p.11 #2 · Official: Voigtlander 15mm f/4.5 III E-mount
Fred Miranda wrote:
BTW: Lightroom sharpening for both lenses was:
Amount: 40, Radius: 0.7, Detail: 25, Masking: 0
Both corrected for vignetting, No CA or distortion correction. (you can see CA very clearly in high contrast areas)
However LaCA is removed almost completely using LR's CA removal.
I'm hoping to get a E-mount (or even M-mount) as good as this one.
Fred have you considered asking Stephen Grandy to rise this with Cosina?
Another question: sometimes differences like this show up as presence or absence of peaking in the corners in viewfinder when focussed on a contrasty subject which, if I it does would make it easy to examine a few copies at a bricks and mortar store. Does it?
p.11 #3 · Official: Voigtlander 15mm f/4.5 III E-mount
Thanks for your efforts Fred and Guy. As Fred I was operation under the assumption that a lens with issues would have asymmetrical corner performance so I certainly have learnt something today. I should really start a lens rental company here in Germany
My blog-partner Bastian has a loaner from Cosinas's German distributor and while his a7s is certainly less demanding than the a7rii I see no performance issues similar to Fred's with his copy. Check out his Voigtlander 4.5/15 rolling review .
Personally my track record with CV lenses isn't the best either. I owned two different copies of the CV 4.5/15 I which both weren't too well centered and the CV 1.7/35 which among the many lenses I reviewed in 2015 was my favorite wasn't well centered either.
p.11 #4 · Official: Voigtlander 15mm f/4.5 III E-mount
I'd be careful using Guy's lens as a reference copy. It could be that during the modification process a lens element(s) was moved. Using a Dremel tool causes a lot of vibration and puts pressure to the front of the lens, that could have changed the position of a lens element(s). There are other lenses that have been known to have better flat field performance on the Sony cameras when the front element spacing is changed. That seems like a more likely scenario to me, than just the absence of the front hood.
Voigtlander came out with an emount version of the lens, but that does not mean it was optimized for the Sony thick cover glass. In fact, your sample data suggests that no change was made.
p.11 #6 · Official: Voigtlander 15mm f/4.5 III E-mount
This is interesting. I also use the III (M-version) on the A7RII and it certainly doesn't perform as good as choppy.
Fred: the corners of your copies don't seem to really improve when stopping down. Mine do improve when I stop down to F8-F11 (but are still so-so).
I'm going to do some tests too.
p.11 #7 · Official: Voigtlander 15mm f/4.5 III E-mount
My mkIII version is sharper into the corners than my FE16-35 is and WATE was. Basically I know I can stick the lens on f8 focus at the 10' mark and everything will be in focus and the corners sharp to a degree that I am happy with
p.11 #8 · Official: Voigtlander 15mm f/4.5 III E-mount
Viramati wrote:
My mkIII version is sharper into the corners than my FE16-35 is and WATE was. Basically I know I can stick the lens on f8 focus at the 10' mark and everything will be in focus and the corners sharp to a degree that I am happy with
Hi David, I was thinking of your lens mkIII.
How do you think does it compare to choppy in the corners, in particular at f/4.5?
TIA.
p.11 #9 · Official: Voigtlander 15mm f/4.5 III E-mount
I completely lost some files. I bought Choppy Jun29th and modified it Sept 19th. In between that time i tested it and drove across country with it with hood. I can't find the raws but I can find the images in my GetDPI gallery which i posted of the funky cars in this thread. Im going to see if i can find a extreme corner for any clue. Not having the raw is REALLY pissing me off.
p.11 #10 · Official: Voigtlander 15mm f/4.5 III E-mount
Okay after looking at least of all the jpegs. Choppy looked good from day 1. So not a hood vignette issue. Just shitty IQ that we are seeing. Im sorry it just makes me mad, we should NOT have to do this kind of testing to find good copies in the industry. I don't want to make this a bitch session as that gets us nowhere but Voightlander needs to be told they have a issue.
And they should send Fred a lens for free for all his trouble. Im not joking either.
p.11 #11 · Official: Voigtlander 15mm f/4.5 III E-mount
GMPhotography wrote:
we should NOT have to do this kind of testing to find good copies in the industry.
I agree with you, but there is another possibility. It's quite possible that the copies Fred tested would be considered good by Voigtlander. In other words, they may perform within specs. And your copy may be an unusual exception. I've heard stories of individual cars that seem magically reliable, due to accidents of manufacturing tolerances working in its favor. Could be the same sort of thing.
p.11 #14 · Official: Voigtlander 15mm f/4.5 III E-mount
GMPhotography wrote:
Okay after looking at least of all the jpegs. Choppy looked good from day 1. So not a hood vignette issue. Just shitty IQ that we are seeing. Im sorry it just makes me mad, we should NOT have to do this kind of testing to find good copies in the industry. I don't want to make this a bitch session as that gets us nowhere but Voightlander needs to be told they have a issue.
And they should send Fred a lens for free for all his trouble. Im not joking either.
Maybe that's true, Guy, but maybe it's not.
What I mean is that they are producing lenses to within a tolerance that they set.
One possibility is that they *could* find a way to manufacture them to a higher tolerance where most would be like Choppy. But that might cost much more: they might have to charge two or three times the price (because a relatively small rise in manufacturing costs gives rise to a small rise in price, which reduces the market so you have to raise the price again to find an equilibrium - it's why we have to pay so much more for incremental improvements)
At their current tolerance maybe most are like Fred's (which he was fairly happy with) and just a few are terrible (like one I borrowed) and just a few are like Choppy.
But of course this is the kind of commercial and engineering info we will never know...
p.11 #15 · Official: Voigtlander 15mm f/4.5 III E-mount
pdmphoto wrote:
Voigtlander came out with an emount version of the lens, but that does not mean it was optimized for the Sony thick cover glass. In fact, your sample data suggests that no change was made.
Based on a few tests I saw last year, it seemed the M-mount version performed slightly better adapted to stock Sony cameras than on Leica M digital, which would lead me to believe it's already designed to account for this factor. It could be that the optimization is a compromise somewhere between best performance on Sony vs. Leica. I.e. for a 1.5mm cover glass rather than 1mm for Leica or 2mm for Sony. But this is just a guess.
I had a re-look at my 15III tests done on Leica M240 and my results are closer to Fred's than Choppy's. By no means are any of my results as good as Choppy's in the image periphery for far distance images, but corners/edges do look good with near distance objects in infinity scenes. This leads me to believe there might be a touch of field curvature (towards the camera) at play with this lens on Leica M digital cameras. My copy also loses a touch of resolution in the left mid zone area, implying it might have a centering issue (though could maybe be camera mount related?), whereas my 15vII is OK in this region in side-by-side tests.
rscheffler wrote:
The cynical side of me thinks this is one reason Cosina does not publish MTFs for Voigtlander lenses.... It leaves individual lens owners somewhat in the dark in respect to a baseline level of expected performance, therefore each of us is more likely to accept a given lens's out of the box performance.
My impression about Voigtlander lenses has also been that they're somewhere between Samyang and Zeiss in respect to build quality and copy variation. The latest Voigtlander designs seem better and much more consistent in this regard, perhaps a benefit of their collaboration with Zeiss in building many of the Zeiss Z* lenses... but if you look at some of their older designs, the discussion happening here isn't anything new. For example, the 21/4 rangefinder lens is a rather old one and a case of copy variation lottery. It's possible to get a pretty good one, but requires testing multiple copies.
With the 15III it probably boils down to whether or not you can live with sub-par extreme corner performance most of the time as a tradeoff against other points in favor of the lens, such as its price, size, etc. ...Show more →
Fred Miranda wrote:
Ron,
Yesterday I would agree with you but today I totally changed my mind about this lens. It took me 4 copies to see how well this lens can perform. I've been testing Guy's lens today and the results surprised me.
His lens has high resolution and contrast from center to the very corners even wide open (f/4.5).
Sweet-spot is really f/5.6 with a bump in micro-contrast throughout the entire frame. There is no detectable field curvature and that is really surprising for such wide angle lens.
The other 3 copies I tested (Including the new E-mount) were terrible at the very corners. So much astigmatism and low contrast. I was always vocal about this. But if Guy's lens is representative of how this lens really performs, I definitely have been testing defective lenses until now. It's perplexing that my bad copies are centered. This is something I have not seeing before. Either there is copy variation or I'm very unlucky.
It competes neck to neck with my Distagon 15/2.8 at f/5.6 and f/8.
There are some issues though. High vignetting even at f/5.6. Lateral CA is also an issue at the corners and stopping down does not help at all. I consider vignetting and CA something minor though as it can be fixed with LR's custom profile for it but it's a negative for sure.
There are so many positives. Incredible resolution center to corner even wide open. Strong flare resistance and beautiful sunstar (subjective). The sunstar is similar to the Loxias and I particularly love it. It is strong with a nice shape even at f/8. Gotta love the 10-blade aperture.
I will post more crops tonight or tomorrow and update this thread. I got some better light here in San Clemente. This lens is a must buy.
Now, I have not seen a good E-mount version yet. The one I received is definitely a lemon and it's going back. I hope there are good copies out there because I want one! ...Show more →
Fred, I'm looking forward to how your replacement copy performs. At the moment my cynicism leads me to believe that Choppy is an exception and that the majority of the 15vIII copies land somewhere between it and yours.
p.11 #16 · Official: Voigtlander 15mm f/4.5 III E-mount
I think lens variation is probably the most frustrating aspect of photography for me (from an equipment perspective.) I realize the trade off would most likely involve higher cost, but viewing my delivery today (assuming USPS actually delivers it) with apprehension just seems wrong.
Part of what undermines the argument that it's too costly to produce to a certain tolerance is the apparent consistency with these "corner under performing" lenses. If they can be made to consistently under perform, the logical (but possibly still flawed) argument could be made that they should be able to make them consistently closer to Choppy's performance.
p.11 #18 · Official: Voigtlander 15mm f/4.5 III E-mount
k-h.a.w wrote:
Why doesn't CV work like the chipmakers?
Sort the lenses after testing and sell the better ones for a higher price.
Probably because it would raise the price for all lenses due to the need to test all of them. My suspicion is for the price point Voigtlander currently hits, they do fairly minimal testing and instead rely on a certain degree of assembly precision that assumes most will be good copies. I.e. let the end user do the time consuming testing. A minority of users will do such testing, and a minority of those will probably be unhappy with their copy and exchange it for another one. Whether or not this is the right approach is of course debatable and depends on your perspective. It's not what one as a photographer would want, but from a business perspective, might be logical.
In any case, if indeed it's found that there's significant copy variation, Cosina should be criticized for it. It could be that up until now they could more or less get away with it because the M-mount market was fairly small, scattered and possibly less technically inclined. Now that they're making native Sony mount lenses, they're selling to many experienced, very technically demanding photographers with high expectations.
p.11 #19 · Official: Voigtlander 15mm f/4.5 III E-mount
Very significant differences between the 2 lenses and I no doubt believe what I see. Forgive me for not reading every post here, but was the hard stop infinity confirmed in the corners on each lens?
p.11 #20 · Official: Voigtlander 15mm f/4.5 III E-mount
Viramati wrote:
My mkIII version is sharper into the corners than my FE16-35 is and WATE was. Basically I know I can stick the lens on f8 focus at the 10' mark and everything will be in focus and the corners sharp to a degree that I am happy with
There were always forum members insisting their CV 15/4.5 III lenses were excellent even towards the corners. I used to think...."what's wrong with these people!"
You were one of them Viramati and I believe you probably got a good copy as well.
A while back I posted crops comparing it to the TS-E 17 and FE 16-35/4 and the CV could not even compete. It was excellent in the center and mid-field areas but starting at the corners towards the edges of the frame, it would fall apart.
Guy's copy performs way better than the my 16-35/4 @16 and similar to TS-E 17/4 and that's a wow for such tiny rangefinder adapted lens. We all thought it was because of distance. Now I know it's not the case! It's really copy variance.
The only optical compromise was higher CA and vignetting to keep the lens compact. I can live with that!
