- Status
When I switched to full frame I ran the calculations and found that for EVERY SINGLE EF-S lens from focal lengths 10mm to 250mm (16mm-400mm equivalent) it's full frame equivalent lens would be always substantially much better, almost always cheaper, could acheive better subject isolation with shallower DOF and be much much sharper. In fact full frame lenses typically have twice the resolution of their crop counterparts.
Example:
10-22mm vs 17-40mm = lens is cheaper, has 202% the resolution, has 1-2 stops shallower DOF and has a longer reach on the long end, and is weather sealed.
17-55mm IS vs 24-105mm IS = lens is cheaper, has 205% the resolution, has 0.5 stops shallower DOF, and is both longer and wider, and is weather sealed.
55-250mm IS vs 70-300mm IS = lens is only $150 more, has 186% the resolution, has 1-1.5 stops shallower DOF, is wider, but also only has 300mm vs 400mm of reach, HOWEVER it is so much sharper that cropping it actually results in a 6% sharper image which has shallower DOF.
18-200mm IS vs Tamron 28-300mm XR (full frame superzoom) = lens is $100 less, has virtually identical resolution, has a 1 stops shallower DOF, is slightly wider but is slightly shorter (29 vs 28mm 320 vs 300mm).
Remember crop lenses need to have their focal length multiplied by 1.6 and 1.5 stops subtracted from their aperture to compare them to full frame lenses. So a 17-55mm f/2.8 lens is the equivalent of a full frame 28-90mm f/4.5.
The point is that there is zero image quality benefit and it would cost more to use EF-S lenses on full frame than their EF counterparts. It's pointless and engineering a body with this function would also be pointless.
Nikon's dogma as a company is compatabilty that's why you can use 50 year old lenses on their bodies which still work. Even if that compatability is pointless.
Keep in mind my argument is before you consider that you'd be working with very few megapixels if you were to make EF-S lenses compatible with full frame bodies. That really sinks the whole idea.
The only major argument for being able to use EF-S lenses on full frame is lens sharing. However if you actually study the possibility of lens sharing it's much better to share EF lenses between a crop and full frame than EF-S lenses between a crop and full frame.
EF lenses on crop tend to be about 10% worse in resolution than their EF-S counterparts.
So if you were to share EF-S lenses you'd lose 50% of your resolution just at the lens nevermind the megapixels on one body. If you share EF lenses you only lose 10%. I'm pretty sure most people would much rather have 10% less resolution than 50% so it's much much much better to only have EF lenses and share those with your crop than to have EF-S lenses and share those with your full frame instead of getting the EF counterparts. Enabling people to share EF-S lenses between their full frame and crop cameras would be pointless, as it's just giving them the option to have much worse and more expensive gear.
Hope that clears things up. If anything making EF-S lenses incompatible with full frame motivates people to not use EF-S lenses on full frame which is better and cheaper for them in every way.
Hope that helps.
Example:
10-22mm vs 17-40mm = lens is cheaper, has 202% the resolution, has 1-2 stops shallower DOF and has a longer reach on the long end, and is weather sealed.
17-55mm IS vs 24-105mm IS = lens is cheaper, has 205% the resolution, has 0.5 stops shallower DOF, and is both longer and wider, and is weather sealed.
55-250mm IS vs 70-300mm IS = lens is only $150 more, has 186% the resolution, has 1-1.5 stops shallower DOF, is wider, but also only has 300mm vs 400mm of reach, HOWEVER it is so much sharper that cropping it actually results in a 6% sharper image which has shallower DOF.
18-200mm IS vs Tamron 28-300mm XR (full frame superzoom) = lens is $100 less, has virtually identical resolution, has a 1 stops shallower DOF, is slightly wider but is slightly shorter (29 vs 28mm 320 vs 300mm).
Remember crop lenses need to have their focal length multiplied by 1.6 and 1.5 stops subtracted from their aperture to compare them to full frame lenses. So a 17-55mm f/2.8 lens is the equivalent of a full frame 28-90mm f/4.5.
The point is that there is zero image quality benefit and it would cost more to use EF-S lenses on full frame than their EF counterparts. It's pointless and engineering a body with this function would also be pointless.
Nikon's dogma as a company is compatabilty that's why you can use 50 year old lenses on their bodies which still work. Even if that compatability is pointless.
Keep in mind my argument is before you consider that you'd be working with very few megapixels if you were to make EF-S lenses compatible with full frame bodies. That really sinks the whole idea.
The only major argument for being able to use EF-S lenses on full frame is lens sharing. However if you actually study the possibility of lens sharing it's much better to share EF lenses between a crop and full frame than EF-S lenses between a crop and full frame.
EF lenses on crop tend to be about 10% worse in resolution than their EF-S counterparts.
So if you were to share EF-S lenses you'd lose 50% of your resolution just at the lens nevermind the megapixels on one body. If you share EF lenses you only lose 10%. I'm pretty sure most people would much rather have 10% less resolution than 50% so it's much much much better to only have EF lenses and share those with your crop than to have EF-S lenses and share those with your full frame instead of getting the EF counterparts. Enabling people to share EF-S lenses between their full frame and crop cameras would be pointless, as it's just giving them the option to have much worse and more expensive gear.
Hope that clears things up. If anything making EF-S lenses incompatible with full frame motivates people to not use EF-S lenses on full frame which is better and cheaper for them in every way.
Hope that helps.
Upvote
0
N
Neeneko
Guest
briansquibb said:
I don't have the link handy, but it was discovered a while back that digital sensors, because of their geometry, do not get the full benefits of low fstops. A much narrower family of angles actually register, so a good chunk of the extra light a really fast lens lets in does not actually register with the sensor. Canon quietly increases the ISO when it detects low apertures to give the appearance the extra light is doing more then it is. If you use manual aperture lenses (with no reporting) the behavior does not trigger and you can see the effect in play.
Upvote
0
Bruce Photography
Landscapes, 5DX,7D,60D,EOSM,D800/E,D810,D7100
wickidwombat said:Nikon cant use EF-S lenseshyles said:sorry just teasing I know you meant the Nikkor DX lenses
Actually I really love the format changes the D800 does so effortlessly. It isn't that I want to ever buy DX lenes (I don't except to try the 18-300), but when using the 2x extender on the 70-200 at 200 I get 400mm. But instead of cropping in post, I can switch to DX format and I'm now using a crop window in the view finder for DX as if I was using a 600mm lens. Now for Canon you are right. As long as 22MP is full frame you can't afford to crop anything. But with 36MP you can choose to crop now or crop later and save all card/disk real estate. When I'm doing birding, this is really quite nice because 15MP is ok if I can use all of it. Actually they have quite a few format crops that I can choose at the touch of a button. Perhaps this will turn out to be the true utility of a large MP sensor in addition to full MP landscapes full of detail for my large prints.
Upvote
0
It depends on the camera. The 7D is worst but it affects all digital cameras (not just Canon.)briansquibb said:
For the 5D2, the effect starts at about f/2.4. It's significant at f/1.4 and between f/1.2 and f/1.4 you get almost no benefit from the extra light. I posted my results on POTN a couple of weeks back. Here's a link http://photography-on-the.net/forum/showthread.php?t=1165164&page=3 and a second plot of my results... http://photography-on-the.net/forum/showthread.php?t=1165164&page=5. A few others also verified the effect with experiments.
The result of the post was that owners of the 50L and 85L started making really unpleasant comments. This was regretable but the fact is that Canon (and Nikon, Pentax, Sony) all increase the ISO to make it look like your fast glass is actually doing something for you.
Upvote
0
so are you saying that say you are shooting a concert in very low light with a 50 f1.4
@ f1.4 3200 iso
you would be better off shooting at say f2.2 @iso 8000
or f2.8 @iso 12800
because from an iso noise POV this is what is happening anyway?
@ f1.4 3200 iso
you would be better off shooting at say f2.2 @iso 8000
or f2.8 @iso 12800
because from an iso noise POV this is what is happening anyway?
Upvote
0
Actually no - the point is that light from the perifery of the lens does not make it to the active part of the sensor and so it cannot contribute to the image or shallow DoF. Mounted on a 5D2, a 50L can barely perform like a 50/1.6 (albeit a very good one.)briansquibb said:
Upvote
0
Not quite - but let's assume the f/1.4 lens collects light like a f/1.7 lens, then you would get the same effect using a f/1.7 lens and boosting the ISO by 1/3 EV. Of course, since a stopped down fast lens almost always outperforms a wide open slow one, you would probably sacrifice vignetting and perhaps image quality.wickidwombat said:
As I commented earlier, it is strongly dependent on the camera and I invite you to read the DxO Labs article that my POTN post links to.
Upvote
0
The reduced light transmission effect that is being referred to with large aperture lenses, is applicable to some extent to all lenses. It exists because the lens f-stop is only a theoretical measure of how much light a lens transmits to the film/sensor. In practice, glass absorbs some of the light and reflects another proportion (hence the importance of good lens coatings - see http://www.canonrumors.com/tech-articles/all-about-lens-coatings/). The effect of this is that less light reaches the film/sensor than the aperture of the lens would theoretically allow in a perfect lens; the measure of actual transmission of light through a lens is known as the t-stop (note that this is very important to cinema camera lenses -hence why they are often specified in t-stops: see the new Canon cinema-EOS lenses).
None of the above is unique to digital it is a fundamental part of lens physics. Thus lenses with a higher number of elements will (all else being equal) have a lower t-stop for an equivalent f-stop than lenses with fewer elements. This has negative t-stop implications for fast zoom lenses and lens based image stabilisation systems. There is another aspect to this, the structure of silicon sensors require light to hit at an angle that is closer to 90 degrees to their plane than was the case with film. This was ostensibly what required many lenses to be redesigned 'for digital' several years ago (believe that if you will). Wider apertures transmit light travelling at more oblique angles than smaller apertures, which is the factor that may reduce transmission of fast lenses more when used with digital sensors vis-a-vis film.
Some might argue that t-stop doesn't really matter for large aperture primes; with the superb high-ISO performance of today's CMOS sensors, it is shallow depth of field that is their primary usage, rather than for increasing shutter speed. This may not hold true because of the loss of detection of the oblique angle rays; to quote The Luminous Landscape:
"The DxO measurements to date prove that the marginal light rays just don’t hit the sensor. The point regarding depth of field is that these rays are also responsible for a larger blur spot when out of focus. If they are lost, they not only don’t contribute to the light intensity at the sensor, but they also don’t blur the out of focus planes as much as you would expect at wide apertures." [http://www.luminous-landscape.com/essays/an_open_letter_to_the_major_camera_manufacturers.shtml]
So far, I haven't seen anyone empirically test a large aperture prime on digital versus film to determine how much of a problem this is in the real world... any takers?
None of the above is unique to digital it is a fundamental part of lens physics. Thus lenses with a higher number of elements will (all else being equal) have a lower t-stop for an equivalent f-stop than lenses with fewer elements. This has negative t-stop implications for fast zoom lenses and lens based image stabilisation systems. There is another aspect to this, the structure of silicon sensors require light to hit at an angle that is closer to 90 degrees to their plane than was the case with film. This was ostensibly what required many lenses to be redesigned 'for digital' several years ago (believe that if you will). Wider apertures transmit light travelling at more oblique angles than smaller apertures, which is the factor that may reduce transmission of fast lenses more when used with digital sensors vis-a-vis film.
Some might argue that t-stop doesn't really matter for large aperture primes; with the superb high-ISO performance of today's CMOS sensors, it is shallow depth of field that is their primary usage, rather than for increasing shutter speed. This may not hold true because of the loss of detection of the oblique angle rays; to quote The Luminous Landscape:
"The DxO measurements to date prove that the marginal light rays just don’t hit the sensor. The point regarding depth of field is that these rays are also responsible for a larger blur spot when out of focus. If they are lost, they not only don’t contribute to the light intensity at the sensor, but they also don’t blur the out of focus planes as much as you would expect at wide apertures." [http://www.luminous-landscape.com/essays/an_open_letter_to_the_major_camera_manufacturers.shtml]
So far, I haven't seen anyone empirically test a large aperture prime on digital versus film to determine how much of a problem this is in the real world... any takers?
Upvote
0
May be a TC can solve the problem. I know someone have used a Kenko 1.4 TC on Sigma UWA's with success (more or less).
http://photography-on-the.net/forum/showthread.php?t=909337
I dont know if there are any TC on the market which accepts EF-S lenses, but perhaps...
From the samples I have seen, the quality is on the same level as with a crop body.
http://photography-on-the.net/forum/showthread.php?t=909337
I dont know if there are any TC on the market which accepts EF-S lenses, but perhaps...
From the samples I have seen, the quality is on the same level as with a crop body.
Upvote
0
maxxevv said:
yeah you'd have a horrible little VF with impossibly small to use coverage.
No the AF arrays have been pretty much the same size in all DSLR frame sizes APS-C,APS_H or FF. Using a full APS-C VF-szied secondary mirror for AF would even fit the 5D3 AF into APS-C.
Anyway you'd get a lot of vignetting and even hard-vignetting so I don't see how it is worth the bother even if it didn't make the VF useless.
Upvote
0
K
knocker
Guest
Wow!
When I posted the question I got 1 reply and thought oh well it must be a bone question never mind. Then I have another look and now I have this lot must be from over the pond me thinks
All of you thank you for the fantastic response I now understand it all a lot better.
The reason for the question was I have 400d, 40d, and the 7d along with
Canon FD 50mm F1.8+FD to ES conv = 1.4
Sigma 10-20mm f/4-5.6 EX DC HSM
Canon EF-S15-85mm f/3.5-5.6 IS USM
Sigma 17-70mm F2.8-4.5 DC Macro
Sigma 18-200mm F3.5-6.3 DC OS
Canon EF-S 18-55mm f/3.5-5.6
Canon EF 28-135mm f/3.5-5.6 IS USM
Sigma 30mm f/1.4 EX DC HSM
Canon EF 50mm f/1.8 II
Canon EF-S 60mm f/2.8 Macro USM
Canon EF 70-300mm f/4-5.6 IS USM
Canon EF 70-200mm f4 L IS II USM
Sigma 150-500mm f5.0-6.3 DG OS HSM
2x EX DG Tele Converter on 70-200 Canon
2x EX DG Tele Converter on 150-500 Sigma
And was wondering if I should get rid of the EF-S and the 3 body's if the new 7D was full frame or do I keep them in case something was invented that could use both but now I understand that the most economical way forward would be to get rid of them
Thanks again for all the VERY informative reply's
I have been on a lot of forums and there is not one bad post in this lot normally on a forum you get at least one dick head who cant resist a p**s take
Thanks.
Great web site as well Craig
When I posted the question I got 1 reply and thought oh well it must be a bone question never mind. Then I have another look and now I have this lot must be from over the pond me thinks
All of you thank you for the fantastic response I now understand it all a lot better.
The reason for the question was I have 400d, 40d, and the 7d along with
Canon FD 50mm F1.8+FD to ES conv = 1.4
Sigma 10-20mm f/4-5.6 EX DC HSM
Canon EF-S15-85mm f/3.5-5.6 IS USM
Sigma 17-70mm F2.8-4.5 DC Macro
Sigma 18-200mm F3.5-6.3 DC OS
Canon EF-S 18-55mm f/3.5-5.6
Canon EF 28-135mm f/3.5-5.6 IS USM
Sigma 30mm f/1.4 EX DC HSM
Canon EF 50mm f/1.8 II
Canon EF-S 60mm f/2.8 Macro USM
Canon EF 70-300mm f/4-5.6 IS USM
Canon EF 70-200mm f4 L IS II USM
Sigma 150-500mm f5.0-6.3 DG OS HSM
2x EX DG Tele Converter on 70-200 Canon
2x EX DG Tele Converter on 150-500 Sigma
And was wondering if I should get rid of the EF-S and the 3 body's if the new 7D was full frame or do I keep them in case something was invented that could use both but now I understand that the most economical way forward would be to get rid of them
Thanks again for all the VERY informative reply's
I have been on a lot of forums and there is not one bad post in this lot normally on a forum you get at least one dick head who cant resist a p**s take
Thanks.
Great web site as well Craig
Upvote
0
- Status
Similar threads
