Nokton 25mm f/0.95 for micro 4/3rds...

This is really quite something. Like having a 35mm f/1.25 lens for 1.5x crop. Now if only you can have it in AF...

When I saw the f0.95, 25mm on the txt. but when I read further and it is a m4/3 lens, I was like... oh well. I don't shoot m4/3. Cosina doesn't make AF lenses. So there won't be an AF ver any time soon. I do hope they can push the envelope and make a few more f0.95 lenses for the FF. -- View: original size

love_them_all wrote: I do hope they can push the envelope and make a few more f0.95 lenses for the FF.
Your wallet is hating you right now.

love_them_all wrote: When I saw the f0.95, 25mm on the txt. but when I read further and it is a m4/3 lens, I was like... oh well. I don't shoot m4/3. Cosina doesn't make AF lenses.
I don't know if they still do, but they made inexpensive AF lenses that were marketed under several names, including Kodak.
So there won't be an AF ver any time soon. I do hope they can push the envelope and make a few more f0.95 lenses for the FF.
A big part of what makes that particular lens possible is the short back focus, the distance from rear element to sensor. Right now, FF lenses are constrained by SLR mirrors, they need at least 39mm from the rear element to the sensor. It is possible to build very fast lenses, given that constraint: Canon made a 50mm f1.0 about 20 years ago. But the end results tend to not be very impressive. The 50mm f0.95 Leica and f1.1 Cosina, both rangefinder lenses, are much more impressive than the 50mm f1.0 Canon SLR lens was.

Joseph S Wisniewski wrote: A big part of what makes that particular lens possible is the short back focus, the distance from rear element to sensor. Right now, FF lenses are constrained by SLR mirrors, they need at least 39mm from the rear element to the sensor.
I meant making an FF 0.95 RF lens. Sooner or later there will be other FF mirorrless to compete with the M9.
It is possible to build very fast lenses, given that constraint: Canon made a 50mm f1.0 about 20 years ago. But the end results tend to not be very impressive.
Canon also made a 50/0.95 for the RF. Their 50/1.2 RF from the 50's is also surprisingly good on the NEX.
The 50mm f0.95 Leica and f1.1 Cosina, both rangefinder lenses, are much more impressive than the 50mm f1.0 Canon SLR lens was.
The Canon EF 50/1.0 was not as sharp as the EF 50/1.2 wide open. It was such a heavy and expensive lens. The Cosina is good bet, but 1.1 is more like a 1.2, not exotic as an under f1.0 lens. The Leica, oh well, always slightly better but the marginal return is at a huge jump in cost...

Joe0Bloggs wrote: This is really quite something. Like having a 35mm f/1.25 lens for 1.5x crop. Now if only you can have it in AF...
Well, there's the issue of how well an f0.95 lens even works on digital (more on that, in a minute). But I've already got a 30mm f1.4 for my 1.5x crop. Yes, it's 1/6 stop slower, but it's also a better focal length. 25mm on four thirds is as bad as 50mm on full frame, 1.15x the image diagonal. That's a compromise between what people actually wanted (0.8-1x the diagonal) and what the camera companies could make easily (1.35x the diagonal, like those 58mm f1.4 lenses that people hated back in the late 50s). Look at rangefinder lenses, where there's no constraints on lens designs from the reflex mirror. You can have a choice of 35, 40, 45, and 50mm normals. If I still had a Bessa, my normal would probably be a 40mm f1.4 Voigtlander. Now, the issue that's been bugging me about this lens. I remember shooting a borrowed Canon 50mm f1.0 wide open a few years ago. It produced some very strange results. f1.0 is a huge angle, +/-27 degrees from perpendicular. It exceeded the acceptance angle of the microlenses of the particular camera involved (I think it was a 1Ds II) and the circles of confusion of the out of focus parts of the image became "rectangles of confusion". I don't have angle of acceptance data from some of the new micro four thirds cameras, but I've run enough goniometer table plots from the current Nikon and Canon cameras to know what modern sensors look like. I can't picture Panasonic pulling off wider acceptance angles while going to smaller pixel pitches, so I'm betting we see some strange things in the specular highlights from a wide open f0.95 lens. f1.4 doesn't seem to be much of a problem, +/-19 degrees from perpendicular is easier to deal with.

Joseph S Wisniewski wrote: I remember shooting a borrowed Canon 50mm f1.0 wide open a few years ago. It produced some very strange results. f1.0 is a huge angle, +/-27 degrees from perpendicular.
Could this be the explanation of why Olympus didn't venture wider than f/2 with their "digital" Zuikos?

I remember shooting a borrowed Canon 50mm f1.0 wide open a few years ago. It produced some very strange results. f1.0 is a huge angle, +/-27 degrees from perpendicular.
Could this be the explanation of why Olympus didn't venture wider than f/2 with their "digital" Zuikos?
Since there was a Leicasonic 1:1.4 lens which performed superbly, I think it's fair to say that this was just another one of Olympus' eccentricities. We'll probably never know what it was all about. I'm just hoping we'll finally see a non-Cosina 1:1.4 in the M43 system. The quest for an AF 1:1.4 lens goes on... -- http://flickr.com/photos/iskender

Fotogejst wrote:
Joseph S Wisniewski wrote: I remember shooting a borrowed Canon 50mm f1.0 wide open a few years ago. It produced some very strange results. f1.0 is a huge angle, +/-27 degrees from perpendicular.
Could this be the explanation of why Olympus didn't venture wider than f/2 with their "digital" Zuikos?
I'd say no, for two reasons. First, the datasheets on the first sensor Oly used, a Kodak KAF series whos number I've forgotten, had acceptance angle curves for the microlenses, and they could easily accommodate f1.4 lenses. That particular sensor had one of the widest acceptance angles in the industry. It also didn't have the "telecentricity" requirement Oly made such a big deal out of. We'll come back to that. Second, and more important, Olympus's f2.0 lenses, the 35-100mm f2.0 and 14-35mm f2.0, are extremely conventional designs: they're a very generic 70-200mm f2.8 and 28-70mm f2.8 design, followed by 2x wide-converters. They did a little math to merge the front elements of the wide converter with the rear elements of the main lens. f2.8 lenses on a 2x wide converter should have resulted in f1.4 lenses, not f2.0 As far as I can tell, based on looking at their designs, and with my own experience designing a couple of wide converters, they ran into a common problem: it's very, very difficult to make a wide converter with a rear section wide enough to have an f1.4 exit pupil. My dreams of turning an 85mm f1.4 into a 58mm f0.9 were to come to naught. But I digress... I'm guessing that Oly found out that there were wide converter problems pretty late in the design and product planning cycles. Had they found out earlier that there was an f2.0 limit on the wide converters, they could have used smaller, but better optical quality, f4 lenses as the main lenses, instead of f2.8. Imagine the PR points that would have been worth: f2.0 lenses smaller and lighter than Nikon or Canon f2.8 lenses. Cash in on their already common practice of quoting equivalent focal length without quoting equivalent aperture. No sauce for the gander. So, where did this f2.0 limit come from? It's actually spelled out in Oly's four thirds patent, U.S. 6,910,814. It is not possible to maintain the 8 degree "near telecentricity" that Oly argues is necessary for digital cameras with lenses any faster than f2.0, given the particular ratios of lens mount size to sensor size and registration distance to lens mount size that Oly patented. But basically, the whole 8 degree telecentricity argument turned out to be a load of marketing fluff. As soon as other optical companies stopped laughing, they started making lenses that stomped all over it. Panasonic launched their 25mm f1.4. Sigma launched their 30mm and 50mm f1.4 in Oly mount. Canon, Nikon, and Sony had large stables of f1.4 (and even f1.2) lenses running on full frame sensors. The death blow? Leica got together with Kodak and they improved upon Kodak's original sensor design so that it not only didn't need any "telecentricity", but it actually performed well with rangefinder lenses, which are the opposite of telecentric, they have exit pupils as close as 28mm from the sensor. There are sensors that benefit from the use of near telecentric lenses. Point and shoot cameras using CCD sensors with electronic "snap" shutters have very small photosites, and therefore, fairly long focal length microlenses with restrictive angles. But those sensors were never used in a four thirds camera, and only used with limited success in APS DSLRs (Nikon D70, D50, D40).

Very interesting!

Joseph S Wisniewski wrote: Well, there's the issue of how well an f0.95 lens even works on digital (more on that, in a minute). Now, the issue that's been bugging me about this lens. I remember shooting a borrowed Canon 50mm f1.0 wide open a few years ago. It produced some very strange results. f1.0 is a huge angle, +/-27 degrees from perpendicular. It exceeded the acceptance angle of the microlenses of the particular camera involved (I think it was a 1Ds II) and the circles of confusion of the out of focus parts of the image became "rectangles of confusion".
We have lots of great example from the Leica 50 mm f/0.95 Noctilux on a FF M9. So, f/0.95 can work perfectly fine on digital. (Admittedly, the M9 is designed for wide acceptance angles via some offset in the outer microlenses and probably also in the general microlens design.)

Joseph S Wisniewski wrote: Yes, it's 1/6 stop slower, but it's also a better focal length. 25mm on four thirds is as bad as 50mm on full frame, 1.15x the image diagonal. That's a compromise between what people actually wanted (0.8-1x the diagonal) and what the camera companies could make easily (1.35x the diagonal, like those 58mm f1.4 lenses that people hated back in the late 50s).
While you appear to have some great knowledge there, I think that's a somewhat idealistic view of it. In reality I've noticed that different photographers look at the world in different ways. Photographers sometimes choose to use one prime as an all-round as we know, and in that kind of situation I've seen or heard of 28, 35, 40, 45, 50, 55 and 60 being used. Hell, some stick an UWA zoom on their camera and need little else, while others cannot find any use for such a lens. Clearly some will be put off by the 50 mm equivalent field of view of a M43 25mm. However, there will be no universal like or dislike. If you personally on the other hand dislike this field of view it's of course good that you acknowledge it - otherwise you could be throwing away a lot of money here!

Iskender wrote:
Joseph S Wisniewski wrote: Yes, it's 1/6 stop slower, but it's also a better focal length. 25mm on four thirds is as bad as 50mm on full frame, 1.15x the image diagonal. That's a compromise between what people actually wanted (0.8-1x the diagonal) and what the camera companies could make easily (1.35x the diagonal, like those 58mm f1.4 lenses that people hated back in the late 50s).
While you appear to have some great knowledge there, I think that's a somewhat idealistic view of it.
My apologies. I sometimes have a tendency to speak of statistical trends as if they were absolutes.
In reality I've noticed that different photographers look at the world in different ways. Photographers sometimes choose to use one prime as an all-round as we know, and in that kind of situation I've seen or heard of 28, 35, 40, 45, 50, 55 and 60 being used. Hell, some stick an UWA zoom on their camera and need little else, while others cannot find any use for such a lens.
That is true. However, it is also true that camera companies (like most companies) tend to succeed when they cater to what the highest percentage of customers prefer. This is why all square format cameras eventually failed in the market place. It suited some photographers' visions, but not enough to sustain the companies. Nikon, Topcon, etc. made 58mm f1.4 lenses. Perhaps a small percentage of customers favored that length, but a much larger percentage said that they wanted something shorter, and the camera companies complied, within the limits of the technology of the time. So, the 50mm went away, but you had Nikon with a 50mm f1.8, 1.4, and 1.2, and a low cost "series E" 50mm f1.4. And the 45mm f2.8. Had they the technology of our time, they would have gone shorter. Nikon promoted a 45mm f2.8 as an "ideal" focal length. It was what people wanted, but it was not compatible with the lens making technology of 1960 in higher speeds. Given year 2010 computing power, modern optical glasses, and aspehric elements, the normal of the 60s would most likely have been a 45mm instead of a 50mm.
Clearly some will be put off by the 50 mm equivalent field of view of a M43 25mm. However, there will be no universal like or dislike.
There is a universal trend, and it's based on psychophysics, the science of human sensor perception. But it's funny you should use the word "universal". I also use the word. For years, I have been describing the focal length range from 50-80mm as "the universal hole". Look at any camera company's lens lineup from the glory days of primes. You have tremendous variety in wides, a progression in 20-30% increments, 20, 24, 28, 35, 50mm, and typically 2-4 different designs available at each of those focal lengths. Then a 70% leap over the "universal hole", and the 20-30% increments pick up again in the portrait range: 85, 105, 135, 180mm. The only things in the "universal hole" were a few macro lenses, like a 55m Nikon, and the scientific lenses like the 58mm f1.2 aspherics from Nikon and Canon. Nothing that photographers used to satisfy a "vision".
If you personally on the other hand dislike this field of view it's of course good that you acknowledge it - otherwise you could be throwing away a lot of money here!
As I pointed out, it's not me, it's the majority. I'm not throwing away anything, as I already have an acceptable light camera normal for my use. And an acceptable heavy camera normal. One of my favorites is a Voigtlander 40mm f2.0 SL II on a Nikon D3. You can see that trend in the evolution of the "normal zoom". The current standard is the 24-70mm f2.8 on FF. That's centered around 41mm, very close to the 43mm image diagonal, and very far from 50mm. (24mm is 1/1.71 of 41mm, 70mm is 1.71x 41mm). 17-50mm on DX (1.5x crop) is centered around 29mm,and that's exactly the sensor diagonal. Cosina are literally the "party animals" of the optical design community. They're doing weird things that only make some sense in the context of smaller sensors. They brought back the 58mm f1.4 Topcor, in Nikon and Pentax mounts, because 58mm might suck on a FF, but on a 1.5x crop camera, it's an 87mm equivalent, and leaps over the universal hole into an area that large numbers of photographers actually like. That's what they did with the 25mm f1.4, they modernized a classic cine 25mm f0.95 like a Berthiot Cinor (which will cover four thirds). wizfaq 25mm f0.95

Joe0Bloggs wrote: This is really quite something. Like having a 35mm f/1.25 lens for 1.5x crop. Now if only you can have it in AF...
I'm trying to find out more about how this lens compares to classic 25mm f0.95 cine and video lenses that people have been adapting to four thirds and micro four thirds cameras. Here's a great thread on another forum. http://forum.getdpi.com/forum/showthread.php?t=7389 Someone did some comparison shots between the Angenieux 'Type M1', the Carl Meyer 'Moviar' (an Angenieux copy), and the Soligor 'ITV'. Others joined in with a couple of other 25mm f0.95 lenses, including a Senko and a Cosmicar. Oddly enough, the Cosina price is right in the ballpark, a lot of 25mm f0.95 lenses tend to go for $800-1000.

Add new comment

Image
More information
  • Files must be less than 2 MB.
  • Allowed file types: png gif jpg jpeg.
Attachment
More information
  • Files must be less than 2 MB.
  • Allowed file types: zip rar.
By submitting this form, you accept the Mollom privacy policy.