The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Camera Movements and RetroFocus Lenses

Lobalobo

Member
It was interesting to read the discussion of the new Phase One IQ3 100mp back for technical cameras. Doug Peterson posted a number of images designed to answer these questions, he said: How does the new full frame CMOS perform when attached to a technical camera? How does it perform when pushed to the edges of the image circle? How does it compare to the CCD 60mp and 80mp sensors in the same scenario? Is there a noticeable change in Dynamic Range and bit depth? I take it from the responses that the answers were mixed, with some photographers arguing that trying to use significant movements with a wide-angle lens on digital back, particularly a full-frame CMOS back, perhaps, is like trying to fit a square peg in a round hole. If this is right, I wonder whether there are photographers who have considered a largely different direction for movements, using the Hasselblad tilt-shift apparatus (despite its 1.5x magnification) with, say, a 28mm H lens. Would this reduce the angle of light hitting the sensor by more than, say, a Rodenstock 40mm retrofocus lens, thus minimizing cross-talk and color problems? Would more be gained by this than would be lost in the presumably inferior quality of the DSLR versus large format lens, which could be used on something like an Alpa A Series camera when movements aren't required? Or are the advantages of a technical camera largely lost when using even a full-frame DSLR? Just curious.
 

torger

Active member
You can replace movements wide angles with ultra-wides and cropping (and optionally keystone correction). For a typical wide angle and typical movement range you lose about 50% resolution. If you have say 200 megapixels and an ultra-sharp wide you will still get output exceeding most/all of today's systems.

I think this is the likely future scenario in a 10 year period, more likely than we will see CMOS with wide angular response or new tech cam lens ranges.

Hasselblad's tilt-shift adapter is elegant, has no color shift problems, movements get stored in the EXIF data and distortion is auto-corrected. But you can only shift in one direction (and tilt in the same) so it's pretty limited in terms of movements. It also assumes that you use the highest resolution back and the optical quality of the 1.5x teleconverter is adequate. If you instead just have a digital back with 1.5x smaller pixel size you of course get better image quality by just cropping. With the upcoming 100 megapixel back I think more and more people will consider cropping a better option than using the tilt-shift adapter.

Of course you can't do tilt with cropping, but tilt is less popular theses days and is often replaced by focus stacking.

For someone like me that uses a Linhof Techno and seven Digitar lenses the weight of all Hassy lenses to match those focal lengths would be totally unreasonable. My camera system is as close as you can get to a drop-in replacement of a large format field view camera, the shooting experience and technique is virtually the same. I don' t think this camera type will be relevant for new purchases in say 5-7 years. The tech cam genre won't survive in its current form with the highest end sensors not working well with the lens range, and if we look historically the current CMOS technology may be a constant for the coming 6-8 years. Don't expect a wide angular response sensor appear anytime soon.

The ultra-high resolution sensors also make cropping more and more relevant alternative to optical lens movements.

The future technical camera is more like an Alpa FPS with retrofocus lenses. However when/if wide angle lens performance of the normal Phase One XF or Hasselblad H series approaches or matches what you can get on the Alpa then I think the mainstream choice for architecture photographers will simply be the standard XF/H system the best ultrawide you can get and crop as needed.

I think the 60/80MP Dalsa CCDs are the last CCDs for digital backs, and I think the current Rodenstock Digaron lens line is final, there won't be any new designs. Possibly we'll see more like the Cambo Actar, some sort of SLR design adapted for a tech cam but it won't be relevant unless it performs better than the XF/H wides. So it shall be interesting to see for how long technical cameras with it's "technology freeze" will be considered to have a lead in image quality. As long as the CMOS stays at 100MP I think traditional tech cams are competitive. But when it's 150 - 200 MP, how many will then prefer a shifted 60MP CCD on a Rodenstock ahead of a cropped ultra-wide 150-200MP image on an SLR ultra-wide... we'll see.

To me seeing that the IQ3 100 perform about the same way as the old IQ250 on the tech wides was a major disappointment. Again forced into pushing squared pegs into round holes. It felt like this was the last chance to revitalize the genre, and it was blown. If the next CMOS technology generation has wide angular response, for example through BSI technology, it will probably be too late. Schneider is already ceasing production, and by then Rodenstock may have as well, which now when Sony is becoming the new Dalsa is in the same situation as Schneider Digitar range when Dalsa became the new Kodak.

I'd love to be more positive, but I really can't. With my own camera system I have no logical upgrade path, all the higher end systems compromise my current shooting experience. Actually moving to large format film seems more logical for me than going to high tech hell. At least until I've drum scanned and dust spotted a few negatives...
 
Last edited:

jagsiva

Active member
I don't believe Doug has published any of the tech cam samples yet. All samples to date are with the XF SLR camera.

There are however, samples from CI and Alpa. For these, there is definitely a quality hit in terms of color cast, corner sharpness, smearing, banding etc. The amount depends on the lens use, amount of movements, and scene conditions. Whether or not these are acceptable will ultimately depend on the use case and the artist.

For me, the gain in resolution over an 80MP CCD is negligible, but the the CMOS sensor does provide much better shadow recovery, less noise and a far more elegant workflow. I would gladly take these at the same level or higher output quality. But so far, even with 9.0.3, this does not appear to be the case. I went down the tech cam route first and foremost for output quality, so not quite willing to cut corners there. I do feel movements in camera, give my work a different look and feel. The LF glass is also better for the kind of work I do when compared to any SLR glass (MF or 35mm).

But all is not lost yet. Phase has a reputation of making subtle improvements over releases of software and firmware. For example, with C1 9.x, I see a significant improvement in LCC processing of CCD files. I am hopeful that this continues with the new CMOS 100MP back as well.

One thing that I like to see, as I'm sure others would too, is a statement from Phase on tech cam use with their 100MP back. Is it something they endorse as a company? Is it something they will continue invest R&D in to improve? The alternative would not mean the death of tech cams, but sadly, it would reduce it to a cottage industry of hacks, lens adapters, and workarounds.

I know that when the 80MP CCD came out, there were some challenges with some symmetrical lenses. Some of these were "officially" not recommended, if not by Phase, certainly by the dealer community. In the meantime, Phase made some improvements in software/firmware, and the solution we have now is very good. I hope there is a similar evolution of tech cam use on the 100MP back with an affirmative endorsement of such use by Phase.
 

Paul2660

Well-known member
Not having used a 100MP back, I can only speak from the Capture Integration series of tech camera testing that I have seen. I have used the 50MP chip extensively in the field, net the 2 chips respond very much the same way to my eyes with movements. On center, pretty much little no color cast, but with 5mm or more movements, you start to see heavy red/magenta cast on the shifts, which is exactly what the 50MP chip does, You always see some red cast on extreme shifts, (at least I do) even with CCD movements, but the Sony chip to me shows more.

There is also faint blotchy banding, that was also present in the 50MP chip, basically again exactly the same. Not micro lens ripple, which is much more definite in definition. This banding will only show up in a solid, like a blue sky or grey wall, (and in a B&W conversion). You can see it in the first Alpa tests of the building with the solid blue sky, if that image has not been pulled by Alpa.

Positives, the Phase One LCC, does get a lot or the red cast out on the Sony chip, in same cases, I found the LCC did better on the red cast, then with my 260 (again on blues). The Sony also has much great range in shadow recovery, at base ISO and past.

I am hoping that Phase will address the blotchy banding issue, but since it was never fixed on the 150/250/350, it may not be possible, so it's something to just be aware of.


Again, the LCC correction to my eyes, gets the job done on the color casting, (pretty amazing when you see just how much red is present in the LCC), and the shifts show much less noise on the shifted edges, all positives.

The CI shots are on their blog, a few entries down now. The 32mm Rodie did not have a CF on, and I feel it would have greatly helped on the shifts. The 40mm Rodie, does overall great even at 18mm of shift. They used fall on the 18mm, which avoided most of the hard edge vignetting created by the IC indicator.

Paul C
 
Last edited:

dougpeterson

Workshop Member
The tests you're referring to were not ours.

At DT we have only recently wrapped up our extensive testing resulting in more than 30gb of raw files to analyze covering more than 200 combinations of back, lens, and movement. We had support in the form of lens loans from Arca Swiss and one of ours clients so that we could test ten Schneider/Rodenstock lenses (as even we do not have all lens models in demo/rental at any given time).

Since I'm in LA this week with Phase One events and training scheduled every day of the week, and since I was in charge of the testing (and therefore am in the best position to lead the processing and posting thereof), it will likely be another week before the results are fully organized and ready to post publicly. In the meanwhile our clients are welcome to contact us for mini reports and subsets of the files based on their specific points of comparison and needs.

Or for those interested in using the IQ3 100mp with an SLR (e.g. XF or H5X) rather than tech camera you can parse through the 30gb of raws posted here: Phase One 100mp vs Canon/Nikon/Sony Raw File Still Life Shootout
 
Last edited:

Lobalobo

Member
Interesting responses. Sorry that I was confused by what CI had tested, was fooled by this site:
https://captureintegration.com/phase-one-iq3-100mp-technical-camera-testing/.

Returning to film was mentioned by one post. Imagine setting up a technical camera to capture a scene with some movements then recording the identical image twice, once with a digital back another swapping the digital back for a roll film holder then scanning. Wonder what the comparison would show.

Also curious as to why Sony, having won over-the-top reviews (to say the least) for its BSI full-frame 35mm sensor would not have used a similar sensor for its new 100mp full-frame MF digital sensor. The tech camera users would have been over the moon, I take it, and because the image quality Alpha 7Rii is so beloved, so would everyone else have been, I'd imagine.
 

torger

Active member
Also curious as to why Sony, having won over-the-top reviews (to say the least) for its BSI full-frame 35mm sensor would not have used a similar sensor for its new 100mp full-frame MF digital sensor. The tech camera users would have been over the moon, I take it, and because the image quality Alpha 7Rii is so beloved, so would everyone else have been, I'd imagine.
These are (somewhat educated) guesses: the BSI technology is not without challenges, lower yield and I think a tiny bit lower DR at base ISO, the gain is more at high ISO. The 15 stop DR would probably not have been possible at this point if they'd use a BSI technology. The largest reason I think is however that they can't scale up BSI to MFD size yet. It took quite a while to get it from the mobile phones and compacts to 24x36mm size.

I've also heard that the angular response of the A7r-II is not as good as many hoped for but I haven't seen any real tests. To make it excellent for wide angle tech cams you'd probably have to remove the microlenses as well.

If you really want to it's pretty easy to make a sensor that supports tech lenses, Kodak did that with theirs, just add some light shields and remove microlenses and don't make the pixels so damn small and you're done, but you need to sacrifice other performance metrics like some DR and high ISO performance and probably some pixel count, and I don't think the tech cam genre is big and important enough these days to support such trade-offs.
 

Stefan Steib

Active member
When I spoke with my partners the first few times about the HCam-B1 in 2009, it was clear to me that the traditional "real" wideangle Concept was hitting a wall with the then 60 Mpix already. The Canon TS-E´s were just introduced and so we decided to make a camera for this. The corner rayangles (of these techlenses)are simply not tolerable any more with 24mm and shorter(and don´t even exist) for MF.
But my Prediction is a bit more extreme: I predict multishot technology doing 8k 33Mpix with 100 Images a second as a sideeffect of video technology (on 24x36mm FF), take this and supersample/interpolate as many images as needed for nearly any wanted resolution.....
The 33Mpix are totally uncritical as we already know for most quality lenses, even with movements.

And this is my bet. We speak again in 5 years.

Greetings from right now Cape Town/Hout Bay
Stefan
 

torger

Active member
Film exists in parallel, and I think it will potentially survive indefinitely into the future in one way or another as it's low tech and there's a small but seemingly constant demand from artists etc to "go back to the roots" of photography. The really high end scanners etc may go extinct though so we may need to satisfy with lesser quality in the future if we shoot film.

I'm hoping for that different digital solutions could exist in parallel too. It surely will in many genres, I don't think a ultra-high-speed multi-shot camera will be used for typical hand-held studio work. I also think that view cameras will survive quite well into the future for product photography in the studio.

The wide angle tech cam space for architecture and landscape seems to be the most threatened. In order to stay unique they need the unique lens designs, and sensor technology is not making it easy for them. Making the lenses even more retrofocus than today I think will hurt some basic properties making them much too similar to standard wides on the XF or H systems.

All it takes is a sensor in a digital back that has wide angular response and traditional tech cam can live in parallel for decades even if all manufacturing of new lenses stops today. I'd love to see that happen to diversify the landscape a bit, so not only the current highest resolution system makes every other solution go extinct.

It's really not game over yet, and the IQ3 100 looks sort of usable, but the trend is very clear. For each new generation compatibility is worsened and you need to make further compromises. If you went from P45+ to P65+ you did compromises, and likewise from P65+ to IQ180, and now even further from IQ180 to IQ3 100. Each step may seem small, but it adds up.

If that high-speed multishot monster ultra-wide angle camera can simulate movements on the rear screen (through realtime cropping and keystoning) the creative process in the field won't be that hugely different though so I guess I'd live. From an artistic standpoint I have a romantic connection to the traditional "one shot frame" though so if a competitive traditional method would exist in parallel I'd probably use that. Film is not unthinkable for me, not in the field at least. The worst drawback of film as I see it is in the scanning and archiving part, plus that running cost is certainly not insignificant.
 
Talking about keystoning, or rather a lack of it, it seems Crewdson possibly used digital again for his new project (Cathedral of the Pines), which appears to involve a lot of trees that are standing nice and vertical!
 

torger

Active member
Talking about keystoning, or rather a lack of it, it seems Crewdson possibly used digital again for his new project (Cathedral of the Pines), which appears to involve a lot of trees that are standing nice and vertical!
I think I read somewhere that he used large format film for this project, but it could be mistaken. Anyway if using digital it's probably a tech cam just as for his previous work Sancutary, as seen here https://www.youtube.com/watch?v=Taf1EvZIY-Q so keystoning wouldn't be necessary.
 

Lobalobo

Member
Film exists in parallel, and I think it will potentially survive indefinitely into the future in one way or another as it's low tech and there's a small but seemingly constant demand from artists etc to "go back to the roots" of photography. ... From an artistic standpoint I have a romantic connection to the traditional "one shot frame" though so if a competitive traditional method would exist in parallel I'd probably use that. Film is not unthinkable for me, not in the field at least. The worst drawback of film as I see it is in the scanning and archiving part, plus that running cost is certainly not insignificant.
Wonder about whether medium format film could be used in conjunction with, say, an Alpa system set up for movements on a CMOS MF back. The idea would be to focus and compose using the back then, when (but only when) movements on a wide angle lens were too extreme for digital, to replace the digital back with a film back. A scanner would be needed, of course, but this setup would require only one camera and one set of lenses, as opposed to running parallel a 4x5 system for significant movements. The question would be, I guess, how much would be lost in image quality using medium format film as opposed to either a MF digital back or 4x5 film. Any thoughts?
 

ErikKaffehr

Well-known member
Hi Anders,

Jim Kasson pointed me to an article on CMOS development and there is a lot of interesting technology there. It is coming from cell phones. They work a lot on cross-talk issues, and also shallow pixels. Development on small sensors tend to migrate to larger sensors when it is beneficial.

One issue with MF is that the technical camera market is a small part of a small market. It is natural that sensors are developed for the larger MF DSLR market.

That may shift if someone introduced an affordable priced mirrorless design. Think Alpa FPS but at Pentax 645 prices, possibly combined with an EVF. I could very well imagine a company like Fuji making such a camera. Or, Phase One could look around in the cellar and find the blueprints for the Mamiya 7?

But, the camera needs to be affordable to create a significant market.

Here is Jim's article: http://blog.kasson.com/?p=12837

And here is the article he was referrng to: http://www.chipworks.com/sites/defa...-the-Art of Mainstream CMOS Image Sensors.pdf

Best regards
Erik


Film exists in parallel, and I think it will potentially survive indefinitely into the future in one way or another as it's low tech and there's a small but seemingly constant demand from artists etc to "go back to the roots" of photography. The really high end scanners etc may go extinct though so we may need to satisfy with lesser quality in the future if we shoot film.

I'm hoping for that different digital solutions could exist in parallel too. It surely will in many genres, I don't think a ultra-high-speed multi-shot camera will be used for typical hand-held studio work. I also think that view cameras will survive quite well into the future for product photography in the studio.

The wide angle tech cam space for architecture and landscape seems to be the most threatened. In order to stay unique they need the unique lens designs, and sensor technology is not making it easy for them. Making the lenses even more retrofocus than today I think will hurt some basic properties making them much too similar to standard wides on the XF or H systems.

All it takes is a sensor in a digital back that has wide angular response and traditional tech cam can live in parallel for decades even if all manufacturing of new lenses stops today. I'd love to see that happen to diversify the landscape a bit, so not only the current highest resolution system makes every other solution go extinct.

It's really not game over yet, and the IQ3 100 looks sort of usable, but the trend is very clear. For each new generation compatibility is worsened and you need to make further compromises. If you went from P45+ to P65+ you did compromises, and likewise from P65+ to IQ180, and now even further from IQ180 to IQ3 100. Each step may seem small, but it adds up.

If that high-speed multishot monster ultra-wide angle camera can simulate movements on the rear screen (through realtime cropping and keystoning) the creative process in the field won't be that hugely different though so I guess I'd live. From an artistic standpoint I have a romantic connection to the traditional "one shot frame" though so if a competitive traditional method would exist in parallel I'd probably use that. Film is not unthinkable for me, not in the field at least. The worst drawback of film as I see it is in the scanning and archiving part, plus that running cost is certainly not insignificant.
 

torger

Active member
Wonder about whether medium format film could be used in conjunction with, say, an Alpa system set up for movements on a CMOS MF back. The idea would be to focus and compose using the back then, when (but only when) movements on a wide angle lens were too extreme for digital, to replace the digital back with a film back. A scanner would be needed, of course, but this setup would require only one camera and one set of lenses, as opposed to running parallel a 4x5 system for significant movements. The question would be, I guess, how much would be lost in image quality using medium format film as opposed to either a MF digital back or 4x5 film. Any thoughts?
You'd use 645 film then I suppose. There's an interesting comparison here done in 2011:

http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html

using the best film scanning techniques. I'd say 645 film corresponds to 20-30 megapixel digital. Another alternative is then to have two digital backs, for example a P45+ or a CFV-50/H3DII-50/H4D-50/H4D-50 (the now discontinued versions with the Kodak 50MP) and drop in that when the CMOS back can't handle the movements. These older backs can be had for "peanut money" these days.

It would only be sane for someone that rarely shoot wides, otherwise its just too messy.
 

torger

Active member
And here is the article he was referrng to
Thanks for that link, I shall read it in depth later.

Indeed there's a lot unseen on which direction image sensors will evolve.

I think the main issue is that due to the tiny size of the wide angle tech cam segment one will not use the progress to optimize for that. For instance if you make shallower pixels you use that gain to be able to make smaller pixels to get higher pixel counts, rather than significantly increasing angular response. If pixels get so shallow and photo diodes so large that microlenses become obsolete then we're talking, but I haven't seen such designs. I doubt it will happen in near time as the photodiode cannot be too close to the neighbor as crosstalk (after capture) would then occur, and with the separation there's borders without light sensitivity, and then you add microlenses to improve ISO and reduce aliasing.

However it doesn't have to be excellent angular response, Rodenstock Digarons are not that demanding.

Today's consumer mirrorless cameras have considerably longer flange distance than tech cams, where the rear lens element is almost touching the sensor. There are gains to make with shorter flange distance, meaning you can have simpler lens designs that become sharper in the wide end, but they will also suffer from heavy vignetting and small aperture which is okay with the center-filter equipped tech cams, but I doubt mainstream cameras will make such tradeoffs -- so flange distance will still be longer than Rodenstock Digarons and lenses more retrofocus and less demanding on angular response, and then sensors are made to optimize other factors than that.

So I'm somewhat hopeful, but my guess is that the coming ten year period will be a desert when it comes to supporting these traditional lens designs, and as a result other alternatives will arise, possibly along the lines Stefan described above.
 

tjv

Active member
I used my Linhof Techno to shoot 6x7 / 6x9 film for years. It's brilliant for that purpose, assuming you have the lenses to cover the format with room for movements. I shifted to a Credo 60 because the cost of consumables and time to scan (with my Imacon 949), not to mention the time it takes to spot all the dust out of the scans, was driving me insane. I'll tell you one thing though, while digital is better in many ways–resolution mainly–it's pretty much a wash for me as to which I'd use given the resources and an assistant to do my grunt work in scanning form me.
 

torger

Active member
I used my Linhof Techno to shoot 6x7 / 6x9 film for years. It's brilliant for that purpose, assuming you have the lenses to cover the format with room for movements. I shifted to a Credo 60 because the cost of consumables and time to scan (with my Imacon 949), not to mention the time it takes to spot all the dust out of the scans, was driving me insane. I'll tell you one thing though, while digital is better in many ways–resolution mainly–it's pretty much a wash for me as to which I'd use given the resources and an assistant to do my grunt work in scanning form me.
I make romantic thoughts about film from time to time. And then I start thinking about scanning, dust spotting and archiving.

Therefore I hope to be able to shoot in a "film-like" way with digital in the future as well, just as I do today.
 
You'd use 645 film then I suppose. There's an interesting comparison here done in 2011:

http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html

using the best film scanning techniques. I'd say 645 film corresponds to 20-30 megapixel digital.
I think that's probably about right.

I've compared a number of my drum-scanned films, mainly 6x7 film off a Mamiya 7. The Mamiya 7 produces a lot of detail for prints (say) to 50" wide, and at a higher resolution than my 24mp Leica M240 for that size print. But 24mp digital is certainly higher rez than 35mm film, where grain clumps up quickly .... so yes, 645 film is perhaps around 20-30mp in terms of what is obviously recorded via a drum scan.

A problem with film is that some very fine detail gets lost (especially the lower-contrast detail), both from the scanning and also due to the film grain, even with films like Acros 100, Portra 160 or Provia 100. And acuity is better on digital, making it look like it has more resolution than perhaps is really the case.

I am now very tempted by the Leica S-E, however, given my trials with it suggest it beats a 120 film (drum scan) in terms of resolution for 50-60" prints, and obviously for acuity. Meaning the Leica S images really have some impressive "snap" compared to my Mamiya 7, but the Leica S' lens & sensor combo is ALSO able to produce really gracious and smoother images compared to 35mm FF digital.

It's probably likely that my romantic notion of film is still there because I have been using NON-medium format digital (I respond very differently with images made with the Leica S vs. M240) -- 35mm FF makes me yearn for 6x7 film, but I think i'd be done with it if I migrate to something like an S.
 
Last edited:

Lobalobo

Member
You'd use 645 film then I suppose. There's an interesting comparison here done in 2011:

http://static.timparkin.co.uk/static/tmp/cameratest-2/800px.html

using the best film scanning techniques. I'd say 645 film corresponds to 20-30 megapixel digital. Another alternative is then to have two digital backs, for example a P45+ or a CFV-50/H3DII-50/H4D-50/H4D-50 (the now discontinued versions with the Kodak 50MP) and drop in that when the CMOS back can't handle the movements. These older backs can be had for "peanut money" these days.

It would only be sane for someone that rarely shoot wides, otherwise its just too messy.
Thanks for the comparison and, yes, there seems to be quite a difference between 645 film on the one hand and 645 Digital or 4x5 film on the other. Currently, as a hobby, I shoot landscape and architecture on 4x5 film on a camera with few movements (save rise, which I use for architecture). I have long fantasized both about becoming proficient at camera movements and switching to a digital back, thus my curiosity and disappointment at the notion that these things are mutually incompatible. The modularity of supplementing a CMOS digital back with 645 film for significant movements seems so attractive, but from what the comparison shows, more might be lost than gained. Then again, maybe not--quite nice large prints can be made from 20-30 mp images. Thanks again.
 

jerome_m

Member
I wonder whether there are photographers who have considered a largely different direction for movements, using the Hasselblad tilt-shift apparatus (despite its 1.5x magnification) with, say, a 28mm H lens. Would this reduce the angle of light hitting the sensor by more than, say, a Rodenstock 40mm retrofocus lens, thus minimizing cross-talk and color problems?
Yes. The angle of light hitting the sensor is dependent of the position of the exit pupil and retrofocus designs move this pupil forward. The HTS, being a teleconverter, also improves that angle.

Would more be gained by this than would be lost in the presumably inferior quality of the DSLR versus large format lens, which could be used on something like an Alpa A Series camera when movements aren't required?
The HTS combined to the HCD 24 or HCD 28 is not really inferior to a symmetric lens.
 
Top