akpo.ca wrote:
Any modern 300/2.8 from Canikon. All of these lenses should cover 44x33mm quite well too
Afraid not. Check the MTF graphs. The Leica 280/4 accomplishes at least equal sharpness and superior apo correction with several fewer elements. The Canikon lenses are doubtless better in the area of autofocus tho. 😜
Luvwine wrote:
Afraid not. Check the MTF graphs. The Leica 280/4 accomplishes at least equal sharpness and superior apo correction with several fewer elements. The Canikon lenses are doubtless better in the area of autofocus tho. 😜
Don't think that the R 280/4 opens up to f/2.8 as well. Sure, many vintage optics are great. That's not the point of my generalization. Modern lenses tend to be better performers than their vintage counterparts. Regarding the 180/2.8, the 200VR and VRII from Nikon are apochromatic. Pretty sure the 200L IS from Canon is also up there for apochromatic correction too.
akpo.ca wrote:
Obviously 200mm and 135mm are very different lenses, on the same camera. I don't know why you would be comparing the two? 120mm f/4 (a 1:1 macro, ok cool) and a 135/2 comparison would be fairer...
Let me know if you have full resolution jpgs of the 200/2.8 or 120/4 apo/fa to enlighten us with. I have 135/2 full resolution gfx samples to show you. Wide open. Infinity focus or portrait distance. Let me know when you are ready.
The reason to compare the Mamiya 200 f/2.8 on the GFX to a FF 135mm f/2 lens is that cropped to 4 X 3 or squarer the Mamiya has the field of view and the depth of field equivalent of a 146 f/2 lens on a FF 35mm camera. So if you want to compare across formats it is a very relevant comparison. And yes a lens like the Zeiss 135 f/2 APO does very well on the GFX and if you look at the first post in this thread you can see my samples with that lens. I am not disputing that works well. I do think the Zeiss Contax 645 f/4 Macro is very much in the same league and has very similar rending, which isn't surprising given that they have similar MTFs similar heritage at Zeiss. When I have time I will post samples demonstrating that.
Don't think that the R 280/4 opens up to f/2.8 as well. Sure, many vintage optics are great. That's not the point of my generalization. Modern lenses tend to be better performers than their vintage counterparts. Regarding the 180/2.8, the 200VR and VRII from Nikon is apochromatic. Pretty sure the 200L IS from Canon is also up there for apochromatic correction too.
Don't think that the R 280/4 opens up to f/2.8 as well. Sure, many vintage optics are great. That's not the point of my generalization. Modern lenses tend to be better performers than their vintage counterparts. Regarding the 180/2.8, the 200VR and VRII from Nikon are apochromatic. Pretty sure the 200L IS from Canon is also up there for apochromatic correction too.
You do realize that the MTF that you posted is theoretical and in no way indicative of performance. You need to correct for diffraction and this theoretical MTF does not. It really tells us nothing. The point that many people are making is that your generalization is an over generalization and much to much of a simplification. I see Stephen was faster than I was in responding, but we make the same point.
Of course MTF charts are theoretical. That's why I mentioned decentering affecting performance, and iq definitely varies across different samples of the same lens. My $400 135/2 is pin sharp on a GFX, my 120/4 FAs are not. If you want to compare lenses please compare similar focal lengths and apertures. This is a thread about adapting lenses to the GFX, no?
I still have those samples if you're interested in seeing them. Let's forget about MTF charts.
akpo.ca wrote:
Of course MTF charts are theoretical. That's why I mentioned decentering affecting performance and iq varies across different samples of the same lens. My $400 135/2 is pin sharp on a GFX, my 120/4 FAs are not. If you want to compare lenses please compare similar focal lengths and apertures. This is a thread about adapting lenses to the GFX, no?
Wrong again. Not all MTF charts are theoretical. The one's produced by Zeiss are for sure made from real tests with production copies of their lenses. More recently the Lens Rentals blog does tests of many lens on the same sort of equipment that Zeiss uses. The old photodo site also had tests with that same sort of equipment. It isn't clear whether Leica MTFs are from actual lens or not, but their MTF charts match the Photodo site quite closely, so if they aren't actual tests they seem to mirror them well. The Canon MTF are theoretical and do not account for at least one huge factor in real testing--diffraction--so they are not informative of real world performance.
As far as what lenses to compare, I think a lot of different tests are relevant when considering adapting lenses to the GFX. It certainly is relevant to a lot of folks how adapted lenses on the GFX compare to the know performance of excellent lenses on the FF 35mm. I started this thread and when I did I had such comparison in mind, so I think they certainly fall within the scope of this thread. Of course comparing different lenses on the GFX also makes a lot of sense for this thread as well. Both types of comparison, in my view, should be welcome and I don't need you to tell me what types of comparisons can and cannot be made here. Let's welcome all such comparisons. So by all mean post your samples. We would love to see what you have shot.
I don't have the MTF charts for those two lenses either, and in fact I never owned the SMC Pentax-A 35/2.8 for 35mm film. However, I did own the SMC Pentax-A 645 35/3.5 and a variety of similar vintage 35mm focal length lenses designed for film. And I tested them against an ISO 12233 chart when I used them on a Sony A7R. And here's the thing, the medium format lens resolved more in the centre, and everywhere in the image, than any of the lenses designed for 35mm film.
My other SMC Pentax-A 645 lenses performed similarly in my testing on ISO 12233 charts when compared to the same focal length on lenses designed for 35mm film. Therefore, I can say with confidence that the SMC Pentax-A lenses I owned were sharper on a Sony A7R than 35mm film format lenses of the same focal length.
Here's the thing that still seems inconceivable to me: I simply can't imagine lens designers from Pentax, Hasselblad, Mamiya or anyone else making professional grade medium format lenses who CAN make them resolve as many line pairs per mm as a lens for 35mm film NOT doing that -- in other words, deliberately making their professional grade medium format lenses less sharp than they could be. What on earth would the rationale be? It's a professional lens that already likely costs more than the film version -- sometimes vastly more. My own testing says they didn't (or at least not in any way that I can detect).
Cheers, Rob
akpo.ca wrote:
The smaller film formats are more demanding on resolution per square area than larger formats. Larger formats do not require lenses to be designed for such extreme resolution for a small area. The design requirements are different; medium format and large format lenses tend to go for image quality across the entire frame rather than for a limited area.
Therefore the generalization is that c-mount, industrial machine vision and m4/3 lenses are sharper and can resolve a higher lp/mm for the centre of the image circle compared to apsc lenses, which are sharper for their respective formats when compared to 35mm full frame lenses, which are sharper compared to 645/66/67/69 lenses, which are sharper compared to 3x4, 4x5, 5x7, 8x10 lenses. The point here is that although a smaller format lens will tend to be sharper, sometimes they are over-designed and cover a larger area quite nicely; that is to say the sharpness goes to the edge of the image circle. Some 35mm lenses will cover 44x33mm with amazing IQ to the edges wide open.
Recently developed lenses tend to be of smaller formats too, so they gain the advantage of optical advancements and higher requirements for digital sensors.
One can see this from MTF charts and if you want comparisons. I don't have a comparison between the two Pentax 35s, but if you have them both you can do it yourself. De-centering has a larger factor in softening up a corner or two when using larger formats, so take that into account. ...Show more →
akpo.ca wrote:
Diffraction happens at f/2.8 or /2.0 on a 35mm or 44x33mm system?
Yes, it affects the MTF values, but not what you will actually see in pictures. Notice the values in the MTF chart you posted suggest a 100% contrast at both 10 lp/mm and 30 lp/mm in the centre. That is not possible with real testing. Diffraction and other effects that aren't modeled in those Canon theoretical MTFs have a small but measurable effect and MTFs never measure at 100% contrast. So, please do not compare those theoretical values with real measured values. The theoretical values only give us a very loose and imprecise suggestion of what real measurements would look like.
Ok, let's compare real sample photographs then, instead of talking theoretical or measured MTFs. Most become moot in real life practice, especially corner performance, when decentering comes into play. Three samples, all 135mm f/2.0, shot wide open, and straight-out-of-camera-jpgs from the GFX. Cost me $400, has a theoretical MTF chart of being pretty much perfect on a 35mm system.
Infinity or near infinity:
www.dropbox.com/s/yxfjmb1o7ro8p29/DSCF2933.JPG?dl=0
Close portrait distance:
www.dropbox.com/s/ajvfad9wj5kuz79/DSCF8341.JPG?dl=0
Looking forward to seeing your superior 200/2.8APO or 120/4APO (which, imo, are both great lenses). Also I'd be more than interested in seeing how your 135/2.0 APO-sonnar performs vs my cheaper alternative.
I tend to agree akpo.ca's generalization. They are lots of exceptions but generally it is true that modern lens are cheaper and have better correction and lens designed for smaller format are designed to resolve much smaller detail because of modern day high pix density. Such as good olympus glass designed for m43.
There are many lengendary MF and 35mm which are exceptions, but most of lenses at those era will not be overdesigned for film.
I don't think worth the effort to illustrate with examples of those exceptions. As I think both party seems all understand the topic well.
I'm also looking forward to seeing if anybody have full resolution samples from the Leica R 280/4 or 180/2.8APO. I can easily grab a 200/2IS or 200/2VR along with a decent 300/2.8 to compare on the GFX.
First, I appreciate your point that the appeal will be different for different photographers doing different kinds of work. For example, in my case, for my landscape photography (though not for my street and travel photography) I often make use of long focal lengths — it is an element of my style. I would have to make some tremendous compromises in order to switch to the miniMF systems for that reason... and my photography would suffer. But for someone who doesn't have that need — for example someone well served by the existing Fujifilm lenses (or some of the third party MF lenses) — the appeal would be different.
Second, it is hard to imagine a meaningful test done in a better way. This was done by a person who had every reason to believe in the superiority of the 80MP Phase One system, into which he had invested a lot of money and a number of years of his photography. And he is — you'll just have to trust me on this (or PM me if you want to know a name) — a master printer, literally among the best today. The examples he shared incorporated his printing skills and, as such, provide an excellent real world test of what these systems are capable of.
Yes, Dan, I have taken his master printing class and he surely is a great fine art landscape photographer, master printer, and great guy. And pretty good pianist!
His test reminds of when Uwe Steinmueller(RIP) showed me 24x36 prints from a Phase One P40 and a Canon 1ds3. There was not much difference in resolution, just a bit better color.
I was shocked then. Scale that to today's 5dsr 50MP vs miniMF at 50MP I doubt you will see much resolution gain until very large prints. I saw an interesting article written by a photographer friend of Michael Reichman(RIP), where he compared shots taken with Phase One P40 using latest at the time Mamiya lenses vs Rodenstock HR lenses and it definitely showed that to get the most out of the P40 you needed the Rodenstock lenses.
Fast forward to today and I am sure you need the sharpeat mf lenses you can find to get the full resolution increase. I think you need to go to phase one 180 to see a resolution difference.
There are plenty of advantages miniMF has over FF like leaf shutter lenses and 4:3 aspect ratio which I like better for portaits so I can see why someone would want to shoot minimf.
So I agree with Steve and Dan!
akpo.ca wrote:
I'm also looking forward to seeing if anybody have full resolution samples from the Leica R 280/4 or 180/2.8APO. I can easily grab a 200/2IS or 200/2VR along with a decent 300/2.8 to compare on the GFX.
I have pics only on my Sony A7RII for both lenses, but am on vacation now and hoping to take eclipse pics tomorrow (with Leica 280/4 apo with extenders).
akpo.ca wrote:
Ok, let's compare real sample photographs then, instead of talking theoretical or measured MTFs. Most become moot in real life practice, especially corner performance, when decentering comes into play. Three samples, all 135mm f/2.0, shot wide open, and straight-out-of-camera-jpgs from the GFX. Cost me $400, has a theoretical MTF chart of being pretty much perfect on a 35mm system.
Infinity or near infinity:
www.dropbox.com/s/yxfjmb1o7ro8p29/DSCF2933.jpg?dl=0
Close portrait distance:
www.dropbox.com/s/ajvfad9wj5kuz79/DSCF8341.jpg?dl=0
Looking forward to seeing your superior 200/2.8APO or 120/4APO (which, imo, are both great lenses). Also I'd be more than interested in seeing how your 135/2.0 APO-sonnar performs vs my cheaper alternative.
Well those are interesting shots, but for my tastes the vignetting is pretty heavy in the corners. Of course that is a matter of taste and I don't know if you tried to correct it or even if you added it in post. These are nice shots, but they don't blow me away. Is this the Canon 135 f/2 or the Samyang 135 f/2? Both are very nice lenses, but for the GFX if vignetting hasn't been added, I personally would like a bit bigger image circle.
zhangyue wrote:
I tend to agree akpo.ca's generalization. They are lots of exceptions but generally it is true that modern lens are cheaper and have better correction and lens designed for smaller format are designed to resolve much smaller detail because of modern day high pix density. Such as good olympus glass designed for m43.
There are many lengendary MF and 35mm which are exceptions, but most of lenses at those era will not be overdesigned for film.
I don't think worth the effort to illustrate with examples of those exceptions. As I think both party seems all understand the topic well....Show more →
Said that way, I have no problem with the generalization. As long as we recognize there are lots of exceptions I am fine with the statement. I will provide samples as I find time to do so, but I would have done so anyway.
Steve Spencer wrote:
Well those are interesting shots, but for my tastes the vignetting is pretty heavy in the corners. Of course that is a matter of taste and I don't know if you tried to correct it or even if you added it in post. These are nice shots, but they don't blow me away. Is this the Canon 135 f/2 or the Samyang 135 f/2? Both are very nice lenses, but for the GFX if vignetting hasn't been added, I personally would like a bit bigger image circle.
I didn't correct for it, they are SOOC jpgs as labelled. I can correct for the vignetting by stopping down the lens a tad bit or fix it in post, but these are just samples to show performance of a centered/properly aligned and modern optic on medium format. I haven't seen any vintage 645/66/67 lenses as sharp as this, nor are they majority of them centered/aligned properly straight out of the box.
akpo.ca wrote:
I didn't correct for it, they are SOOC jpgs as labelled. I can correct for the vignetting by stopping down the lens a tad bit or fix it in post, but these are just samples to show performance of a centered/properly aligned and modern optic on medium format. I haven't seen any vintage 645/66/67 lenses as sharp as this, nor are they majority of them centered/aligned properly straight out of the box.
Again, I think there is a big overstatement here. The majority of 645/66/67 lenses aren't properly aligned straight out of the box? Where is your evidence for such a statement. Seems totally made up to me. Oh, and what lens did you use in those shots?
Sharpness is just one factor, rendering style is another important factor of lenses and just like a lot of us in this forum enjoy shooting older less sharp, lenses with character on our FF cameras, why can't people enjoy shooting older glass or non native lenses on miniMF? Seems like normal alt forum behavior to me.
