New 7.5mm f2 announced! - Venus Laowa

[ijm5012]

I agree. I honestly wish it was a 9-10mm. I suspect I'd be cropping most of the photos I took with this lens. I guess the center is very sharp even wide open and cropping from 7.5 to 10mm still leaves about 9MP from 16 or 11MP from 20. So that's very doable. Even more so if you have a camera with Hi-Res mode.

(I already defish and heavy crop most of my shots from the 7.5mm fisheye and they turn out pretty good, usually)

Yeah, I really think 9mm is the sweet-spot for a UWA prime lens, as it's substantially wider than 12mm, without being crazy wide like the 7.5 is. A no-compromise, 9mm f/2 lens would be pretty sweet IMO. I briefly looked at the 10.5 Nokton, but the performance of it isn't anything special until it's been stopped down, and 10.5mm doesn't offer enough wide angle for many of the times I need something wider than 12mm.

As for the resulting image size after cropping, it all comes down to final use. For web or small print, 9MP will be more than enough. However, if someone wants to do a larger print, then that's really pushing it. As you mentioned, there's the hi-res mode for those who own a camera that has the capability, and the opportunity to use it is correct, but the vast majority of people will be working with a 16MP sensor. Personally, I have a lot of "large" prints hanging up in my house (30x20 or 36x24), so losing that much native resolution really isn't an option for printing that big (without using hi-res). Therefore, the 7-14 PRO is of better use to me, since I can frame the shot in-camera to maximize the available resolution. But for many others, shooting a scene extra wide and then cropping as needed will be sufficient, and the smaller size & weight is worth it over one of the 7-14 zooms.

 

[tkbslc]

As for the resulting image size after cropping, it all comes down to final use. For web or small print, 9MP will be more than enough. However, if someone wants to do a larger print, then that's really pushing it.

Obviously for large prints, the more MP the merrier. But it's really not that bad. Pro cameras were 10-12MP not very long ago and many large prints were made. It requires a little more care, sure. (And if we were cropping to 9mm as you said above, we'd only drop from 16 to 11.1MP.)

TBH, though, I probably wouldn't be shooting m4/3 if my primary output medium was 40" prints.

 

[EarthQuake]

I love, absolutely love that it is 7.5mm. I wish it were 7mm or better yet 6mm (I miss my FF Sigma 12-24mm, that was a gloriously wide lens). It's easier to crop in a bit than to "uncrop" (stitching is the only way to do that and it's problematic for various reasons), and I love shooting as wide as I possibly can. I've used the 9-18mm and 9mm isn't very appealing to me, my 7-14mm spends the vast majority of its life at the wide end. I realize that I'm probably not the norm here. Not only do I shoot a lot at 7mm, I probably use my 7-14mm about 30% of the time when I shoot, perhaps more.

Theres the 8.5/2.8 and 10.5/0.95 for those that don't want such a wide lens, though admittedly those are larger and likely significantly more expensive. Maybe someday Samyang will make an M43 version of the 12/2, by that I mean a 9/2 - shouldn't be too hard for them, if they think there is enough interest of course.

 

[tkbslc]

Theres the 8.5/2.8 and 10.5/0.95 for those that don't want such a wide lens, though admittedly those are larger and likely significantly more expensive. .

I was very excited about both of those lenses until the size, price and performance were revealed. They are big, expensive, and not very good.

I'm still really excited about the 7.5mm f2. I will probably buy one once the dust settles. I'd just have ordered up a 9.5mm f2 instead if I had my lens factory out back.

Maybe someday Samyang will make an M43 version of the 12/2, by that I mean a 9/2 - shouldn't be too hard for them, if they think there is enough interest of course.

I'd be so stoked if that happened. I've actually almost bought a cheap Sony or FUji so I can use the 12mm f2 as it was intended. But I've been just biding my time with defished 7.5mm shots.

 

[EarthQuake]

Venus optics confirms this lens will cost $500 or less.

Laowa 7.5mm MFT lens will cost a bit less than $500 – 43 Rumors

 

[Johnny The Greek]

Whoa! Depending on how much this costs and how well it performs, it might just see my fisheye heading out the door.

Wow, good job Laowa. Very impressed with everything this company has come out with so far, to be honest...

I doubt I would ever get rid of my fisheye, but yeah. I'd be hard pressed to find a way not to buy this.

 

[Egregius V]

I wouldn't replace my fisheye, either - except with another fisheye. That's a special kind of lens.

Because I love both primes and zooms and prefer them to be small and light, most of the f/2.8 zooms haven't interested me. I made an exception for the O. 7-14 f/2.8 when I was able to get one cheap. There are times when it's more capable than my O. 9-18 and 7.5mm fisheye. Its quality at f/2.8 sets the bar quite high for me. I look forward to finding out how well the Laowa does at f/2 with stars and architecture. The Photozone review isn't clear enough about that.

 

[Growltiger]

I love, absolutely love that it is 7.5mm. I wish it were 7mm or better yet 6mm (I miss my FF Sigma 12-24mm, that was a gloriously wide lens). It's easier to crop in a bit than to "uncrop" (stitching is the only way to do that and it's problematic for various reasons), and I love shooting as wide as I possibly can. I've used the 9-18mm and 9mm isn't very appealing to me, my 7-14mm spends the vast majority of its life at the wide end. I realize that I'm probably not the norm here. Not only do I shoot a lot at 7mm, I probably use my 7-14mm about 30% of the time when I shoot, perhaps more.

I am in complete agreement with you. I have the 7-14 Pro and the 9-18 and 90% of all photos with either lens are taken at the fully wide setting. 9mm is nowhere near true ultra-wide. I'm disappointed this lens is only 7.5, and like you would prefer 6.5 or wider. Does anyone know just how wide it would be possible to make a good rectilinear lens? Come on lens makers, go for 5mm!

 

[EarthQuake]

I am in complete agreement with you. I have the 7-14 Pro and the 9-18 and 90% of all photos with either lens are taken at the fully wide setting. 9mm is nowhere near true ultra-wide. I'm disappointed this lens is only 7.5, and like you would prefer 6.5 or wider. Does anyone know just how wide it would be possible to make a good rectilinear lens? Come on lens makers, go for 5mm!

Voigtlander makes a 10mm lens for 35mm, which is as far as I know, the widest rectilinear lens for any system. I can't think of any reason why it would be harder to design a 5mm lens for M43.

Folks often state that designing very short lenses is difficult, and while this is true it tends to be in relation to the system. For instance, designing a 10mm lens for 43 should be nowhere near the technical channel it is for 35mm.

 

[RichDesmond]

Voigtlander makes a 10mm lens for 35mm, which is as far as I know, the widest rectilinear lens for any system. I can't think of any reason why it would be harder to design a 5mm lens for M43.

Folks often state that designing very short lenses is difficult, and while this is true it tends to be in relation to the system. For instance, designing a 10mm lens for 43 should be nowhere near the technical channel it is for 35mm.

Especially given the short register distance of a mirrorless system.

I'm not the ultrawide enthusiast you two are, but I'd sure like a photo-oriented 8mm. These drone lenses aren't getting me too excited.

 

[wimg]

Voigtlander makes a 10mm lens for 35mm, which is as far as I know, the widest rectilinear lens for any system. I can't think of any reason why it would be harder to design a 5mm lens for M43.

Folks often state that designing very short lenses is difficult, and while this is true it tends to be in relation to the system. For instance, designing a 10mm lens for 43 should be nowhere near the technical channel it is for 35mm.

Designing a lens is one thing. However, it has to be manufactured as well, and effectively, because of the much higher resolution and smaller parts, tolerances need to be much smaller, by at least a factor 2, and likely by a factor 4. This is true both for the lenses as for the mechanics.

From that POV, it is much harder .

Kind regards, Wim

 

[EarthQuake]

Designing a lens is one thing. However, it has to be manufactured as well, and effectively, because of the much higher resolution and smaller parts, tolerances need to be much smaller, by at least a factor 2, and likely by a factor 4. This is true both for the lenses as for the mechanics.

From that POV, it is much harder .

Kind regards, Wim

Well, this isn't something that is mutually exclusive to ultrawides, it's a fact for all 43 lenses. Additionally, 43 lenses only have to cover an image circle 1/4th the size as well, so that should be much easier to design (and manufacture, less material is needed) for certain types of lenses (wide more so than telephoto). There are certainly trade offs.

I'm not sure how big of a concern small tolerances are to modern manufacturing these days. I mean, if major corporations had difficulty dealing with small tolerances, there would be no way we could fit such good cameras into our cell phones that are mass produced, and quite inexpensively at that. That's the logical continuation of the tolerance argument, cell phone lenses need to resolve many factors more than 43 to be usable, but I don't think many would argue that makes them proportionally harder to manufacture or design.

Anyway, my main point was that there is a common theory that lenses shorter than say 24mm are very difficult to design regardless of sensor size, but this is unequivocally false. The focal length of the lens without knowledge of the size of sensor or flange distance doesn't inform much at all about how difficult it will be to design or manufacture.

 

[wimg]

Well, this isn't something that is mutually exclusive to ultrawides, it's a fact for all 43 lenses. Additionally, 43 lenses only have to cover an image circle 1/4th the size as well, so that should be much easier to design (and manufacture, less material is needed) for certain types of lenses (wide more so than telephoto). There are certainly trade offs.

I'm not sure how big of a concern small tolerances are to modern manufacturing these days. I mean, if major corporations had difficulty dealing with small tolerances, there would be no way we could fit such good cameras into our cell phones that are mass produced, and quite inexpensively at that. That's the logical continuation of the tolerance argument, cell phone lenses need to resolve many factors more than 43 to be usable, but I don't think many would argue that makes them proportionally harder to manufacture or design.

Anyway, my main point was that there is a common theory that lenses shorter than say 24mm are very difficult to design regardless of sensor size, but this is unequivocally false. The focal length of the a lens without knowledge of the size of sensor or flange distance doesn't inform much at all about how difficult it will be to design or manufacture.

All is relative, really. To make excellent lenses for MFT, tolerances in lens elements need to be much, much better than their FF equivalents. Whereas you can get away with 1/4 lambda tolerances in curvature, that will be much less with regard to MFT, plus an even better distribution of curvature discrepancies, as there is much less tolerance available to have optical defects corrected.

As to cellular phone lenses: those are so short in FL, that they have an extremely large DoF, and as they also have a relatively small aperture, they can get away with very simple designs. It has started to change now, however, with the latest designs and patents on zoom lenses for cellular phones .

I guess to a degree that holds true for MFT lenses when you get to, say F/5.6 and smaller apertures, but once you get large apertures, it becomes increasingly difficult. With FF companies struggle to get very good resolution across the entire image field once you get above, say 30MP sensors, and the lenses as a result becoem extremely large in order to cope with optical and other errors. In the MFT world we are currently at a resolution of 80MP sensors compared to FF, so that makes it a increasingly more difficult, especially with wider aperture lenses, and even more so with (U)WA lenses. These do become retro-focus designs as well, because of their inherent size and the need to keep light projected on the sensor at a relative straight angle compared to the actual AoV.

Typical short flange distance lenses, i.e, film era type designs, just do not cut it anymore with relatively large, high MP sensors. This is another reason why the latest generation of large aperture MFT lenses are so big compared to their smaller aperture brethren. More glass is needed for corrective measures, etc., also lenghtening the lens, generally requiring rethinking of standard designs as well, making these lenses even more complex. If you look at these lenses, like the Nocticron, the Panny 12 F/1.4, the Oly 25 F/1.2, it is really quite amazing, IMO, how low the pricing of these lenses is, compared to their FF equivalents.

Just my 2c.

Kind regards, Wim

 

[Turbofrog]

I'm not sure how big of a concern small tolerances are to modern manufacturing these days. I mean, if major corporations had difficulty dealing with small tolerances, there would be no way we could fit such good cameras into our cell phones that are mass produced, and quite inexpensively at that. That's the logical continuation of the tolerance argument, cell phone lenses need to resolve many factors more than 43 to be usable, but I don't think many would argue that makes them proportionally harder to manufacture or design.

Actually, smartphone lenses are incredibly sophisticated. But they they are all medium-wide primes with a fixed aperture, so several of design challenges are removed. The physical size of smartphone lenses also gives the designers the opportunity to use fantastically complicated aspheric elements that optical engineers working on larger sensors could only dream of. On top of this, smartphones are manufactured in huge quantities, so very complicated designs can be implemented and still have sufficient margin to amortize the complex tooling.

Just take a look at a few smartphone lenses and tell me again that they are simple to design or manufacture:

Subscribe to see EXIF info for this image (if available)

Subscribe to see EXIF info for this image (if available)

When was the last time you saw elements that looked anything like that in your "real" camera?

 

[EarthQuake]

All is relative, really. To make excellent lenses for MFT, tolerances in lens elements need to be much, much better than their FF equivalents. Whereas you can get away with 1/4 lambda tolerances in curvature, that will be much less with regard to MFT, plus an even better distribution of curvature discrepancies, as there is much less tolerance available to have optical defects corrected.

As to cellular phone lenses: those are so short in FL, that they have an extremely large DoF, and as they also have a relatively small aperture, they can get away with very simple designs. It has started to change now, however, with the latest designs and patents on zoom lenses for cellular phones .

I guess to a degree that holds true for MFT lenses when you get to, say F/5.6 and smaller apertures, but once you get large apertures, it becomes increasingly difficult. With FF companies struggle to get very good resolution across the entire image field once you get above, say 30MP sensors, and the lenses as a result becoem extremely large in order to cope with optical and other errors. In the MFT world we are currently at a resolution of 80MP sensors compared to FF, so that makes it a increasingly more difficult, especially with wider aperture lenses, and even more so with (U)WA lenses. These do become retro-focus designs as well, because of their inherent size and the need to keep light projected on the sensor at a relative straight angle compared to the actual AoV.

Typical short flange distance lenses, i.e, film era type designs, just do not cut it anymore with relatively large, high MP sensors. This is another reason why the latest generation of large aperture MFT lenses are so big compared to their smaller aperture brethren. More glass is needed for corrective measures, etc., also lenghtening the lens, generally requiring rethinking of standard designs as well, making these lenses even more complex. If you look at these lenses, like the Nocticron, the Panny 12 F/1.4, the Oly 25 F/1.2, it is really quite amazing, IMO, how low the pricing of these lenses is, compared to their FF equivalents.

Just my 2c.

Kind regards, Wim

While I don't necessarily disagree with any of this, I think it's very hard to say how much affect it has on the actual design and production of the lenses. So, yes, a 43 lens has to resolve 4x the detail to match the resolution of a 35mm lens. Sure, that makes sense. However, I would imagine this is mostly a challenge when it comes to the initial design - which is computer aided and something that only needs to be done once, not every time a lens is built, so it's a fixed R&D cost. Manufacturing a lens to higher tolerances again may or may not be a significant cost. I highly doubt manufacturing at the higher tolerance level required is anything close to a proportional difference, ie, 4x higher tolerance = 4x higher cost. Even then, I think 35mm gear is already manufactured to very high tolerances and I would be surprised if the tolerance demands for 43 are really that much higher, I would expect the biggest difference would be in the design phase.

As to lens pricing, I don't know how much raw manufacturing costs plays into the equation, certainly it does, but lenses are priced to economic and market factors, they are not commodity goods.

 

[Turbofrog]

Manufacturing a lens to higher tolerances again may or may not be a significant cost. I highly doubt manufacturing at the higher tolerance level required is anything close to a proportional difference, ie, 4x higher tolerance = 4x higher cost.

You are correct, but perhaps not in the way you think. Typically at the higher ends of the spectrum, you see a point of diminishing returns kick in harshly. Often times it is nowhere near a proportional difference, and getting a small increase in tolerances costs tremendously more. So 4x higher tolerance = 10, 20, 100x higher cost. In many cases you need whole new kinds of machines to even measure those tolerances, let alone calibrate for them.

Obviously depends on your starting point. I'm not sure where the camera lens world lies in this spectrum of precision.

But remember that for M4/3, the pixels that we're trying to satisfactorily resolve here are ~0.004 mm. With a 50 degree FoV (conveniently, ~50mm equivalent) and a 16MP sensor, your angular tolerance is 0.011deg/pixel.

In the smartphone world camera lens and sensor modules are assembled together and then tested afterwards as a completed unit. The yield rate of acceptability is maybe only 1/3, and it is more expensive to try and recalibrate the system than to simply assemble a new module and hope it falls within the stringent tolerances required.

I agree with you that raw materials costs and the amount of required material likely have a relatively small impact on the overall cost of the finished product.

 

[EarthQuake]

Actually, smartphone lenses are incredibly sophisticated. But they they are all medium-wide primes with a fixed aperture, so several of design challenges are removed. The physical size of smartphone lenses also gives the designers the opportunity to use fantastically complicated aspheric elements that optical engineers working on larger sensors could only dream of. On top of this, smartphones are manufactured in huge quantities, so very complicated designs can be implemented and still have sufficient margin to amortize the complex tooling.

Just take a look at a few smartphone lenses and tell me again that they are simple to design or manufacture:

View attachment 517008

View attachment 517009

When was the last time you saw elements that looked anything like that in your "real" camera?

Whoa that is super cool, I had no idea.

 

[tkbslc]

So anyway..................

at under $500 this is looking like a pretty attractive purchase.

 

[MarkRyan]

at under $500 this is looking like a pretty attractive purchase.

Just need to be for sale...

 

[atarijedi]

Haha, at the risk of sounding especially obnoxious, I almost think 7.5mm is too wide. Would have been happier with something in the 9-10mm range for more day-to-day shooting. That said, this lens should be absolutely exceptional for astrophotography, assuming it can keep coma under control. Even if it has to be stopped down to f2.8, it's much smaller, lighter, and undoubtedly more affordably than the 7-14/2.8 it would be competing against in those circumstances.

There is always the Kowa Prominar 8.5mm f/2.8, also rectilinear, supposedly very sharp as well.