoms of Coke bottles, and a few better lenses, most of which are 10+ years old, designed when it wasn't clear full frame was going to be viable. There are a few new, nice, non-Fuji APS-C lenses, but only a few, and I can't think of ANY that goes in the category of really special lenses right now. Frankly, most of the special lenses I've seen released lately (apart from the Touits) are for one mirrorless system or another - Olympus PRO, Fuji, Sony Zeiss...
Canon and Nikon are resting on their laurels with modest redesigns of (very nice) full-frame lenses whose predecessors have been out for generations. Nikon doesn't even have a standard FX zoom with fixed maximum aperture and VR, other than a 24-120 that is pushing the boundary between a standard zoom and a superzoom. Neither Nikon's nor Canon's bread and butter 24-70 f2.8 pro zoom has image stabilization! Both are bazookas to carry and use. I've used the Nikon extensively (owned it for years, before going all-mirrorless), and it's a good lens, but I prefer the Sony Zeiss 24-70 f4, and possibly even the Fuji 18-55 f4 and the Olympus PRO 12-40 f2.8 (neither of which yet works with a sensor with the resolution of some of what I paired the Nikkor with). I never loved the bokeh of the Nikkor, not to mention that it is an incredibly clunky lens to shoot with - big, heavy, poor close focusing and pretty much requires a tripod much of the time (and a tripod at least one size heavier than I usually use with my mirrorless systems, maybe two!).
Dan
-
First, I think many mistook 'plateau' for 'pinnacle'! By plateau, I meant that improvements in image quality would slow and other aspects of development would supersede. A plateau is a breathing point and is usually not constrained by technology development, but cost, manufacturing or other practical considerations. It is a point where the user might say 'good enough for now'. Additionally, we are talking DSLR. That is using a 24mm x 36mm sensor.
If we use MP (proxy for captured resolution), dynamic range, color depth and high ISO performance as our measure of image quality. Then a poster in this thread indicated the D4 represented that threshold for him. And I believe there is a lot of others that would agree. That is, 16MP was good enough and other considerations out weigh his desire for additional MP. If you asked most D4 shooters, however, if they could get everything they have now, would they prefer the camera to output 24MP, I think most would respond positively. If, however, you move that to 36MP, what would the answer likely be? I know for certain many photographers have shied away from the D810 specifically because they do not need, nor want to deal with 36MPs worth of data. Are there photographers that would like to have more than that, absolutely, but the questions become is there a better, more cost effective solution than a DSLR sensor for that and is it likely to sell well enough to make it feasible for the manufacturer, at least in the near term?
I would love to see the sales projections versus actual sales for the D800/D800e twins. I suspect total sales for the twins well outstripped Nikon's sale projections and that the D800e demand outstripped projections to a greater degree than the D800. I'm pretty sure the D810 is the result.
I also agree with the poster that mentioned processing power as the next area where advances could be made (allowing for onboard power constraints). Current generation RADAR algorithms, a computationally intensive process, are much less sophisticated than they used to be precisely because the amount of processing power available is so much larger. Processing power obviously would help frame rates, focusing and metering and overall usability. I personally have 2 things I would love to see from cameras that processing power/memory would make possible.
1. User save-able presets (no I don't mean U1 and U2 on Enthusiast Nikons or the half-assed shooting banks). Every setting on a modern camera, even if there is a mechanical switch like live view or the mode dial on the D750 is set electronically. What I want is the ability to setup the camera how I want and save with a name those settings (complete state of the camera which would preclude mutually exclusive options). A camera might say hold 10 such settings. You could get at them from a menu or a button/dial setup. Maybe there would be 3 or 4 buttons in a group on the camera that you could save 3 or 4 presets for direct access.
2. The ability to run 3rd party apps directly on the camera. Thinking something like Helicon Remote, for example. In studio, I'd prefer running this tethered to a PC with a large screen and more efficient keyboard/mouse interface, but in the field it would be better done in camera than tethered to my Samsung Note II.
Just my thoughts!
-
1. User save-able presets (no I don't mean U1 and U2 on Enthusiast Nikons or the half-assed shooting banks). Every setting on a modern camera, even if there is a mechanical switch like live view or the mode dial on the D750 is set electronically. What I want is the ability to setup the camera how I want and save with a name those settings (complete state of the camera which would preclude mutually exclusive options). A camera might say hold 10 such settings. You could get at them from a menu or a button/dial setup. Maybe there would be 3 or 4 buttons in a group on the camera that you could save 3 or 4 presets for direct access.
.
Just my thoughts!
Leica in the "T" and "S" do this. As does my 6D to some extent
can't speak to Nikon
-
One of my favorite crazy ideas (admittedly for a relatively small market) has always been sensor-based tilt! Shift would require the camera to increase in size, but tilt (and swing) should be relatively easy - we already have actuators that can move the sensor for stabilization and cleaning. It shouldn't be that much harder to add tilt to that!
We could also improve getting everything into one camera. I shoot an A7r, a GH4, and an E-M1. Between the three, they're a perfect camera! I'd love to see a camera that has the still resolution and dynamic range of the A7r, video like a GH4 and the build of an E-M1. This camera could bin the pixels 4 to 1 for 4k, and 16 to 1 for Full HD, both with near-perfect pixel counts - the A7r is within a few percentage points of being standard 8k (twice Ultra HD 4k) across, with some extra pixels in the short dimension to account for the different aspect ratio - treat each 2x2 block as a pixel and cut off the top and bottom, and you have UHD 4k. Treat each 2x2 block of THOSE pixels as one and you have 1080p FHD (incidentally, at a ridiculously high max ISO). This requires some serious processing power (and write speed), especially if you enable the accidental 8k video mode it probably spawns.
Dan
-
Pentax has a star tracker based on IBIS. I have also thought about tilt. Maybe we run into some heat dissipation issues and a whole lot bigger sensor to move precisely. But why not, it could be done. Imagine using multiple touch points on screen to focus a tilted plane.
Reading the entire sensor for video is precisely what that Samsung NX1 does. It's 4k but we'll get to 8k soon enough.
I would like backlit programmable buttons. They could change color with a led. Iso is red, mode is blue, focus is green and so on.
Running 3d party apps should already be there. Why you have to get a different accessory to run camranger or similar just blows my mind. People hack Canons with magic lantern already. It can be done. It should be a feature, not a hack.
But they still make image capture devices without touchscreens and WiFi so I'm not holding my breath for any of this.
-
Maybe my standards are a bit low, but the Canon EF-S 15-85mm f/3.5-5.6 IS is a reasonable landscape walkabout lens for APS-C. My guess is that there just isn't the market for very high quality zooms for APS-C costing more than $1,200.00. Even in full frame world, there are a lot of people bypassing the expensive Canon 24-70 f/2.8 II.
-
Maybe my standards are a bit low, but the Canon EF-S 15-85mm f/3.5-5.6 IS is a reasonable landscape walkabout lens for APS-C. My guess is that there just isn't the market for very high quality zooms for APS-C costing more than $1,200.00. Even in full frame world, there are a lot of people bypassing the expensive Canon 24-70 f/2.8 II.
Not just expensive... heavy and bulky too
Better sensors are just making the Big Glass less necessary in terms of light capture. Quality may be better in the 2.8 vs. the 4.0 but the physical factors just push towards the more compact designs
-
No question about that - most of the recent high-end zoom designs have been f4.0 (or even variable in a tight range around f4.0, like the Fujis that are f3.5-4.5). The exception has been the Olympus PRO lenses, but those have the advantage (an f2.8 lens can be smaller, both because of the reduced coverage requirement and because an "80" is really a 40, reducing the front element size) as well as the disadvantage (need to be faster because of bokeh, DOF and ISO on the smaller sensor) of their smaller sensor. Apart from the Olympi, most f2.8 zooms are recycled designs, and the new f4.0 lenses, sometimes in the same manufacturer's line, often beat them on image quality, as well as compactness and features. If we were seeing truly new f2.8 zooms, someone would have come out with a truly top-end stabilized 24-70 or similar...
Dan
-
IMO cameras oughta be able to interact as easily & seamlessly with all our devices—smartphones, tablets, laptops, desktops, NAS—as they do with their CF or SD storage cards.
Re. photosite count…I'd like enough to be able to map an R, G & B from each Bayer array directly to each R, G & B sub-pixel of a 4K (4096x2160) electronic display. 36mp doesn't cut it for this, nor does our current RAW processing software. :)
-Dave-
-
What would make more sense is a camera system built around a 24x24mm or 30x30mm sensor.
I think it is complex to decide what "makes sense" in the general... sense. If you have a large batch of lenses that are optimized for exactly 24x36mm (including light-blocking non-circular objects), it might not make sense to stray outside that sensor size. If you are free to design lenses along with sensor systems from scratch, you might have some freedom. If your users demand 16:9 images to go with their tvs or 1.41:1 ratio for A4 prints, then this might bias the "optimal" selection.
If ever optics performance vs cost becomes dominant to sensor performance vs cost (I don't think we are there yet), it might make sense (for some users) to spend a lot on sensor in order to maximize optics performance. That would (perhaps) be a sensor convering the entire usable image circle. The simplest shaping achieving that would be a square sensor of height/width equalling the diameter of the optics image circle.
Of course, the image from such a sensor would have to be cropped one way or another to make sense for most photography applications. Thus you are "wasting" sensor area and putting some (minimal) demand on end-user cropping skills in order to maximize the quality/flexibility that can be had for a given image circle.
-h
-
With the 36MP D810, we have already hit the plateau in terms of raw megapixels. A significant portion of DSLR photographers deem 36MP too much, at least compared to trade-offs in other attributes like FPS, total frames buffered, file size storage and bandwidth issues.
Those would seem like trade-offs that will change over time. 64MB was an investment in memory cards when I bought it many years ago, now it is worthless.
While recording, processing and storing 36MP at sufficient framerates might be at the limit today, I am guessing that inexpensive cell phones will be able to do just that in a few years (they might not use the quality components that gives you "high" quality, though).
A full Frame sensor with 3.9μm pixels would be approximately 56MP! While obviously, technically feasible, I doubt from a practical standpoint we 1)would ever see such a sensor, or 2)if it would actually yield better image quality.
I would be willing to bet quite a lot that 1) is wrong.
2) may well be right for many users. I don't think that it will be right for everyone.
There are, of course, other attributes to image quality beyond resolution as proxied by MPs. Dynamic range and higher ISO performance being two biggies. Any body have ideas as to which attribute might be the next to plateau?
Everyone have their ideas. Many of us will be wrong. The answer may well be a "socio-economic" one rather than a technical one. (Our understanding of) physics may limit what can be done at any one time, but the products that are available to customers may be limited by other factors, such as:
1. What is the cost of developing some new technology
2. What is the number of customers willing to lay out cash on that technology
3. What is the expected return of investment doing this as opposed to investing in something else
Not sure this has a whole lot of significance to a professional or the wealthy as they will purchase incremental upgrades to quality based on value or status. But for the amateur, somewhere in the plateau is where investment in new camera bodies make sense. The plateau can be caused by technical or practical limitations.
Perhaps the amateur (French amateur "lover of" (http://en.wikipedia.org/wiki/Amateur)) is the one that buys new tech "no matter what", while professionals (to a larger degree) buys "whatever will get the job done"?
The larger point is that once the DSLR platform stabilizes, similar to how PCs have, the manufacturers can go back to doing what they should be doing, making them function better and consumers can be confident that their purchase isn't going to be 'outdated' in a few years time. I like that because I hate switching cameras. Once I spend the time an effort to learn a camera, I much prefer making images than spending time relearning a replacement tool.
Is there some contradiction between increasing the number of megapixels and making cameras function better? Of course, if manufacturers succeed at this, it means that people will upgrade more often and used cameras will loose their market value more rapidly. I don't see a problem with that, the opposite would be a stagnated market where people keep their cameras for 10 years because manufacturers are unable to develop significant improvements. That means less money put into R&D and less improvement in image quality.
-h
-
Regarding the thread title, it depends on the type of photography you do. For some you are already there.
I think a lot of people have taken Jim's analysis of the Otus to represent what the future camera system can do. He never said you could do that. The analysis for the Otus is for that size of lens at that focal length.
On the other hand you take the very fine nikon 600 f4, drop a 2x tele on it, put it on a 36MP sensor, you have already passed diffraction.
The manufacturers dance around the physics so you don't really know what is going on. The simple fact is you will hit the wall when you try to extract too much angular resolution from a lens of a given size. The simple formula is (for green light) 134/Diameter in mm gives you the diffraction limit in seconds of arc. You know your field of view from the fl on 35mm.
The front filter on the Otus is 77mm. The 600 f4 is 52mm (not really f4 is it?).
field of view for the Otus is about 39 degrees, 3.4 for the 600, 1.72 with a 2x tele
1.72 degrees x 3600 = 6192 arc seconds. Diffraction is 134/50mm = 2.68 arc seconds. The d800 is 7360 pixels across. You are resolving to line pairs, not pixels. Notice how even the best long lenses in the DxO database have low resolution numbers?
Go through the same thing for the much wider and bigger otus, you can have a massive number of fine pixels.
What saves your tiny cell camera? The wide angle.
So the answer to the question in the title is in what angle you need, along with how big is the lens (assuming it is good).
-
On the other hand you take the very fine nikon 600 f4, drop a 2x tele on it, put it on a 36MP sensor, you have already passed diffraction.
The manufacturers dance around the physics so you don't really know what is going on. The simple fact is you will hit the wall when you try to extract too much angular resolution from a lens of a given size. The simple formula is (for green light) 134/Diameter in mm gives you the diffraction limit in seconds of arc. You know your field of view from the fl on 35mm.
The front filter on the Otus is 77mm. The 600 f4 is 52mm (not really f4 is it?).
The 52mm filter for the Nikon 600 goes inside the lens.
Jim
-
The nikon website lists the protective front filter as 52mm. That is what I am going on. If someone says the front is a 150mm lens I stand corrected.
-
The nikon website lists the protective front filter as 52mm. That is what I am going on. If someone says the front is a 150mm lens I stand corrected.
From the specs page on the Nikon web site:
(http://www.kasson.com/ll/nikon600 filter.PNG)
(http://www.kasson.com/ll/600mmsize.PNG)
I don't have the Nikon 600, but I do have the 400mm f/2.8, which has a similar-sized, but slightly smaller, front element. It is one big lens,
Jim
-
Thanks Jim,
Before correcting my above post, is there anything else needing fixing?
-
Thanks Jim,
Before correcting my above post, is there anything else needing fixing?
I question the premise. While the focal length does indeed play into the angular resolution, the linear resolution in the sensor plane for a diffraction-limited lens is dependent on only the wavelength of the light and the f-stop.
Here's a table I created a few years ago, before I started thinking of this problem in terms of MTF multiplication:
http://blog.kasson.com/?p=432
I haven't looked at it in a while; I hope there are no errors.
Here's a classical image system engineering look at the problem:
http://blog.kasson.com/?p=5742
The reference system in the "System Q" metric is monochromatic, unfortunately. The concept does approximately extend to Bayer sensors, but it gets a little sloppy, since Bayer resolution varies with direction, demosaicing algorithm, CFA filter characteristics, as well as wavelength.
Jim
-
I question the premise. While the focal length does indeed play into the angular resolution, the linear resolution in the sensor plane for a diffraction-limited lens is dependent on only the wavelength of the light and the f-stop.
Here's a table I created a few years ago, before I started thinking of this problem in terms of MTF multiplication:
http://blog.kasson.com/?p=432
I haven't looked at it in a while; I hope there are no errors.
Here's a classical image system engineering look at the problem:
http://blog.kasson.com/?p=5742
The reference system in the "System Q" metric is monochromatic, unfortunately. The concept does approximately extend to Bayer sensors, but it gets a little sloppy, since Bayer resolution varies with direction, demosaicing algorithm, CFA filter characteristics, as well as wavelength.
Jim
The f stop is a handy way of bypassing dealing with the fl and the angle of view. Your table has the pixels across at diffraction which says the same thing.
I think the premise does work because that is how we work as photographers. Personally my two biggest styles of photography are landscape then wildlife. For landscape I don't worry about the details. For wildlife I very much worry about the angle of view and the pixel level resolution. To get undisturbed wildlife shots I use either a 300 f4 on aps-c (85mm front element) or a 1200fl 222mm (10"-1.75" secondary) telescope. That lens has angular resolution that will work on my ff or my finer apsc for 1800fl equivalent from the crop. I can also use a 2x tele with both. I am not blowing $12k on a 600mm lens.
Given most photographers have one goto long lens for wildlife, the angle of view is central to them placing themselves to fill the frame with their quarry. Mess that up you miss it all.
Going back to my original post the premise is what you are shooting dictates the field of view you want. We then translate that into sticking a lens of the proper FL on the camera. For the people that use long FLs they already have to start thinking about the lens with tele, matching their pixel density. For others using wide angles, it will never become an issue.
On the other hand it does not work to go to a scene then pick an f2.8 lens or an f5.6 lens. I bet your table is correct. I do not work that way so IMO that premise is wrong.
-
The f stop is a handy way of bypassing dealing with the fl and the angle of view. Your table has the pixels across at diffraction which says the same thing.
I think the premise does work because that is how we work as photographers. Personally my two biggest styles of photography are landscape then wildlife. For landscape I don't worry about the details. For wildlife I very much worry about the angle of view and the pixel level resolution. To get undisturbed wildlife shots I use either a 300 f4 on aps-c (85mm front element) or a 1200fl 222mm (10"-1.75" secondary) telescope. That lens has angular resolution that will work on my ff or my finer apsc for 1800fl equivalent from the crop. I can also use a 2x tele with both. I am not blowing $12k on a 600mm lens.
Given most photographers have one goto long lens for wildlife, the angle of view is central to them placing themselves to fill the frame with their quarry. Mess that up you miss it all.
Going back to my original post the premise is what you are shooting dictates the field of view you want. We then translate that into sticking a lens of the proper FL on the camera. For the people that use long FLs they already have to start thinking about the lens with tele, matching their pixel density. For others using wide angles, it will never become an issue.
On the other hand it does not work to go to a scene then pick an f2.8 lens or an f5.6 lens. I bet your table is correct. I do not work that way so IMO that premise is wrong.
To each his own. It's the results that count. If it works for you, you're doing the right thing.
Jim
-
Apart from Fuji, the world of APS-C lenses is a sad desert of 18-something zooms with elements made from the bottoms of Coke bottles, and a few better lenses, most of which are 10+ years old, designed when it wasn't clear full frame was going to be viable.
I don't know Nikon well, but my Canon EF-S 17-55 f/2.8 USM IS is decent. I hear that the 15-85 is optically good as well. The EF-S 60mm macro is said to be good. The new EF-S 11-18 appears to be a solid performer at a low cost.
What is really missing (in my view) is top-notch normal primes with large aperture (and ideally IS). My guess is that we will never see those, since the customers willing to pay for (and carry) those kind of lenses can easily opt for FF.
I have gotten the impression that there is little reason to do e.g. an EF-S 70-200 since the reduction in size/weight/price would be minimal for a given performance, vs just using the excellent EF versions?
-h
-
I don't know Nikon well, but my Canon EF-S 17-55 f/2.8 USM IS is decent. I hear that the 15-85 is optically good as well. The EF-S 60mm macro is said to be good. The new EF-S 11-18 appears to be a solid performer at a low cost.
What is really missing (in my view) is top-notch normal primes with large aperture (and ideally IS). My guess is that we will never see those, since the customers willing to pay for (and carry) those kind of lenses can easily opt for FF.
I have gotten the impression that there is little reason to do e.g. an EF-S 70-200 since the reduction in size/weight/price would be minimal for a given performance, vs just using the excellent EF versions?
-h
Many factors including ones you cited prevent Nikon or Canon from doing too much in DX lenses. I have owned 2 DX bodies D300 and D7100 and I have ever only purchased 1 DX lens - 18-200mm f/3.5-4.5 VR II which is surprisingly good, distortions and all. The DX Nikkor 35mm f/1.8 looks like a steal, but I already have a 35mm f/2D. There is nothing preventing you from using FX size lenses with the benefit that as FX cameras come down in price you can use them on both. On the wide end, it is just too much cheaper and easier to do fast lenses for FX than equivalent FOV on DX.
-
There is nothing preventing you from using FX size lenses with the benefit that as FX cameras come down in price you can use them on both.
But designing a lens specifically for crop sensors evidently allows for smaller, lighter and cheaper lenses (in some cases). So should one buy lenses for what one needs today, or for what one might need in the future?
On the wide end, it is just too much cheaper and easier to do fast lenses for FX than equivalent FOV on DX.
Crop wide-angles tends to be cheaper than FF wide-angles. The EF-S 10-18mm f/4.5 - f/5.6 with IS costs $299 and gets good reviews.
http://www.the-digital-picture.com/Reviews/Canon-EF-S-10-18mm-f-4.5-5.6-IS-STM-Lens.aspx
An equivalent FF lens would have to be 16-29mm f/7.2-f/9 IS. Such a lens does not exists AFAIK, but if it did, would it cost $299? I doubt it. The Canon 16-35mm f/4.L IS costs about 4x as much. Granted, it is a very different lens.
We could talk about the Sigma 18-35mm f/1.8 vs 24-70mm f/2.8.
-h
-
Please do!
Best regards
Erik
We could talk about the Sigma 18-35mm f/1.8 vs 24-70mm f/2.8.
-
Please do!
Best regards
Erik
http://www.photozone.de/canon-eos/872-sigma1835f18_canon?start=1
"The Sigma 18-35mm f/1.8 DC HSM | A produced nothing short of stunning resolution figures in the MTF lab.
...
this Sigma lens is also as expensive, as big and as heavy as its full format counterparts"
Tested on a (15MP, 3168-line) Canon 50D. MTF50 : 2500 LW/PH @ 18mm and 24mm f/1.8 in image centre. >2000 LW/PH at centre/border/extreme at all apertures and focal lengths tested.
http://www.photozone.de/canon_eos_ff/773-canon2470f28mk2ff?start=1
Tested on a (21.1MP, 3744-line) Canon 5Dmk2. MTF50 : 3500 LW/PH @ 24mm and 40mm f/2.8 in image centre. >2300 LW/PH at centre/border/extreme at all apertures and focal lengths tested.
If you want the same image quality, light-gathering capability/DOF etc, it seems that similar weight, size, price etc is needed in order for a product to be feasible (constrained by >f/0.5 not being practically available at any crop factor). The wildcard, it seems, is that manufacturers _choose_ to make products based on what they believe will sell, and so far they have often prioritized lower price/size/weight on the smaller formats, and more light/DOF-flexibility for larger formats.
-h