Sparrow
Veteran
em? no it doesn't ... anyway, many seem to confuse distortion with the normal planar projection as it is ... and I've not had much luck explaining the difference to them
Are you OK with lens corrections on Leica Q?
- Thread starter Avotius
- Start date
- Latest activity Latest activity:
- Replies 222
- Views 10K
em? no it doesn't ... anyway, many seem to confuse distortion with the normal planar projection as it is ... and I've not had much luck explaining the difference to them
Black
Photographer.
Cal,
I think I made the mistake of claiming to be a purist or something similar, once. Once. Got called out on it and I'll never do it again.
Its good to call people out on such matters. Character building and all that 😉
uhoh7
Veteran
In a forum which loves film so much I find the "who cares, good enough" attitude refreshing, if curious.
GaryLH
Veteran
Black
Photographer.
Calzone
Gear Whore #1
The images from the "Q" look pretty good to me and I own a 28 Cron (most used lens). All of my cameras utilize MF lenses, even my Nikon D3X, so autofocus would be new to me. The slight shutter delay can't be as bad as the Contax T3 I found by almost stepping on it in an abandoned part of Brooklyn where I use to live.
One of the reasons I own so many cameras is to keep me fresh.
Cal
One of the reasons I own so many cameras is to keep me fresh.
Cal
H
Hunter S Thompson
Guest
$4000 for a camera with a poorly designed lens. What a joke. It might be understandable if there were some benefit for the end user. Many lenses need optical correction when they are designed to be extremely small. That's a benefit for the end user.
Clearly Leica is skimping on manufacturing costs buy not correcting for the distortion. This is a test case. If successful, we'll see more and more manufacturers simply designing poorer lenses, at the same or higher cost, and then correcting for these poor designs in the digital realm.
This is a very very bad trend IMO.
Nope, not a test case at all. The cat is already well and truly out of the bag when it comes to in-camera correction. All the big name camera manufacturers have been doing it for some time. Even Leica is already doing it with its 6-bit coding on M-mount lenses. The camera reads the 6-bit coding so it knows what in-camera corrections to perform for the attached lens. Ain't no stopping it now, and it certainly doesn't mean the lens was poorly designed.
H
Hunter S Thompson
Guest
No, that's incorrect. Most lenses that need lens correction are either el-cheapo lenses, or designed for some other advantage like very small size.
Leica is doing it just to save on their own costs. Nothing more.
The lens is absolutely poorly designed. That's a fact. Who here would use this lens without any assistance from software?
Lss
Well-known
Sony RX1 does apply distortion correction. The barrel distortion is very noticeable without the correction. You can choose to not apply it, which can be useful with certain subjects.
I was hoping the Leica Q would also allow not applying the correction. From some of the images I have seen since, it seems the correction is truly an integral part of the system as the images otherwise show black corners. Even the image circle is thus optimized with the software correction in mind. Given this, I have no problem this option is not meant to be available to the user. (On the Sony camera, you do not get black corners.)
I have never bothered to compare the corners of the uncorrected and corrected RX1R image in the same file, although I have meant to do it. The corners of the corrected image have certainly been good enough for me.
Godfrey
somewhat colored
My Summilux-D 25mm f/1.4 ASPH for FourThirds SLR cameras, hardly small, hardly cheap, and a very very high quality performer, included lens correction metadata injected into the raw files when used on a Panasonic G1.
Results with a raw processor that honored the lens correction meant that all lateral chromatic aberration was completely eliminated even wide open (a small amount was visible wide open without the correction applied) and the entire image has a crisper, cleaner rendering feel. A small amount of barrel distortion was also corrected.
Lens correction is used on multimillion dollar space telescope camera lenses as well as by criminal investigation and intelligence agencies to improve the quality of the data they collect. On that basis, I'd say there's plenty of merit to be had from applying modern technology to the problems of best imaging.
A lens designed to be used with an embedded correction algorithm may well be designed to have more simple, easily corrected aberrations in order that there are fewer complex, difficult to remove aberrations, the end result being better overall performance.
So to say "The lens is absolutely poorly designed. That's a fact. ..." when it obviously produces lovely results from the work posted by every reviewer and owner so far is quite incorrect, IMO, at the very least.
G
mfunnell
Shaken, so blurred
On the contrary, I strongly suspect it's quite well designed, though probably with design parameters you would not have chosen or don't prefer.
But I think that's probably the point: it was designed to be used with the assistace of software correction, so it isn't intended to be used without that assistance. If that matters to you then I guess it matters. I think it would matter to me if it were an interchangable lens, especially if I wished to use it on both film and digital bodies or on camera bodies from different manufacturers. On a fixed-lens camera it doesn't bother me so much - especially a camera I'm unlikely to buy (for reasons quite other than the presence or absence of lens distortions and software corrections).
...Mike
H
Hunter S Thompson
Guest
Well designed? Who wants distortion? It's as if a Ferrari were designed to have one of the front tires out of alignment, but with an auxiliary 5th tire available to correct this.
I agree with you here, but tell me, where is the benefit to the customer for having this poorly designed lens attached to the camera that needs software help? How do you, me, or anyone benefit from this? What are you paying $4000 for in this case? You cannot look at this lens and admire its design because it is in fact loaded with distortion that needs software correction.
If this camera were $2000, then perhaps my objections would fade away. It's an entirely different scenario of Leica is producing a substandard lens that needs software correction in order to lower their costs and pass it on to the end user. This I can get behind.
What's next, artificially made bokeh? Imagine buying a 50mm f/1.4 where all the bokeh is made from software. We can laugh at it now, but such software already exists. It looks awful or at least fake today but it wont stay that way forever. Will you or anyone else be interested in a $4000 50mm lens (which is really a ho hum design) but has great software correction?
f16sunshine
Moderator
The benefit is most certainly having a lens with higher speed in a smaller package than it would be if designed for optical perfection without software correction.
Have you taken a look at the Zeiss Otus or Sigma Art lenses? Not exactly compact .
The Q having a fixed lens and therefore dedicated sensor, I prefer a smaller faster lens with SW correction to a perfect optical path in a much larger and unwieldy package.
It's the final output I want to see be perfect.
mfunnell
Shaken, so blurred
Sounds like an odd design for a motor car, though it would invite the question "but how does it drive?" If it drives well, perhaps better than more traditional designs at the price-point, the odd design might be justified. However this, from earlier in the thread, sounds closer to my guess regarding the motivation:
It's guesswork on my part but they seem to have made a rather wide aperture lens for its focal length and size. The use of software correction may be part of what enables that. I'm neither jumping for joy nor recoiling in horror. I am interested in what they've achieved in terms of the sample photographs I've seen, despite having no interest in buying the camera for myself.
...Mike
nongfuspring
Well-known
You're dramatically oversimplifying. With the exception of a few lens designs at certain focal lengths all lenses have distortion. Some have more, others have less, and generally the less distortion a lens has the more elements it must have and the larger and slower it tends to be. Lens design is always a balancing act between compromises. I'm not saying it's a good or a bad lens, but to call it poorly designed out of hand even when nobody here even seems to know exactly how much distortion it actually has seems more than a bit premature.
Personally I don't care - the only thing that I find important is the image at the end of the processing chain, and from what I've seen so far it seems pretty beaut. My only quibble is that from what I've seen so far you can't turn it off.
H
Hunter S Thompson
Guest
Over at FM I've seen folks there claim that the distortion is in excess of 2%. I can't be certain that's true but assuming it is, that's where my position comes from.
Of course all lenses have distortion. That's meaningless in the context of this discussion. I'm talking specifically of extreme levels of distortion. If a lens has to resort to software correction merely to present an acceptable picture, that's extreme distortion and in my view, a very poorly designed lens.
Mark T
Established
Sparrow
Veteran
... no, it doesn't matter there's a bit of distortion.
... there was a chap on the forum a few years back that took it upon themselves to point out the unacceptable distortion of every CV lens ever made, after a while it started to become aggravating ...
... when I eventually when I started questioning him it quickly became clear that what he was calling barrel or pincushion-distortion on the lens in question was not the aberration he was claiming but in fact simply the converging verticals of the classic Planar projections
I have taken care ever since to ascertain just what people are calling distortion and it seems converging verticals are seen as distortion far more often than perspective is, despite the fact that they are exactly the same thing
nongfuspring
Well-known
It's not meaningless, it is relative - you can't look at a single aspect of lens design out of context with other competing factors. Like Mark says 2% is hardly extreme, if 1 extra percent of distortion over the F2.8 ASPH is the compromise for a faster lens I'm sure many would be happy to make such a minor compromise.
But all this is really is meaningless unless it makes an impact on the final image. If the software correction is ruining the photos, then sure, it's bad; if not, then who cares. To dismiss a camera based on it's technology, not on it's usability or output is putting the cart before the horse.
