AlanG Posted June 27, 2011 Share #561 Posted June 27, 2011 Advertisement (gone after registration) To borrow from the vernacular of usenet... Alt.thread.die.die.die As I consider this to be unacceptable, please excuse me for asking if these problems and stair stepped edges on some angled lines is typical of what one can expect when shooting buildings with an M9. Or is this just a bad example that I stumbled onto? Do you know? Can some raw converters prevent this because I have seen some that do better than others on this type of angled line with various other cameras. Is there some other place I should be asking this? Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/153862-open-letter-to-leica-%E2%80%94-10-ways-to-improve-the-m9-rangefinder/?do=findComment&comment=1716820'>More sharing options...
Advertisement Posted June 27, 2011 Posted June 27, 2011 Hi AlanG, Take a look here Open Letter to Leica — 10 Ways To Improve the M9 Rangefinder. I'm sure you'll find what you were looking for!
mjh Posted June 28, 2011 Share #562 Posted June 28, 2011 I don't know if calling this moire is correct as it is a simple edge and not an interference pattern. It really is a zigzag edge that is getting some color artifacts While it isn’t moiré in the strict sense but rather a demosaicing error, it is basically the same phenomenon as colour moiré. See my forthcoming article on moiré in the next issue of LFI. Link to post Share on other sites More sharing options...
mwilliamsphotography Posted June 28, 2011 Share #563 Posted June 28, 2011 Does anyone use a M9 to professionally shoot architecture work? Just curious. "I've had my M9 since February." Interesting. Is this your first rangefinder? How much have you used the M9 and in what circumstances? I'm a pretty good decisive moment photographer. I have an iPhone. It's cool. It takes fun photos. I could probably shoot some decisive moment shots with it. If I did that for one of my wedding clients and they looked like the sample links provided, I'd have plenty of time to play with my iPhone, because I'd be out of work : -) Speed? iPhones blink, rangefinders do not. iPhone pics can look arty because they can't take a real photo in many situations ... so we call it arty. Anyone can make a real photo look arty, but you can't do the opposite. Laughable arguments. Keep them coming, it cheers me up a lot. -Marc Link to post Share on other sites More sharing options...
AlanG Posted June 28, 2011 Share #564 Posted June 28, 2011 Does anyone use a M9 to professionally shoot architecture work? Just curious. Like others, I tend to shoot a lot of buildings when I travel even when not shooting professionally. But I think if the edges get stair stepped in buildings this will have similar results on edges of all sorts of subjects. Thus it would be a good idea to minimize this. Three years ago I tested C1, DXO and Canon's DPP converter and found that DXO did the best job on angled edges at that time. I am just questioning if anyone has found if using C-1, Lightroom, PS CS5, Aperture, or possibly some other converter minimizes this on the M9. Link to post Share on other sites More sharing options...
Jamie Roberts Posted June 28, 2011 Share #565 Posted June 28, 2011 Here is a link to the entire photo. You can see the lines have disappeared even in the low res version. Al four of his Getty Museum photos have line drop outs. (Even the areas in shadow.) I can't think of any way to explain this. The b/w ones clearly show some aliasing on the edges. I don't have the raw files and I know some of this might be in the raw conversion so I'll withhold conclusions. But I shoot a lot of buildings and have not run across either the "moire" on sharp edges or the drop outs. (I don't know if calling this moire is correct as it is a simple edge and not an interference pattern. It really is a zigzag edge that is getting some color artifacts){snipped} Alan, I really do think you're noticing an anomaly of processing OR a physical difference in the light hitting the "panels" OR an actual physical anomaly in the building. I've never seen the M9 just "drop" detail out of a shot under any circumstances whatsoever. I'd be very surprised if anyone else had seen that either... ETA: I just looked at a dozen or so non-M9 shots of the same musem available from Google images. The lines "drop out" in the same places: this is an actual panel artefact (if we're talking about the same thing). It's not the M9 at all. I also followed the link you provided and don't really see much aliasing on BW edges though what is there could easily be a JPEG or even a resizing artefact and have nothing to do with the camera. I mean, it's hard to know what we're even looking at here, isn't it? I also think that perhaps the panel edges might be subject to moire because the lens / sensor is actually focusing or discriminating actual physical variance between the foreground and background. This often happens with mesh fences in buildings or screens as well (and with Canons too ) In any case, unless we have a control camera shooting the same thing at the same time, and the RAW files that shot this, I think it's safer to say that the M9 records what is there, yet certainly is subject to digital artefacts like other high resolution digital cameras. Link to post Share on other sites More sharing options...
Jamie Roberts Posted June 28, 2011 Share #566 Posted June 28, 2011 {snipped}I am just questioning if anyone has found if using C-1, Lightroom, PS CS5, Aperture, or possibly some other converter minimizes this on the M9. C1 does an excellent job minimizing colour artefacts with the M9. As I said, while I'm not a architecture shooter, I just don't see much moire or CA using C1 and M9 raw files. I'm sure the RAW converter has a lot to do with the final output. Link to post Share on other sites More sharing options...
AlanJW Posted June 28, 2011 Share #567 Posted June 28, 2011 Advertisement (gone after registration) I also do a lot of buildings when I am traveling and I have not seen this (moire) problem. I use ACR and CS5 because I have a workflow that I understand and don't want to expend the effort to try something else. I am not sure therefore whether it is the raw developer or something else that caused it in the image Alan discusses. For work like this I would be at ISO 160 by the way, just to plug in that variable. Link to post Share on other sites More sharing options...
Jamie Roberts Posted June 28, 2011 Share #568 Posted June 28, 2011 As I consider this to be unacceptable, please excuse me for asking if these problems and stair stepped edges on some angled lines is typical of what one can expect when shooting buildings with an M9. Or is this just a bad example that I stumbled onto? Do you know? Can some raw converters prevent this because I have seen some that do better than others on this type of angled line with various other cameras. Is there some other place I should be asking this? Alan, I have the original DNG from Jeff's thread and I can confirm the following: 1) C1 V6 simply does not show the moire / CA very much--it is still there, but it's an order of magnitude less than the posted JPEGs IMO. I did nothing to the file but open it up in C1. I doubt anything would show in print. Just to be sure, with the purple fringing checked on the lens tab, even the very small amounts overall are essentially gone. 2) there are NO line drop outs. Those are actually variances in the actual panel details, and the M9 does a fabulous job of showing them actually. You can see the 3d variation in the curvature of the museum if you look at the whole shot 3) There is no stair-stepping of detail or line jaggies throughout the shot. 4) The detail, here at ISO 160 at 100%, is quite astonishing actually, for the lens at f8, but I really wouldn't expect anything less from an aspherical 35 Lux at f8. I'd like to see it a 5.6 5) BTW--One thing that will really exacerbate the issue is if you look at these shots through a non-color-managed browser (like Chrome or IE9) with a wide-gamut monitor set to aRGB. The profile mis-match will increase the saturation of colour (and therefore of any colour artefact). Only when I'm looking at the C1-generated files under those non-colour-managed conditions do I see anything like a lot of CA, but as I said, that's due to much oversaturation. So download them and look at them in a color-managed application like Photoshop (they should have the sRGB profile embedded; I'm not sure what posting them here really does in those terms). I can also post a 100% JPEG from C1 if you'd like to see it; the full-res, full bit-depth TIFF is better, but it's over 100MB In the meantime, here are some 100% crops from the JPEG (and re-saved as JPEG...not as good as a TIFF but they make the point very well, I think): Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/153862-open-letter-to-leica-%E2%80%94-10-ways-to-improve-the-m9-rangefinder/?do=findComment&comment=1716975'>More sharing options...
bill Posted June 28, 2011 Share #569 Posted June 28, 2011 So this is a non-issue then. Or at worst a limitation of today's digital technology that is evidenced in some degree under certain circumstances by sensor-based imaging systems which is easily minimised or eliminated in post-processing software. Interesting to note how practical, hands-on experience of the Leica equipment in the dock debunks disengaged theory and observation "from a distance". My thanks to those who lent their time and expertise to demonstrating the extent to which this is a storm in a teacup. Regards, Bill Link to post Share on other sites More sharing options...
mwilliamsphotography Posted June 28, 2011 Share #570 Posted June 28, 2011 Nice detective work Jamie. There is a lot of junk that shows up in files even from cameras a lot better than the M9 ... which are usually non-issues due to screen resolution conflicts, bad exposure on parts of the image, conversion issues, or poor post work with the RAW files. If I intend blowing up a file to the kind of pixel peeping ratios shown here, it sure isn't as an 8 bit jpeg and sRGB color space ... and I usually use something like C1 to do it if there are any initial visible issues because it makes them a real world non-issue or eliminates them entirely. I've seen the kind of issues Alan has pointed out when shooting buildings in relatively harsh light (and not just with the M9). None of it shows up in the print after just standard post adjustments are applied. Time to move onto another M9 fault to debunk : -) -Marc Link to post Share on other sites More sharing options...
Jamie Roberts Posted June 28, 2011 Share #571 Posted June 28, 2011 Bill & Marc-- The "moire" is even less of an issue with a TIFF from C1 colour balanced for a neutral white point (the white metal on the building, not the panels) and a neutral black point in the shot. Then any residual gradient noise (from the underexposed and under-"developed" red channel) goes away as well. Yes, the building and rocks then look warm, but it was taken (evidently) at 4PM in the afternoon under a cloudless sky in LA I'd expect it to be warmer than the "as shot" as posted, where the sky is pure cyan (you're not going to be able to print that anyway!). Oh--and in the original thread, it looks like Zlatko found out the OP was actually turning up luminance sharpening when processing in LR (something I know nothing about)--but that would certainly exacerbate the artefact problem, to say the least! Link to post Share on other sites More sharing options...
AlanG Posted June 28, 2011 Share #572 Posted June 28, 2011 Thanks Jamie. That looks much better (not perfect - still some stair stepping and color in some lines) but I'm only viewing on my laptop right now and will look later. I'll download the raw and check it out in C1. I haven't used Lightroom, but I figure when you get down to this level, some differences in raw converters can show up. I know DXO does gives me less stair stepping in lines than this, but I'll compare this photo and one of my own in C1 to standardize things a bit. Note: when I say stair stepping in lines it is not in all angular lines but just ones that go at certain angles. The second photo has wavy lines and the lower part of the handrail is also getting color aliased. I know I am being much more critical then many people here but that is what I look for at this point because cameras, lenses, and raw processing have gotten so good that these tiny differences may be all that separates them. Link to post Share on other sites More sharing options...
bill Posted June 28, 2011 Share #573 Posted June 28, 2011 I know I am being much more critical then many people here but that is what I look for at this point because cameras, lenses, and raw processing have gotten so good that these tiny differences may be all that separates them. So, from this do we conclude that the reason you do not have a Leica of any description is because they don't make anything that meets your critical standards? Do we further conclude that every one of us has by definition lower standards than you? Regards, Bill Link to post Share on other sites More sharing options...
AlanG Posted June 28, 2011 Share #574 Posted June 28, 2011 I downloaded the raw image and would have to say that I'd like to take back any concerns I had about the image quality. I think the issues have to do with converting in Lightroom, or at least how this person converted in Lightroom. I don't have Lightroom and the closest Adobe converter I have is camera raw 5.7 (Which I don't use much.) Anyhow with my limited ability in camera raw 5.7, I couldn't get the image looking right. In C1 it was possible to get it much better and I'd say it's as good as I think one could get from an 18 megapixel 35mm image. (In my experience.) I used the C1 default sharpening for M9 and did not see any problems vs. having this off. I did find that the color moire had to be set at around 60 for a clean image. I tried various patterns and didn't see much change as these probably don't change edges much. If I turned the moire power up to 80 or higher some bad things happened. In general, I did not see too many other things being impacted, at least on this image, by using a moire power of 60. I can't imagine not using this as a default unless one could find it causes color changes in some images. Maybe something like this moiré correction could be in the basic firmware that makes the DNG file so that other converters, that don't have moiré correction, and in camera jpegs could look this good. I do think the line drop-outs may have to do with the actual building panels as they look to be installed unevenly. This seems strange in this high level of construction but perhaps a photo doesn't lie Anyway, I apologize if I pissed off anyone and do agree the image quality looks superb when you get the most out of the conversions. Here are two crops, one with the moiré filter at 60 and the other at 80 so you can see what happens if you overdo it. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/153862-open-letter-to-leica-%E2%80%94-10-ways-to-improve-the-m9-rangefinder/?do=findComment&comment=1717629'>More sharing options...
AlanG Posted June 28, 2011 Share #575 Posted June 28, 2011 So, from this do we conclude that the reason you do not have a Leica of any description is because they don't make anything that meets your critical standards? Do we further conclude that every one of us has by definition lower standards than you? Regards, Bill I think I was clear and I don't know why you'd form this conclusion. If I am going to buy any camera and lens system, I want to look extremely closely at various aspects of an image to see if I can justify the costs over other less expensive choices. At minimum if I agree to pay a premium because I prefer how a certain camera works or want features provided by that camera, I'd also like to make sure that the image quality is at least equal to what I am already utilizing. But in general, I think all of the higher res full frame cameras (and perhaps many APS cameras) and good lenses are capable of giving very good highly detailed results that are usable in all but maybe some extremely critical situations that might require MF gear. In terms of actual image quality, not camera features, differences may come down to really narrow specific items such as which particular lens performs better wide open, or which lens + converter has the least distortion, etc. And of course your camera and subjects may have to be pretty locked down to see small differences at that level. Link to post Share on other sites More sharing options...
Rick Posted June 28, 2011 Share #576 Posted June 28, 2011 QED, The End.................................... Link to post Share on other sites More sharing options...
Jamie Roberts Posted June 28, 2011 Share #577 Posted June 28, 2011 {snipped}In C1 it was possible to get it much better and I'd say it's as good as I think one could get from an 18 megapixel 35mm image. (In my experience.) I used the C1 default sharpening for M9 and did not see any problems vs. having this off. I did find that the color moire had to be set at around 60 for a clean image.{snipped} Thanks Alan--it's worth pointing out that I didn't even use the Moire tool (on noise reduction tab). You're right--it cleans things up totally. Link to post Share on other sites More sharing options...
AlanG Posted June 28, 2011 Share #578 Posted June 28, 2011 Just so you can see a comparison, here is a 100% crop from a 5DII using a 24 TSE at 1/160th at f10 hand held. These are fairly comparable sized objects and edges. Both were processed at straight C1 defaults, sharpness 130 for the Leica and 180 for the Canon, except for the moiré filter on the M9 shot. I think they are pretty similar in quality but "maybe" a bit less aliasing in the 5D shot still. The slight difference in angles may matter although I doubt it. I don't think this is enough difference to be a deal breaker and I think that it illustrates that at least stopped down, the systems have comparable detail and edge sharpness. So I don't think the AA filter deteriorates sharpness significantly. While the 5DII has 3 more mega pixels, I don't think this comes into play for this example that simply shows edges at 100% not overall resolution which would require a controlled comparison of the same subject. I think it comes down to any camera only shows its best detail when the image has been adjusted correctly. (With the best raw processor.) And surely all of us let this slide some and don't really get the most out of our current gear every time. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/153862-open-letter-to-leica-%E2%80%94-10-ways-to-improve-the-m9-rangefinder/?do=findComment&comment=1717692'>More sharing options...
Rick Posted June 28, 2011 Share #579 Posted June 28, 2011 Alan, did you ever look at the FAQ before you started this? See post 40 and 41 here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/130720-m9-faqs-frequently-asked-questions-answers-2.html Link to post Share on other sites More sharing options...
AlanG Posted June 28, 2011 Share #580 Posted June 28, 2011 Alan, did you ever look at the FAQ before you started this? See post 40 and 41 here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/130720-m9-faqs-frequently-asked-questions-answers-2.html I understand that about moiré. But these are aliased edges that are not caused by interference patterns and are obviously more common if they can occur on all lines of a given size in a building. At least they can easily be corrected (better than some moiré patterns) with the right software so I don't see why this can't be eliminated in camera firmware some day. Throwing you a total side issue. I found this type of c/a purple fringe from film images I shot with a Rollei 6006 using a Zeiss 40mm lens and scanned some years ago. I never noticed it at the time and no client ever complained. So things really have improved a lot today but many of us have also gotten a lot more critical. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/153862-open-letter-to-leica-%E2%80%94-10-ways-to-improve-the-m9-rangefinder/?do=findComment&comment=1717711'>More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.