adamdewilde wrote:
I didn't even care until you all brought it up. Really I'm not to offended.. I'm actually thinking that my LR is broken or something though, because files are coming into LR looking like that RAW converter I'm using. The JPGS look totally different as of now.
You have to update Lightroom to a version that includes the Q profile for correction. Maybe install the one that came with the Q if you haven't yet.
Tariq Gibran wrote:
You have to update Lightroom to a version that includes the Q profile for correction. Maybe install the one that came with the Q if you haven't yet.
Yeah I'm guessing I have to move up to LR 6 then? I'm on LR5, wasn't going to bother moving up..
Anyway, just got back from an outdoor event with lots of strobes etc. I think that the AF is 1-series DSLR lvl, but I don't know if having a wide helps or doesn't help. I honestly think the wide puts it at a disadvantage. So they'd probably do good things with a 50 or 75 version of the Q.
And even though I find this a bit pointless. Here's a f/5.6 image that I've stripped the corrections out of.. I basically just moved all the numbers to 0. I might have not done the right thing on some of the sliders, but it seems like what you'd expect an uncorrected file to look like right?
JPG then RAW stripped of profile.
BTW, the JPG files are really very reasonable as far as look is concern. I just feel the 28mm makes the camera seem like a point and shoot.. A very expensive but capable, FAST, good AF point and shoot.
adamdewilde wrote:
Anyway, just got back from an outdoor event with lots of strobes etc. I think that the AF is 1-series DSLR lvl, but I don't know if having a wide helps or doesn't help. I honestly think the wide puts it at a disadvantage. So they'd probably do good things with a 50 or 75 version of the Q.
And even though I find this a bit pointless. Here's a f/5.6 image that I've stripped the corrections out of.. I basically just moved all the numbers to 0. I might have not done the right thing on some of the sliders, but it seems like what you'd expect an uncorrected file to look like right?
JPG then RAW stripped of profile.
BTW, the JPG files are really very reasonable as far as look is concern. I just feel the 28mm makes the camera seem like a point and shoot.. A very expensive but capable, FAST, good AF point and shoot.
Good example, particularly with the balloon in the lower right corner as we can see how the shape of a sphere actually gets distorted and elongated by the automatic correction.
I'd guess that this is technically a wider lens, say 25mm or so, which is corrected to 28mm. Optically, I believe this is done by increasing focal length toward the corners while retaining a theoretical 28mm lens' angle of view, which is to say that there's no such thing as a "pure" 28mm lens that doesn't have barrel distortion (and gobs of other distortions). Anyhow, I don't have any prejudice against digital corrections as opposed to optical corrections per se; it is the results that count.
Also, the technical documentation for the camera doesn't include MTF charts, which is disappointing but in line with all of Leica's fixed-lens cameras.
I see light but minimal lateral CA in the uncorrected file, but more interesting to me is the moire correction in the JPG, which appears to incur minimal to no resolution penalty although that's hard to gauge here.
I prefer for my wide angle lenses to have a healthy dose of barrel in order to retain accurate shapes. Sometimes you want the balloon to be round, sometimes you want the buildings to be straight, and it is nice to have that choice. I'm also reminded of why I like Capture One over Lightroom; in C1, you can back off the factory vignetting and distortion corrections, so that you start off with a corrected image but can recover some of the original characteristics to the desired degree. Then again, most modern lenses have wavy or pincushion distortion, and I can't think of a reason to prefer those...although portrait flattery hadn't occurred to me before, so maybe I'll just retain my prejudice against wavy distortion. :-)
I'd guess that this is technically a wider lens, say 25mm or so, which is corrected to 28mm. Optically, I believe this is done by increasing focal length toward the corners while retaining a theoretical 28mm lens' angle of view, which is to say that there's no such thing as a "pure" 28mm lens that doesn't have barrel distortion (and gobs of other distortions). Anyhow, I don't have any prejudice against digital corrections as opposed to optical corrections per se; it is the results that count.
Jon
It's achieved by increasing the focal length (perhaps to something like 25-26 as you suggest) of the lens and cropping to the 28mm focal length after distortion correction. The focal length is the focal length though. You can't have one for the center and one for the corners.
Even the new LEICA SUMMILUX-M 28mm f/1.4 ASPH has only around 1% distortion, an amount that would almost never need correction unless doing really critical work.
The amount of native distortion present in the Q's "28" lens is an order of magnitude greater than the Leica M 28's (my guess is between 7-12%). It's of an amount that really must be corrected, not only due to the distortion itself but due to those black corners from using the entire image circle of the lens (they must be cropped off unless one is going for a truly "special" look). Doing so does compromise the technical quality of the final image outside of the center - something very un-Leica like imo.
Also noteworthy in this context is that the Sony RX1's Sonnar, which is really around 32mm's before any correction, shows far less native optical distortion than this Q. In practice, it hardly ever needs correcting.
Well, I'm looking for some samples to check the corners....not that I was in the market, but curious. Here a few which might give an idea.
As usual finding a decent infinity test shot is not easy With the Q I think we have the added wildcard of AF, which is locked who knows where. So a real test would be in MF at the infinity stop, long view, with distant details in at least one corner. Please post links if you have them.
The expectation at DPR of corners as sharp as the center is laughable, especially at 28mm, and corners are only part of the story, mid-zone dips another issue. For example, with the ZM 35/2 you do have falloff in extreme corners at all apertures, but mid zones are superior to any Leica lens.
It's a choice.
Extreme corners are one thing, but edge performance in general is critical for me in landscapes, and the Q edges look pretty good on first look.
But forming an opinion based a few shots is dangerous.
uhoh7 wrote:
Well, I'm looking for some samples to check the corners....not that I was in the market, but curious. Here a few which might give an idea.
As usual finding a decent infinity test shot is not easy With the Q I think we have the added wildcard of AF, which is locked who knows where. So a real test would be in MF at the infinity stop, long view, with distant details in at least one corner. Please post links if you have them.
The expectation at DPR of corners as sharp as the center is laughable, especially at 28mm, and corners are only part of the story, mid-zone dips another issue. For example, with the ZM 35/2 you do have falloff in extreme corners at all apertures, but mid zones are superior to any Leica lens.
It's a choice.
Extreme corners are one thing, but edge performance in general is critical for me in landscapes, and the Q edges look pretty good on first look.
But forming an opinion based a few shots is dangerous. ...Show more →
It's a paid site, but Sean Reid's review does show landscape/ infinity examples with crops and compares them directly with the RX1. The difference in sharpness in the corners is day and night between the two. I'm surprised he did not get into the cause (maybe it's coming) as he has in previous reviews shown native vs digitally corrected distortion and has also written an article about the detrimental image effects of digital correction for distortion. If anything, Reid is also kind to Leica in general as a Leica user himself (I still find his opinions very objective).
The Q would not, imo, be the best choice if one needs decent corners for landscape work. The compromise in this case is entirely due to the design of the lens, which requires a lot of native distortion correction. This is most often found in cheaper lenses. If not for the existence of the RX1, one might claim it was necessary to achieve the size in a fixed lens FF camera.
Tariq Gibran wrote:
It's a paid site, but Sean Reid's review does show landscape/ infinity examples with crops and compares them directly with the RX1. The difference in sharpness in the corners is day and night between the two. I'm surprised he did not get into the cause (maybe it's coming) as he has in previous reviews shown native vs digitally corrected distortion and has also written an article about the detrimental image effects of digital correction for distortion. If anything, Reid is also kind to Leica in general as a Leica user himself (I still find his opinions very objective).
The Q would not, imo, be the best choice if one needs decent corners for landscape work. The compromise in this case is entirely due to the design of the lens, which requires a lot of native distortion correction. This is most often found in cheaper lenses. If not for the existence of the RX1, one might claim it was necessary to achieve the size in a fixed lens FF camera.
Ming implies the RX1 has the corners but the Q is well ahead mid-frame.
The question looming and which I fear will evade definitive answer for several months: the new backlit A7R2 sensor and 28 cron vs the Q, which would answer Louis question
Tariq Gibran wrote:
The focal length is the focal length though. You can't have one for the center and one for the corners.
My understanding is that reducing linear distortion in wide angle lenses is achieved by increasing magnification off axis, and that a lens' focal length is defined by the distance of the theoretical thin lens that would achieve the same magnification at a subject distance of infinity. Look at any brick wall test of any rectilinear, wide angle lens, and you can see that those at the corners are larger than those on axis, which is a physical necessity to keep lines relatively straight as the subject matter itself is located further from the lens. Notwithstanding nominal focal lengths, if the magnification is higher, then the focal length is longer, no?
I'm not trying to argue; I enjoy learning about this stuff and would appreciate any clarification of where I'm going wrong.
@uhoh: RX1 - three frame vertical stitch, dump it into APP (or now LR) and pick your focal length from ~24mm-~40mm, you are over sampling over the middle of the frame.
Don't think this Leica camera was meant for hikes in the wilderness, more for a wrist-strapping PJ or NY street shooter.
Well, from what I've seen so far, the lens may not be the best option for shooting landscapes or architecture, but it has a very nice rendering and its wide open performance is stellar. It would be perfect for many kinds of subjects.
But what is the most important part for me, is that the camera is really excellent in many respects, as highlighted by Adam. And inevitably, in a year time or less, we will see the same body that will take interchangeable lenses, AF, and all our M mount lenses and whatever mount you wish, via adapters, with a very thin sensor cover and no AA filter, uncompressed raw, and all the goodies of the Q. It will be all what you hoped the A7 would have been, and it wasn't.
JonPB wrote:
My understanding is that reducing linear distortion in wide angle lenses is achieved by increasing magnification off axis, and that a lens' focal length is defined by the distance of the theoretical thin lens that would achieve the same magnification at a subject distance of infinity. Look at any brick wall test of any rectilinear, wide angle lens, and you can see that those at the corners are larger than those on axis, which is a physical necessity to keep lines relatively straight as the subject matter itself is located further from the lens. Notwithstanding nominal focal lengths, if the magnification is higher, then the focal length is longer, no?
I'm not trying to argue; I enjoy learning about this stuff and would appreciate any clarification of where I'm going wrong.
The focal length of a lens is generally considered the distance (a single distance) between a point within the lens and the image that is formed when the lens is focused at infinity. I have not heard of a "single-focal-length" lens ever described as having multiple focal lengths. "Zoom" lenses of course allow one to change the lenses magnification but the only way to change (optically) the image a single-focal-length lens creates is by either moving the camera or moving yourself. That's the way the concept is taught anyway (text book photo 101).
edwardkaraa wrote:
Well, from what I've seen so far, the lens may not be the best option for shooting landscapes or architecture, but it has a very nice rendering and its wide open performance is stellar. It would be perfect for many kinds of subjects.
But what is the most important part for me, is that the camera is really excellent in many respects, as highlighted by Adam. And inevitably, in a year time or less, we will see the same body that will take interchangeable lenses, AF, and all our M mount lenses and whatever mount you wish, via adapters, with a very thin sensor cover and no AA filter, uncompressed raw, and all the goodies of the Q. It will be all what you hoped the A7 would have been, and it wasn't. ...Show more →
+1 Agree with all of that! I'm very anxious to see the body only version and I also wonder if one might be able to rip out the existing lens of the Q and....probably an expensive exercise that could result in 'crash and burn". Wouldn't it be cool though if one found an M mount lurking underneath.
Tariq Gibran wrote:
+1 Agree with all of that! I'm very anxious to see the body only version and I also wonder if one might be able to rip out the existing lens of the Q and....probably an expensive exercise that could result in 'crash and burn". Wouldn't it be cool though if one found an M mount lurking underneath.
Don't give this idea to Adam he might actually do it
edwardkaraa wrote:
Don't give this idea to Adam he might actually do it
Haahah.. When the RX1 first came out, my initial thought was to buy one and hack off the lens. Then I realised that none of my engineer friends would come to my aid and I had to scrap the idea.
I'm in the same boat now.. Willing to take a saw to the Q, but nobody to come to my rescue
I just got back from the movies BTW (that's why I wasn't chatting).. Going to get an early-ish night (it's 3:30am). Will upload some random snaps tomorrow, but man I'm really not use to the 28mm focal length. So everything looks "messy" to me.
I wouldn't worry about the corrections so much. I think it's a very usable camera for city streets, vacations, etc. I'm at least committing to keeping the camera for one wedding. And I'll bring it along on my other shoots, incase I need something wide (I never do).
Going to have to make a note to look at some 28mm street photographers, try to inspire myself tomorrow.
corposant wrote:
@uhoh: RX1 - three frame vertical stitch, dump it into APP (or now LR) and pick your focal length from ~24mm-~40mm, you are over sampling over the middle of the frame.
Don't think this Leica camera was meant for hikes in the wilderness, more for a wrist-strapping PJ or NY street shooter.
It's funny how Leica and everyone else totally ignores the raison d etre of the oringal ur-Leica, which was exactly that: a camera for hikes in the wilderness. The candid camera lineage can be claimed by zeiss, still nice and small but with the incredible 50/1.5, direct descendant of the original high speed Ermanox glass.
In fact, when asked about the slow speed of the Elmar, Leica voices replied something to the effect: "we don't want our climbers fooling with advanced settings, we just want them to bring home the images without fuss."
Today the Rx100 rules the backcountry, but we know the vast gap in image quality between that camera and a good FF. The fragile RX1 is also not meant for the backcountry, yet it's one of the few serious options, unless you want to distort the whole experience with alot of heavy gear.
The Nex 5 remains an option better than the Rx100 and more versatile than the RX1 but, a shadow of good FF performance. The M9 and three lenses is my current comprimise, and as fas I'm concerned the lenset is great, mature and ready for action: not too big if you are choosy. But the M9 body is too big, obviously and too heavy, as are all the A7 cameras, the newer the worse The Q is about 60% of the mass of M9 and a 28 cron.
So, in the real world of available small FF cameras, the Q will end up in some remote areas, or should, I think ti's reasonable to want to see how it can do. For me, the jury is still out.
edwardkaraa wrote:
And inevitably, in a year time or less, we will see the same body that will take interchangeable lenses, AF, and all our M mount lenses and whatever mount you wish, via adapters, with a very thin sensor cover and no AA filter, uncompressed raw, and all the goodies of the Q.
How's the AF gonna work with all of the M-mount lenses? Or will it be like Nikon - a mount that works with all of the old glass but also develop a whole new line of AF lenses?
Ray S. wrote:
How's the AF gonna work with all of the M-mount lenses? Or will it be like Nikon - a mount that works with all of the old glass but also develop a whole new line of AF lenses?
-Ray
I suspect we may see a whole new line of AF lenses form Leica. How compatible the M mount lenses will be is the big question. Clearly, AF is not gonna work with the manual focus M-mount lenses. My guess is we see this at Photokina next year.
With the Q I think we have the added wildcard of AF, which is locked who knows where. So a real test would be in MF at the infinity stop, long view, with distant details in at least one corner. Please post links if you have them.
. Wouldn't it be cool though if one found an M mount lurking underneath.