It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
Well then, I can assure you that this is not a problem with that great lens and that world-class amazing little FF 50 MP camera. The image fidelity on the Q2 is absolutely amazing. It is a nice 50MP FF sensor combined with one of the very best prime lenses on Planet Earth. The Q2 is incredible. Why do you think so many discerning GFX shooters have one? It's not for the Red Dot. Well.... Maybe a little.I'm unfamiliar with the lens on the Q2, but the issue would be more likely distortion than field curvature. When the digital correction is applied to distortion it wreaks havoc on the image quality, most notably in the corners.Is field curvature the reason for the softness in the bottom corners. For a lens that’s supposed to be considered special I’m shocked at how bad it is stopped down to f7.1. I would expect that wide open but not at that aperture.I am not a programmer or computer scientist and I can't write code. But I am a PC hobbyist and have been since the dawn of the PC age (build my own rigs). I have never used Apple stuff but think they are amazing. Anyway, I am not technical about windows discussions so I have to be careful, but I have followed the heated discussions about windows being way behind Mac on scaling from the dawn of the 4K age. But they fixed that and Windows is as good as the Apple OS now when it comes to scaling. But a lot of it is still software and website specific. 8K is going to be interesting. Can't wait.Your files posted here always look fine to me. I don’t know about PCs, but Mac browsers display images at 200% when an image is set to 100% in the browser or opened “bare” as a JPEG into a new window.
Adobe jumped on this 4k scaling thing early with Microsoft because they had pro content developers and photographers using LR and PS with PCs and servers and there were these new 4K monitors. Adobe nailed it early and pretty well. So in Windows when you ask for 1:1 with LR or PS, that is what you get. You don't get 1:2.5. Maybe you do when you open a jpeg out of DPR at what they call 100% view. If so, DPR needs to fix that. But anyway, Windows will suggest that you set to 150% for almost any 32 inch 4K IPS pro monitor in order to change the size of text, play well with most apps and to assist with the proper scaling overall. Anyway, I'm not technical on that aspect so thanks for the info. I could always be wrong about some of that explanation.
Yes I did and I didn't like what I saw, but it was better than viewing it straight up on DPR. But, 100% in PS is really true pixel res. Same for LR. They talk about that all the time on the LR and Adobe boards. Adobe does that well. By the way, I'm not bashing DPR for this but this is the leading photo equipment website and this is the MF Board where we talk the nuances, compare shots and post some of the highest res images in the world. They could do better.When looking at high res files in your browser, you have to zoom out your browser (Command and “-“ keys. If you look at samples on Lloyd Chambers website, it’s the same way. On a Mac at 100% zoom (using the drop-down menu below his images) they will look blurry/pixelated. I have to view them at 50%.
Have you tried downloading the JPEG generated here and opened it on Photoshop then viewed at 100%? It should look ok minus some detail for compression. But also notice how 100% in PS is less magnified than they show up here.
I export all jpegs to DPR as full-size jpegs at 85 to 88% quality and have for a long time. Jim and I have had this discussion many times and he guided me through that minefield three years ago. I was asking whether to downsize the jpeg export from LR or just make it lower quality in order to get the jpeg at low enough MP to post on DPR. 45 MB files will usually post and 48 will not. I try to stay around 30 MB by lowering the quality setting. Some have said you can drop to 80% in LR exports and it is indiscernible and you get much smaller jpegs. That algorithm is something isn't it? The limit on DPR is around 48. You try 50 and your post locks and you have to crash out of it. They should have a warning....Another thing to note is always save your JPEGs with at least some compression before uploading here and to many other sites. I use 90% quality (11 in PS).
Yes, especially DPR. I think they need to lighten up and give up some of that massive server space or borrow some from their owner and master - Godzilla. That is not an insult. Godzilla let King Kong live in the last movie. He had him down and out....Sometimes sites will bypass compression if the image is already compressed some. If you try to upload a JPEG at max quality, most sites will crush your image with their heavy default compression.
I know all of the equivalencies of course and am aware of the challenges of posting side by side comparison shots with different systems, sensor sizes and lenses. The variables are endless. You know the list....Regarding Q2 vs. GFX 100 DOF:
1. The Q2 lens is wider than 28mm (so adjust your thinking when trying to compare to the GF 30).
I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied? I would like to go peep one uncorrected. So FC with that awesome little mm prime is really "massive: and distortion is massive? I thought that 28 on the Q2 was one of the top 3 lenses in the World. That's what everyone tells me. That lens is that bad? I need to go back and read the Leica Q2 reviews and see what they great reviewers said about that. I honestly don't remember and I read them all.2. Q2 has massive field curvature and distortion (corrected in software) that cheats DOF for certain landscape scenarios.
Would love to see it. Tell me how to peep a Leica Q2 DNG uncorrected.If you look at an uncorrected Q2 image, it almost looks like a fisheye lens.
Wow! I believe you but I did not know that. How can that never be mentioned on the Leica Board? That said, that board is probably the absolute least technical of all the boards, which is pretty amazing because you would think it would be otherwise.3. The mandatory Q2 crop applied to all Q2 files means it’s technically not a full frame image.
Amazing. Smaller than FF. The Leica Q2 is smaller than FF? Would someone elsew like to comment on that? How much smaller?You’re comparing smaller than full frame DOF with 44x33.
Yes, that I am very aware of and has been talked about a lot since we GFX shooters are shooting at 4:3. I love 4:3. That has been noted by Erik and Jim before as being one of the nuances that is "tricking" me into believing the difference between GFX and FF DOF is way more than one stop. But I don't harp on that number anymore and don't care so much. I just say I notice a big difference. Sometimes bigger than big. Like on that shot.4. A horizontal landscape with the Q2’s 3:2 image ratio does not have to cover as much distance with detail as 4:3, especially if you keep the same amount of sky for the shots with both formats.
That was a fine post Sir. You should start a thread with this information on the Leica Board. The Q2 shooters there would appreciate it. And a Hell of a lot of them are GFX shooters too because the Q2 and GFX go together like Michael Jackson's hand in a sequined glove in 1985.
The Q2 is my all-time absolute most favorite camera. Now I could never just use it alone for a long period of time. Maybe for a day, or a hike. The beautiful Q2.... When I have it around my neck, which is a lot, it makes me happy. When I hold it in my hand, it makes me happy. When I look at it, it makes me happy. Sometimes I pet it and moan softly while on the airplane after gently taking it out of the shoulder bag piled on top of the GFX 100 body sand EVF.... I say to it, Darling - I can't wait to get you alone on Sicily...Ohhhh My Love ... the things we are going to do together....Those little emojis are me smooching with my sexy little Q2. Her name is Ansell, and she has some bumps and bruises from banging up against my GFX 100 for two years.
My problem is my GFX 100 body and my Q2 body are all banged up because they are constantly bashing into one another around my neck, in my bag, and in the back seat and floorboards of my truck. They sound like two bowling balls bumping up against each other as I drive or stroll down the road or walk through parks and alleyways.. Luckily, most of the scratches and dents on my GFX 100 are on the RRS L Plate and on the Q2 the scratches and dents are mostly on that awesome twist-on metal hood that is permanently on the 28mm lens that has so much horrible distortion and FC.
On another note of interest to the Board, I just sent my Q2 to Germany because they are going to help me with a little problem I have. I also just mailed my GFX 100 to Steve in NJ because I have a couple of little issues that they are going to help me out with. These are all little bumps in the road, but I bet it takes me two months to get that Q2 back from Germany. The 100? 5 days I bet. Those guys are great.
While doing a couple of minor repairs (including replacing my totally no-friction tilt adapter), they are also going to check out my GFX 100 as it has been over some rocky road lately. I even asked them to check for a tilted sensor, even though I know mine is not tilted because I pixel-peep every shot and I know what right looks like.
But here's the thing.... I also know what wrong looks like.
That was a good post Man. Thanks.
This has been explained to you several times. You're not getting it. I'll try again. DPR prohibits repetitive links to commercial sites, identifying that as spam. The two sites you mention are commercial sites. They sell ads and monetize eyeballs. A link here and there to either of those seems to be acceptable. Many such links would be spam.But ask yourself this. What if Fred or Ken The Angry posted here 15 times a day with strong opinions and linked to their platforms hundreds or thousands of times? There would be a riot and people would go berserk. The Board would blow up and chaos would ensue.
It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
I know. So do yopu think the 28 has a lot of FC? I don't think it does. Do you? I could be wrong.It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
So are you guys telling me that or the guy who said it?Jim is correct, of course. The correction is to distortion, not field curvature.It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
This reddit thread has some images showing the distortion and cropping.Is field curvature the reason for soft corners? No. Field curvature doesn’t make lenses softer—it just changes the plane of focus. Bottom edges of images tend to be closer to the camera, so field curvature (generally closer as you move away from the center) actually helps the lower edge and corners. Field curvature would hurt sharpness at the top of most images, but that is often “sky” that is devoid of detail so one doesn’t notice.
Lack of sharpness in the corners is a combination of ordinary optical defects and, in particular, distortion correction.
When you fix distortion in software, you obviously need to make bigger corrections in the corners than anywhere else. Effectively, you might smear the data from one pixel across the area that should be covered by two. This can degrade sharpness somewhat. To see how much you would have to compare an uncorrected and a corrected image as it will vary from one lens design to the next. The more the distortion correction, the greater the loss in detail. However, it’s usually pretty minor with most lenses, and it only is noticeable in the extremities. Lightroom and Adobe won’t let you disable the distortion correction in the Q. Some other packages will.
The way Leica chose to design the Q was as a combined optical/software system. They were able to achieve a fast lens with insane resolution overall in a compact size with OIS by NOT fully correcting for distortion optically. Instead, software takes care of that particular issue. Saves a lot of bulk while maintaining excellent detail throughout most of the image.
Is the Q less than full frame? Of course not. It’s a full frame camera. Heck, most cameras have regions of the chip that are not used—just for calibration as they are optically black. That’s why pixel counts in the final image sometimes don’t match spec sheets—at least it’s one reason why. It’s still a full frame camera. Does it throw away some data in the corners? Absolutely. Distortion correction necessarily does that. All profiles with distortion correction throw away some corner data. The Q has more distortion than most so loses more corner data than most. You lose some field of view in the corners as well as some resolution due to the “stretching” of data. It’s a good compromise, though obviously not ideal for all pictures.
DPR specifically allows linking to your portfolio, but cautions you that overdoing it could be considered spam. Flickr is a commercial site, so that moves it towards a gray area. I think that the number of links would be dispositive.So I guess I'm doing the same thing, except its just in the sig block and I don't try to push traffic to Flickr. I post my images on the MF Board that I want you guys to see. I don't link to Flick and say, "see this image." Should I? Because I can do that. That's easy. But I didn't think DPR liked that.
I used to have a blog, but got tired of the content management tool I was using.That's not true at all. I am saying that there's a reason I post comparison images on my blog, and that's because I can control the compression. It's also because I can organize posts in the way that I see fit.I know that. Your blog is linked thousands of times a year on DPR and multiple times daily on this Board. I don't avoid your blog Jim, and I like it and have gone to it a lot. But this is DPR. I'm not sure that DPR should constantly link to your blog so much and am surprised you and others who do it constantly get away with it here on this DPR MF Board. But I have no beef with that. Carry on. I have never once complained about it and it's not a big deal but you brought it up openly here and are actually complaining that I post here instead of on your blog.
I never suggested that you comment on my blog. I have suggested several times that you look at comparison images on my blog. For the most part, you have refused to do so. The last time I remember was when I posted DOF comparison images.
Who said it? Where? When?There are good blogs all over the place but this is DPR. Some might think your blog was an actual competitor to DPR. I don't, but its been said.
From the DPR Guidelines: "In the spirit of encouraging discussion, it’s fine to post links to other websites. However, posting links to explicitly commercial sites may be considered spam."
My site is not a commercial site. There are no ads there.
This is not the first time you have brought this up. I'm not sure why it bothers you.
Jim
Who said it? Where? When?Who said it? Where? When?There are good blogs all over the place but this is DPR. Some might think your blog was an actual competitor to DPR. I don't, but its been said.
I didn't see evidence of significant FC in the image you posted, but that is my only point of reference. The link to reddit above did demonstrate some distortion that is corrected in camera, but not an egregious level, especially for the world of fixed lens cameras.I know. So do yopu think the 28 has a lot of FC? I don't think it does. Do you? I could be wrong.It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
Yeah, from the links further down in the thread it is clear that the level of distortion isn't bad, and I absolutely believe on the image quality you can get out of that little camera. I shoot a Fuji X100F and its a great little camera, but I wouldn't mind if it was a Q2 instead.Well then, I can assure you that this is not a problem with that great lens and that world-class amazing little FF 50 MP camera. The image fidelity on the Q2 is absolutely amazing. It is a nice 50MP FF sensor combined with one of the very best prime lenses on Planet Earth. The Q2 is incredible. Why do you think so many discerning GFX shooters have one? It's not for the Red Dot. Well.... Maybe a little.I'm unfamiliar with the lens on the Q2, but the issue would be more likely distortion than field curvature. When the digital correction is applied to distortion it wreaks havoc on the image quality, most notably in the corners.Is field curvature the reason for the softness in the bottom corners. For a lens that’s supposed to be considered special I’m shocked at how bad it is stopped down to f7.1. I would expect that wide open but not at that aperture.![]()
Then why did you ask if you could turn off correction in Lr?I know.It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
Never tested it, so I have no opinion.So do yopu think the 28 has a lot of FC? I don't think it does. Do you?
I could be wrong.
I'm commenting on what you said. To my knowledge, no one ever said in this thread that there is software correction for FC.So are you guys telling me that or the guy who said it?Jim is correct, of course. The correction is to distortion, not field curvature.It is impossible to correct for field curvature in post.I have not noticed FC on my Q2 (maybe because it is so well corrected when I pop open the DNG file in LR). Can I turn that off in LR and see it without the corrections applied?
I'd go a step further and suggest that Jim isn't overly concerned with the number of views on his site either.Well, I don't know Jim K personally although I feel like I do, but I know enough to bet good money with anyone that he is not seeking to make a few bucks on his website / blog.If they were linking to their work in a way that was helpful to the discussion at hand and brought greater clarity and understanding to the community it wouldn't be a problem, it would be welcomed. There is a very clear difference between blog spam and contributing to the discussion. I rarely post links to my own website here because 1) it often isn't related to medium format photography and 2) I actually do monetize my website and want more clicks and views. No offense to Jim, but it is pretty obvious that he doesn't run his site in a way to optimize clicks. 'The Last Word' isn't particularly easy to navigate, isn't monetized, and unlike myself, the entry titles aren't exactly sensationalized. If Ken the Angry has work that is relevant to the discussion at hand, he, or anyone else is welcome to link to it. As a forewarning though it is likely to be met with some derision as his scientific instincts are, to put it mildly, poor.I have no problem with people linking a blog to a post and who would care if I did? I'm not a Mod and have no right to tell anyone what they can and can not post.Most of the content on DPR is in the form of forum posts, not reviews. DPR paid no one for the free content they supply that drives thousands and thousands of visits to DPR. I don't think DPR are in any position to claim that linking a blog in a forum post is in some sesne piggy backing on DPR's success...
But ask yourself this. What if Fred or Ken The Angry posted here 15 times a day with strong opinions and linked to their platforms hundreds or thousands of times? There would be a riot and people would go berserk. The Board would blow up and chaos would ensue.
There is no problem with linking to your work in flickr when appropriate.So I guess I'm doing the same thing, except its just in the sig block and I don't try to push traffic to Flickr. I post my images on the MF Board that I want you guys to see. I don't link to Flick and say, "see this image." Should I? Because I can do that. That's easy. But I didn't think DPR liked that.
Yes, the Q2 sensor is a full frame sensor, but the lens is projecting about a 26mm image onto that sensor and the DNG is flagged for C1/LR to crop the image down to just below 28mm. So it's true the Q2 is full frame, but it's also true the final images are not. This is much more aggressive software correction and cropping than we are used to seeing. The Q2 images are cropped on all sides, but most heavily on the long sides due to the 3:2 image ratio.Is field curvature the reason for soft corners? No. Field curvature doesn’t make lenses softer—it just changes the plane of focus. Bottom edges of images tend to be closer to the camera, so field curvature (generally closer as you move away from the center) actually helps the lower edge and corners. Field curvature would hurt sharpness at the top of most images, but that is often “sky” that is devoid of detail so one doesn’t notice.
Lack of sharpness in the corners is a combination of ordinary optical defects and, in particular, distortion correction.
When you fix distortion in software, you obviously need to make bigger corrections in the corners than anywhere else. Effectively, you might smear the data from one pixel across the area that should be covered by two. This can degrade sharpness somewhat. To see how much you would have to compare an uncorrected and a corrected image as it will vary from one lens design to the next. The more the distortion correction, the greater the loss in detail. However, it’s usually pretty minor with most lenses, and it only is noticeable in the extremities. Lightroom and Adobe won’t let you disable the distortion correction in the Q. Some other packages will.
The way Leica chose to design the Q was as a combined optical/software system. They were able to achieve a fast lens with insane resolution overall in a compact size with OIS by NOT fully correcting for distortion optically. Instead, software takes care of that particular issue. Saves a lot of bulk while maintaining excellent detail throughout most of the image.
Is the Q less than full frame? Of course not. It’s a full frame camera. Heck, most cameras have regions of the chip that are not used—just for calibration as they are optically black. That’s why pixel counts in the final image sometimes don’t match spec sheets—at least it’s one reason why. It’s still a full frame camera. Does it throw away some data in the corners? Absolutely. Distortion correction necessarily does that. All profiles with distortion correction throw away some corner data. The Q has more distortion than most so loses more corner data than most. You lose some field of view in the corners as well as some resolution due to the “stretching” of data. It’s a good compromise, though obviously not ideal for all pictures.
I have no clout with DPR. And you are mischaracterizing the effects of their compression. If you don't like their compression, link to an ad-free site with your images on it.I have a better idea. Use your clout to get DPR to allow us to post some MF shots that don't look like my wife's iPhone shots blown up on my studio monitor.Post the raw on Dropbox. There’s single pixel detail in the center of that crop.Nope. That garbage you posted from the DPR crop is nothing remotely like I see on my monitor with the raw or my jpeg at 1:1.Greg, you haven't responded to this post , though you've responded to others since. I think this whole thread is overblown. I downloaded the JPEG of your original post from DPR. I looked at it with extreme pixel-peeping settings, and posted the results in the link above.
There is pixel-level detail there, but not in the foreground.
Sure there's compression going on, but it's not enough to invalidate judging the sharpness of your image, and indeed it's enough to show that there is about the expected amount of DOF-caused degradation of image sharpness.
And here's a tight crop of what I see in my browser window at 100% on a 4K monitor;
This is a PNG file so there's less compression.
Let's flatten that mountain and call it a molehill.
Not even close. And after shooting probably 50,000 GFX shots, I can tell you for damn sure that no GFX shot on a scene like that will have anywhere near that DOF. Not even close. Like I said, it slaps you in the face it is so obvious.
Go back and look at your old images; are they similarly afflicted?
blog.kasson.com
Kodalith.Nope. That garbage you posted from the DPR crop is nothing remotely like I see on my monitor with the raw or my jpeg at 1:1.Greg, you haven't responded to this post , though you've responded to others since. I think this whole thread is overblown. I downloaded the JPEG of your original post from DPR. I looked at it with extreme pixel-peeping settings, and posted the results in the link above.
There is pixel-level detail there, but not in the foreground.
Sure there's compression going on, but it's not enough to invalidate judging the sharpness of your image, and indeed it's enough to show that there is about the expected amount of DOF-caused degradation of image sharpness.
And here's a tight crop of what I see in my browser window at 100% on a 4K monitor;
This is a PNG file so there's less compression.
Let's flatten that mountain and call it a molehill.
Not even close. And after shooting probably 50,000 GFX shots, I can tell you for damn sure that no GFX shot on a scene like that will have anywhere near that DOF. Not even close. Like I said, it slaps you in the face it is so obvious.
blog.kasson.com