p.67 #1 · Official: Sony A7 and A7R Fullframe Mirrorless
alundeb wrote:
The point was not to compare, but to illustrate that categorical rejection based on MTF charts with a drop in tangential 40 lp/mm lines is neither safe, nor consequential.
p.67 #2 · Official: Sony A7 and A7R Fullframe Mirrorless
Steve Spencer wrote:
This is convincing to me that the FE 24-70 is likely to not be a great solution for landscapes at the wide end.
That is exactly the point. MTF can be misleading in the sense that a lens may have stellar MTF, but performs poorly at close range, or vice versa.
p.67 #3 · Official: Sony A7 and A7R Fullframe Mirrorless
Steve Spencer wrote:
Come on you know that isn't a fair comparison. The FE is theoretical and the C/Y is measured. You also know that is cherry picking numbers in the extreme and creating a summary that distorts the overall pattern shown in the charts. It is looking at the MTFs as a whole that I see even the theoretical MTFs for the FE (which don't account for diffraction and thus are grossly inflated) the corners look pretty bad especially the 40mm lines. This is convincing to me that the FE 24-70 is likely to not be a great solution for landscapes at the wide end. Perhaps it isn't convincing to you, that is fine, but I think quite a few other people were worried about these MTFs as well ...Show more →
Have you really looked closely at the MTF chart for the 24-70 FE at 24 mm f/8?
The Sagittal 40 lp/mm is flawless across the frame.
The Tangential 40 lp/mm is above 50%(theoretical) up to 12mm. That is a very high value and extends to the bottom of the frame in landscape. It keeps above 20% until about 18mm, that is the edge of the frame.
Regarding differences between theoretical and measured values, those differences are most apparent at high MTF. At low MTF, in my experience, we often get what the chart promises.
The MTF charts indicate that the lens will resolve well for landscape at infinity, but the micro contrast will be asymmetrical, exceptionally high in one direction.
p.67 #4 · Official: Sony A7 and A7R Fullframe Mirrorless
Yikes! My first impression is not good on the wide end but extraordinary at the 70mm end. I compared this MTF to my Leica 35-70 R f/4 (which I love), the Canon 24-70 f/4 L (which I don’t have), to the C-Y Z 35-70 f/3.4 (which I don’t have – prefer the Leica construction) and the Canon 28-70 f/2.8 L (which I have and love). I did a quick down and dirty comparison of the Meridional/Tangential values from 15mm to the end (21.6mm) for wide angle wide-open & f/8 and the the telephoto end wide-open & f/8. Always difficult to determine the Canon since they measure micro-contrast at 30 lines/mm and not the 40 lines/mm that Zeiss and Leica use. However the Canon 28-70 MTF was from Photodo and was measured at 40lines/mm. This is how they stack up.
Lens_______________MTF_____WA-W/O___WA-f/8___T-W/O___T-f/8
24-70 ZA f/4________40/mm____45-8_______38-10____60-40____70-50
L 35-70R f/4________40/mm____45-11_______________40-34
C EF 24-70 f/4______30/mm____55-15______64-24_____53-29___75-62
CYZ 35-70 f/3.4_____40/mm____25-18______53-39_____40-22___59-35
C EF 28-70 f/2.8 ____40/mm____24-8_______17-14_____20-14___55-38
Comparing my treasured 28-70L to the 24-70 ZA does give me some hope. I think rather than cancel my order for the 24-70 I will see how much I like the Leica 35-70 on the A7r and decide from there.
Dave
Nov 14, 2013 at 11:53 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.67 #5 · Official: Sony A7 and A7R Fullframe Mirrorless
alundeb wrote:
Have you really looked closely at the MTF chart for the 24-70 FE at 24 mm f/8?
The Sagittal 40 lp/mm is flawless across the frame.
The Tangential 40 lp/mm is above 50%(theoretical) up to 12mm. That is a very high value and extends to the bottom of the frame in landscape. It keeps above 20% until about 18mm, that is the edge of the frame.
Regarding differences between theoretical and measured values, those differences are most apparent at high MTF. At low MTF, in my experience, we often get what the chart promises.
The MTF charts indicate that the lens will resolve well for landscape at infinity, but the micro contrast will be asymmetrical, exceptionally high in one direction. ...Show more →
Yes, I have looked closely at the MTF chart for the 24-70 FE at 24mm and f/8. Yes, the Sagittal line is flawless at 40 lp/mm, but it is actually too flawless at over 80% it shows just how skewed theoretical MTFs can be.
Although I do agree that differences between theoretical and measured MTFs are greatest at high MTFs, it is not as if they don't still create quite a bit of overestimation at low MTF. My experience is definitely not that you get what the chart promises. If that is the case then the 40 lp/mm sagittal line for this zoom is notably better than the Zeiss 55 Otus. If you really believe the measured MTF would be that good, no wonder you think this lens would be so strong. You also have to keep in mind that one of the things that theoretical MTFs don't take into account is diffraction which will of course be higher at f/8 than at f/4, so the overestimation is high at the smaller aperture as well.
So I think a huge part of our disagreement is whether we trust the theoretical MTFs or not. I think they are big overestimations, whereas you seem to think they are pretty accurate. This difference leads to a marked difference in interpretations of the charts, so it is not all that surprising that we disagree.
Beyond that disagreement, however, is how we interpret the tang line drop toward the corners at 40 lp/mm. I still think this is an overestimate but the line drops quite low in my view. I like to see this line around 60 percent across the whole frame, although a little drop past 18mm or so doesn't bother me much. From 15mm on out this line is lower than I would like to see and I am assuming it is an overestimation because it is estimated. So, for me this line suggests there will be some corner weakness. You aren't worried about it or so it seems.
I hope I have been fair in pointing to where we disagree, and we can get beyond whether we are each looking closely at the charts.
p.67 #7 · Official: Sony A7 and A7R Fullframe Mirrorless
And because of this going back and forth with a lot of discussion I ordered the A7 with the kit lens...
The 24-70 f4 has a high price (Zeiss name plate is expensive to make ;-) ) and not really nice mtfs.
For a third of that price you get the kitlens 28-80 f3.5-5.6 which is labeled as a G lens!
And constant f4 versus 3.5 till 5.6 is not that much of a difference...
For the times I wanna use a zoom the kitlens will do it. (I hope)
In 90% of my camera time I use primes anyway.
p.67 #8 · Official: Sony A7 and A7R Fullframe Mirrorless
MTF is good indication about how certain lens perform. I have found most of my Leica and Zeiss lens match MTF. It is very useful to understand particular lens without endless test.
24-70 is a very important lens to me, I always want a zoom lens around to do quick snap. I am a little bit disappointed for its MTF. but mainly its perfect Sagittal make Tangential looks worse than what it really is
as for compare Zeiss to kit lens, color, flare, contrast are real difference not just resolution. Have you seen Philip's test with 16-70 NEX zeiss with other lens, the color and contrast difference is so obvious. The same I see from magazine test between Canon L glass and their consumer stuff, difference is astounding make you think Canon use plastic for their consumer zoom and prime.
This MTF remind me leica 80-200 at tele-end, which is not as strong as its 100-135, but still very good performance to cover my need.
p.67 #9 · Official: Sony A7 and A7R Fullframe Mirrorless
Steve Spencer wrote:
I hope I have been fair in pointing to where we disagree, and we can get beyond whether we are each looking closely at the charts.
I did not mean to be rude. The reason I mentioned it was the possibility of some kind of optical illusion. The slope of the falling line looks bad, but it also starts very high. When looking closely myself, I was abit surprised at how high the MTF actually is far out in the frame.
p.67 #11 · Official: Sony A7 and A7R Fullframe Mirrorless
fotoingo2 wrote:
For a third of that price you get the kitlens 28-80 f3.5-5.6 which is labeled as a G lens!
And constant f4 versus 3.5 till 5.6 is not that much of a difference...
Honestly I was thinking the same thing, but with A7r kit is not even an option, and will the lens be sold stand alone?
But maybe I can get a used one from an A7 owner who doesn't like the chart
and TY @alundeb for chiming in on this. Some really do not realize or care that attacking a lens may lead some readers to actually change their orders, etc.
Of course awhile back I was a bit ignorant of just how good the FE 55 seems to be, but I never came close to infering it may be a dog, which has been done here to the 2470z.
Is it so hard to say: hey what about the way that line is looking? Draw discussion first, before the "back to the drawing board" dpreview forum style remarks.
p.67 #14 · Official: Sony A7 and A7R Fullframe Mirrorless
'I find comparing lenses based only on MTF charts to be a fool's errand.'
I have a lot of time for this view having been caught out a number of times. The main thing is consistent methodology, and these estimated data of Sony's are sometimes not far removed from quasi-fiction. They are neverthless a fair guide to the general behaviour. All very good lenses on a range of subject matter have excellent MTF, perhaps we can agree.
Now zooms of this nature will vary a fair bit from end to end, the designer's choice comes into play. Quite high curavature is very common at the short end. For Steve's benefit I posted a 28mm CY image that has quite poor corners in its chart, yet it produces good images, due to the oft-seen curvature that seems to be a part of Zeiss D28mm DNA. But the lines *fall together* and they are off CZ's optical bench - very reliable in general. Curvature in a moderate wide angle simply modifies the plane of sharp focus to an arc of sharp focus, an often benign aberration in the actual image.
This 24-70 zoom has wildly separated 40 lpmm lines on a calculated chart (therefore optimistic in general). Separated lines, from my imperfect understanding, is not so much a problem if they occur smoothly and steadily across the frame, as seen in many short tele Sonnars - the subject is generally off in the distance or such that quite flat sagittal data are much more important - the viewing eye moves front to back, not around the frame for typical subjects. But in today's world people want flat planes rendered well and a lens with these 24-70 data will not do that, for say, building facades, indoor scenes, etc - radial imaging of subjects will be worse than it should be. And on 36Mp people will give it all close scrutiny, as we are not printing 12x8s anymore to be happy.
I think people would expect something similar in the charts as those for Dave's 35-70/4 R or the Contax 35-70/3.4. Maybe a 28-70 or 35-70 would have been a better idea, if you have to end up using it at 35mm at the short end. Zooms are general lenses of course and need to do everything pretty well.
p.67 #16 · Official: Sony A7 and A7R Fullframe Mirrorless
ebrandon wrote:
Thanks for the link.
I'm definitely no expert in reading MTF charts, but don't the charts for the ZA 24-70mm f2.8 look way worse than the charts for the ZE 24-70mm f4?
And the ZA 24-70mm f2.8 is a brilliant lens! I'm really looking forward to using mine on the A7R.
Depends on what end you are looking at. On the Sagittal values the ZA 24-70 f/4 is significantly better in all settings. However, on the Tangential (micro-contrast) the ZA 24-70mm f2.8 scores much better (also consider it is f/2.8 vs f/4) on the wide end and at 70mm it lags behind the 24-70 f/4 ZA.
Nov 14, 2013 at 04:18 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.67 #17 · Official: Sony A7 and A7R Fullframe Mirrorless
ebrandon wrote:
Thanks for the link.
I'm definitely no expert in reading MTF charts, but don't the charts for the ZA 24-70mm f2.8 look way worse than the charts for the ZE 24-70mm f4?
And the ZA 24-70mm f2.8 is a brilliant lens! I'm really looking forward to using mine on the A7R.
I should probably shut up, because I am not sure anyone will listen to what I have to say anymore, but IMO the ZA 24-70 has much better MTF at least at 24mm (even though it is theoretical and again in my view a gross overestimation of performance). At 70mm they are closer and the FE probably has a slight edge. If you look at the ZA at f/8 you will see that there are very high sagittal lines at 10, 20, and 40 mm/lp all the way across the field. In comparing sagittal lines the ZA is only a bit better than the FE, partly because both are bumping up again the ceiling in possible performance (and again as these are theoretical values they are unrealistically high, much higher than any lens has every shown in actual measurements) so there is little room to show differences. Looking at tang lines you see the ZA out performing the FE by quite a lot especially in the outer frame. The ZA only drops to 40% out in the corners (which if it were a real measurement would be quite good), but the FE drops from 12mm on to below 50% and is at 10% in the corner. From my perspective that suggests probable corner problems for the FE that are not going to be nearly as a big of an issue with the ZA. So, I am glad people are optimistic, but the Sony/Zeiss theoretical MTFs suggest that at 24mm the ZA ought to be better, and the FE will probably struggle, and although I definitely believe theoretical MTFs should be taken with a huge grain of salt we can be quite certain that they represent the best possible performance and measured MTFs will be considerably worse.
p.67 #18 · Official: Sony A7 and A7R Fullframe Mirrorless
ebrandon wrote:
Thanks for the link.
I'm definitely no expert in reading MTF charts, but don't the charts for the ZA 24-70mm f2.8 look way worse than the charts for the ZE 24-70mm f4?
And the ZA 24-70mm f2.8 is a brilliant lens! I'm really looking forward to using mine on the A7R.
A year ago I made the mistake of buying the 24-70mm f2.8 ZA, it works well with cropped sensors and the FS100 video camera, as well the NEX-6.
However, my concern being how it will perform with the a7R and LAEA4 adapter. Vignetting can be dealt with, but edge resolving may be an issue.
At any rate, it won't be long now and it'll prove itself or find another owner.
Cheers
Michael
p.67 #19 · Official: Sony A7 and A7R Fullframe Mirrorless
cputeq wrote:
Maybe it's me, but I find comparing lenses based only on MTF charts to be a fool's errand.
It's not entirely without merit. It's certainly no different than the conversations that occur in other geek forums when all the members have to obsess over are specifications without any hope of immediate practical experience, so it seems to satisfy a social need if nothing else.
p.67 #20 · Official: Sony A7 and A7R Fullframe Mirrorless
Steve Spencer wrote:
I should probably shut up, because I am not sure anyone will listen to what I have to say anymore, but IMO the ZA 24-70 has much better MTF at least at 24mm (even though it is theoretical and again in my view a gross overestimation of performance)..
Your points are well made and understood. When I stop contradicting, it is usually because I accept the points you or somebody else makes Myself I got a bit anxious that nobody would understand my point, that I have seen good real world perfomance from some lenses with that characteristic drop in Tan MTF.
