Jeff wrote:
Does anyone know if Canon's 'calculated' MTF curves ended up matching what was measured (by others) for the 70-20/2.8L IS II after it was released?
It's doubtful they would match charts made on an optical bench. I don't think anyone has done that sort of testing on Canon lenses for ages anyway. But it is true that the MTF charts said it would be better than the original version and it was measured to be so.
Bruce Sawle wrote:
You may want to check out the sites yourself before you make auc statements. Photozone and lenstip are the only reliable test sites photozone has the Nikon easily beating the canon at all focal lengths and apertures. Lenstip has the canon slightly ahead at f2.8. Dpreview is garbage as far as I'm concerned and the rest results are basically equal. SLR gear same as dpeview no chrys graphs or numbers to compare so basically a wash. Even the digital picture shows these lenses dead even.
You can't really compare lens+sensor results across brands or even bodies very well though. An optical bench type test lets you compare lenses to lenses but nobody does those anymore that I know of. You might be able to come up with a semi-reliable conversion factor for lens+sensor tests but I don't see that anyone has.
skibum5 wrote:
You can't really compare lens+sensor results across brands or even bodies very well though. An optical bench type test lets you compare lenses to lenses but nobody does those anymore that I know of. You might be able to come up with a semi-reliable conversion factor for lens+sensor tests but I don't see that anyone has.
Roger that. You have to account for the fact that the Nikon was tested using a higher resolution sensor. The MTF numbers were practically identical with the Nikon just ahead for the apertures I use most. If the Canon is that close on a lower resolution sensor, I would imagine it probably is slightly sharper optically.
Of course, I generally don't just peer through my lenses. I typically slap it on a body and shoot images, in which case if you're looking for sharp IMAGES and not just sharp lenses, the Nikon 70-200 still wins, but only just.
I think the fact that the 24-70 f/2.8 L II has been updated to such sharpness is a strong signal that Canon intends to release a high megapixel camera soon in the range of 30-36 MPx, conceivably more.
You may have too many pixels if the sensor pitch is less than one half the size of a line pair at your best lens's 50% MTF. For example, 36 MPx out-resolves the best Canon lenses, as far as I can tell. The 85 mm f/1.2 L does about 80 line pairs per millimetre at 50 % MTF and that is 12.5 micrometres. A 36 MPx full-frame sensor has sensel (pixel) dimensions of 4.9 micrometres: less than half the line pair size, and therefore meeting the Nyquist frequency limitations delineated by the Sampling Theorem.
Monito wrote:
I think the fact that the 24-70 f/2.8 L II has been updated to such sharpness is a strong signal that Canon intends to release a high megapixel camera soon in the range of 30-36 MPx, conceivably more.
You may have too many pixels if the sensor pitch is less than one half the size of a line pair at your best lens's 50% MTF. For example, 36 MPx out-resolves the best Canon lenses, as far as I can tell. The 85 mm f/1.2 L does about 80 line pairs per millimetre at 50 % MTF and that is 12.5 micrometres. A 36 MPx full-frame sensor has sensel (pixel) dimensions of 4.9 micrometres: less than half the line pair size, and therefore meeting the Nyquist frequency limitations delineated by the Sampling Theorem. ...Show more →
Maybe in the corners, but if you have used other APS-C cameras with similar resolution (like the D7000) it is clear that high MP can produce more detailed images.
EB-1 wrote:
Maybe in the corners, but if you have used other APS-C cameras with similar resolution (like the D7000) it is clear that high MP can produce more detailed images.
Yes, of course. That's the point.
Canon has been updating the resolution of their lenses to take advantage of what they know are the high resolution sensors coming down the road. 85 II, 16-35 II, 24 TSE II, 70-200 f/2.8 IS II, and now the 24-70 II.
Canon showed a 50 MPx sensor in full-frame size, several years ago.
jdben622 wrote:
Lucky guy. So would anyone REALLY notice you taking the gaffers tape off the test body and sticking in your own card?
There was a Canon rep stationed in front of each camera and they seemed to be watching pretty closely. Now that I think of it there were no signs saying not to mount lenses on your own camera so maybe I can get away with a few shots before they take it away from me
Bruce Sawle wrote:
You may want to check out the sites yourself before you make auc statements. Photozone and lenstip are the only reliable test sites photozone has the Nikon easily beating the canon at all focal lengths and apertures.
Have you looked at the charts. The Nikon beats te canon regardless of what there final verdict stars say if you buy a lens based on stars then you are in trouble. Look at the charts and the numbers don't lie.
Photozone and Lenstip make it abundantly clear that one cannot compare the absolute test values of one lens manufacture to another. It is obvious that Nikon and Canon lenses are tested on Nikon and Canon bodies, which have different sensors and pixel count amongst other variables including crop vs FF. Thus it is totally inappropriate to compare test charts, unless the test camera body is identical. The above interpretation of the test data is way out of line.
Mike K wrote:
Photozone and Lenstip make it abundantly clear that one cannot compare the absolute test values of one lens manufacture to another. It is obvious that Nikon and Canon lenses are tested on Nikon and Canon bodies, which have different sensors and pixel count amongst other variables including crop vs FF. Thus it is totally inappropriate to compare test charts, unless the test camera body is identical. The above interpretation of the test data is way out of line.
Totally agree about cross comparisons.
However, back of the envelope calcs show from the MTF50s at the center that both lenses are getting 0.96 linewidths per pixel, and drop to 0.8-0.9 at the edges. This suggests that the test results are sensor limited (no real surprise there) and a whole lot more testing and meridional and sagittal etc etc at MTF20 will be needed to declare a "winner"
Bruce Sawle wrote:
Have you looked at the charts. The Nikon beats te canon regardless of what there final verdict stars say if you buy a lens based on stars then you are in trouble. Look at the charts and the numbers don't lie.
Like many Nikon fanboys, you obviously do not know how to interpret charts with the corresponding texts. On Photozone, Nikkor FF lenses are tested on a 24 MP FF sensor while Canon FF lenses are tested on a 21 MP FF sensor.
Express the resolution of the edges wrt center to remove discrepancies due to different sensors, different RAW conversion etc:
@ Nikkor 70 mm f/2.8, ratio (edge/centre) = 2845/3701 = 77%
@ Nikkor 135 mm f/2.8, ratio (edge/centre) = 2822/3676 = 77%
@ Nikkor 200 mm f/2.8, ratio (edge/centre) = 3054/3554 = 86%
@ Canon 70 mm f/2.8, ratio (edge/centre) = 3070/3550 = 86%
@ Canon 135 mm f/2.8, ratio (edge/centre) = 2954/3454 = 86%
@ Canon 200 mm f/2.8, ratio (edge/centre) = 3100/3265 = 95%
"Wide open at F2.8, sharpness is excellent in the center of the frame, but it tends to fall off towards the edges; again best results are obtained at 200mm."
"Sharpness is extremely high, and impressively consistent right across the frame from center to corner. Truly excellent results are obtained at all focal lengths from wide open..."
You are nothing more than a loud but empty vessel. Show your numbers or some REAL side-by-side test shots, else forever be silent.
Addendum:
a) From Photozone site itself
"Please note that the tests results are not comparable across the different systems! This does also apply for the new EOS tests based on the EOS 50D because of differences in the sensor system (e.g. AA-filter) as well as different RAW-converters"
b) There are also sample to sample variations, both lens and cameras
