JimKasson Posted January 28, 2014 Share #4241 Posted January 28, 2014 Advertisement (gone after registration) Not to my eye. See here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/318679-monochrom-high-iso-test.html#post2615983 Jay, thanks for doing the experiments, and thanks for starting a new thread, as this high ISO stuff was drifting away from the a7R. I'll comment on the new thread, but I wanted to say a few things here that are not MM-specific. The first is that the only useful way to compare noise performance by looking at images is to either look at uncompressed processed images at sensor resolution or look at the raw files themselves. Raw files are awkward because of their size. TIFFs cropped from demosaiced images are acceptable substitutes, if all of the raw conversion settings are stated so that the results can be reproduced. Even TIFFs can be large, so what I have occasionally done in the past is to enlarge a section of an uncompressed, processed image by 200% or 300% using nearest neighbor an JPEGing the result. The discrete cosine transform in the lossy part of the JPEG standard removes noise (and other high-spatial-frequency information) by design, but if the pixels are expanded sufficiently, the noise doesn't look so high-frequency to the DCT algorithm. It's also helpful to make it easy for the viewer to save the file on her own machine so that she can use the image editor of her choice to do the viewing, and perhaps set up images to be compared as layers in a single file. The second it that you may be entirely right in your conclusions with respect to the MM. I don't dispute them, as I haven't gotten my mitts on one, and I can't look at your files in sufficient detail to see what's going on. The third, and last, is that, for those of you with a technical bent, there's an active thread right now on Lula on high ISO. Thanks again, Jim Link to post Share on other sites More sharing options...
Advertisement Posted January 28, 2014 Posted January 28, 2014 Hi JimKasson, Take a look here The Sony A7 thread [Merged]. I'm sure you'll find what you were looking for!
Jeff S Posted January 30, 2014 Share #4242 Posted January 30, 2014 Roger from lensrentals pulls a 'marknorton' and tears down the A7R here. Jeff Link to post Share on other sites More sharing options...
dwbell Posted January 30, 2014 Share #4243 Posted January 30, 2014 Roger from lensrentals pulls a 'marknorton' and tears down the A7R here. Jeff And pointed out that replacing the cover glass (with a third party thinner one?) is only three spring clips away. In theory. Link to post Share on other sites More sharing options...
chris_livsey Posted January 30, 2014 Share #4244 Posted January 30, 2014 And pointed out that replacing the cover glass (with a third party thinner one?) is only three spring clips away. In theory. And would potentially spoil the native glass results as they are "tuned" to that combination as shipped? Link to post Share on other sites More sharing options...
dwbell Posted January 30, 2014 Share #4245 Posted January 30, 2014 And would potentially spoil the native glass results as they are "tuned" to that combination as shipped? I would imagine so, yes. The speculation (educated guess?) here at the moment is that the M240 has much thinner cover glass which allows for better corner performance with WA Leica M lenses, but at the cost of IR sensitivity showing in the blacks (as purples etc). So if one wanted to match the A7R specifically to the M lenses, then a thinner 3rd party glass might offer an option. Link to post Share on other sites More sharing options...
k-hawinkler Posted January 30, 2014 Share #4246 Posted January 30, 2014 What the article demonstrated was the relative simplicity of the A7R design and its components. Certainly commensurate with its price tag. Exquisite engineering also with regards to repairability. I wish to see in a follow on model IBIS and a better shutter. IBIS would increase complexity and size. But well worth it IMHO. Link to post Share on other sites More sharing options...
dwbell Posted January 30, 2014 Share #4247 Posted January 30, 2014 Advertisement (gone after registration) What the article demonstrated was the relative simplicity of the A7R design and its components.Certainly commensurate with its price tag. Exquisite engineering also with regards to repairability. I wish to see in a follow on model IBIS and a better shutter. IBIS would increase complexity and size. But well worth it IMHO. I'd welcome a cutting edge IBIS, as my higher shutter speeds to accommodate the 36MP resolution almost negate any high ISO gain the sensor has on other cameras. So for me the IBIS would be kind of a necessity for every shot to come back to 1xfl, rather than a trick tech for very low light. I'm confident though, that the next iteration will again be a technological tour de force, and will be not long in coming. Link to post Share on other sites More sharing options...
k-hawinkler Posted January 30, 2014 Share #4248 Posted January 30, 2014 I agree. A state of the art IBIS, as in the E-M1, would compensate for about 4 stops of shakiness on the part of the camera operator. Then choice of shutter speed would be driven by object motion at more moderate ISO levels. The future is bright. Link to post Share on other sites More sharing options...
JimKasson Posted January 30, 2014 Share #4249 Posted January 30, 2014 I wish to see in a follow on model IBIS and a better shutter. IBIS would increase complexity and size. But well worth it IMHO. However, I think it would work only with native lenses. IBIS looks at the lateral and rotational motion of the camera, calculates the image displacement, and moves the sensor to compensate, right? In order know what's happening to the image, it needs to know the lens focal length. If the subject is within, say, 20 or 30 focal lengths, it needs to know the subject distance. If the lens uses internal focusing in whole or in part, it needs to know how the effective focal length changes with distance. I can't figure out how the camera could know those things with 3rd party lenses. One of the unique -- in my experience, anyway -- characteristics of the a7R is the high-frequency nature of the shutter shock. An IBIS tuned to deal with handheld motion wouldn't be able to deal with that. An IBIS tuned to deal with the shutter shock runs the risk of exciting sympathetic vibrations in the tripod assembly, and having to deal with them, and exciting further sympathetic vibrations in the tripod assembly, and on and on. So, while IBIS may help some things, I don't think it's the magic bullet for dealing with the a7R's current shutter problems. BTW, we now know we can blame Copal for the shutter shock. Jim Link to post Share on other sites More sharing options...
chris_livsey Posted January 30, 2014 Share #4250 Posted January 30, 2014 BTW, we now know we can blame Copal for the shutter shock. Jim Or Sony for how they have mounted it Link to post Share on other sites More sharing options...
k-hawinkler Posted January 30, 2014 Share #4251 Posted January 30, 2014 However, I think it would work only with native lenses. IBIS looks at the lateral and rotational motion of the camera, calculates the image displacement, and moves the sensor to compensate, right? In order know what's happening to the image, it needs to know the lens focal length. If the subject is within, say, 20 or 30 focal lengths, it needs to know the subject distance. If the lens uses internal focusing in whole or in part, it needs to know how the effective focal length changes with distance. I can't figure out how the camera could know those things with 3rd party lenses. One of the unique -- in my experience, anyway -- characteristics of the a7R is the high-frequency nature of the shutter shock. An IBIS tuned to deal with handheld motion wouldn't be able to deal with that. An IBIS tuned to deal with the shutter shock runs the risk of exciting sympathetic vibrations in the tripod assembly, and having to deal with them, and exciting further sympathetic vibrations in the tripod assembly, and on and on. So, while IBIS may help some things, I don't think it's the magic bullet for dealing with the a7R's current shutter problems. BTW, we now know we can blame Copal for the shutter shock. Jim Hi Jim, May I suggest that you inform yourself a little better how the IBIS works in the E-M5 and E-M1? Link to post Share on other sites More sharing options...
JimKasson Posted January 30, 2014 Share #4252 Posted January 30, 2014 Hi Jim, May I suggest that you inform yourself a little better how the IBIS works in the E-M5 and E-M1? I'd be pleased to do so, and, in the service of that end, I'd appreciate any links you could pass along. Thanks, Jim Link to post Share on other sites More sharing options...
k-hawinkler Posted January 30, 2014 Share #4253 Posted January 30, 2014 I'd be pleased to do so, and, in the service of that end, I'd appreciate any links you could pass along. Thanks, Jim Thanks Jim. I replied to your PM and listed some articles to that effect. Link to post Share on other sites More sharing options...
chris_livsey Posted January 30, 2014 Share #4254 Posted January 30, 2014 Jim, forgive me if this appears in your blog entries somewhere I regret work (paid) prevents reading every day. I understand that both high and low speeds will be least affected by the shutter, so what are the practical limits? ie how low, how high, where does it become 'iffy" assuming the cut is not clean but progressive? EDIT Looks like I am a day or so early reading your latest :-) Link to post Share on other sites More sharing options...
JimKasson Posted January 31, 2014 Share #4255 Posted January 31, 2014 I understand that both high and low speeds will be least affected by the shutter, so what are the practical limits? ie how low, how high, where does it become 'iffy" assuming the cut is not clean but progressive? Chris, I wish there were a simple answer to your question. It depends on the lens length and mass, the plate, the head, the tripod, and a lot of other things. I’ll get to that, but first I want to talk about something on which it depends that for the most part, goes unnoticed: the lens resolution. I have a group of technically-interested photographers, among them Huntington Witherill and Mike Collette. Both have made guest posts – here's Mike's – in my blog about the a7R shutter shock. I’ve also been corresponding with other people about the problem. Recently, someone announced that, using the Witherill lens mounting method, that he had completely tamed the a7R shutter shock for a Canon zoom for all shutter speeds and all focal lengths out to 200mm. Well, I was amazed. Being from Missouri, I asked to see files from shots of ISO 12233 targets. When I looked at them, I saw that the lens was not resolving detail to the sensor’s pitch, or anywhere near it. Thus, the shutter shock motion was buried in the lens fuzziness. Thus, the photographer who claimed to have solved the shutter shock problem was right, for that lens. I’m pretty sure he would have had the problem with an optic of greater resolving power. I’ll give you some general guidelines. As the lens and ancillary gear gets more massive, the resonant frequency tends to drop. This has two effects: the high-frequency component of the shutter shock tends to be restrained by the mass, and the wobbling tends to go on longer. The first effect makes the high safe shutter speed lower, and the second effect makes the low safe shutter speed lower. Mounting the camera and lens to a rail and supporting the lens barrel tends to make the high safe shutter speed lower, and the low safe shutter speed higher, if the head is clamped to the rail under the camera, not under the center of gravity of the assembly. Leaving out the mass effect of the lens, shorter lenses make the high safe shutter speed lower, and the low safe shutter speed higher. With the Zony 55mm lens, the RRS L-bracket, a head like the Arca Swiss C1 or the big RRS ball head and a thick set of carbon fiber legs, you want to stay away from 1/100, and the further the better. Since it’s a darned sharp lens, it’s hard to find high shutter speeds that are completely uncontaminated. 1/1000 and up is pretty safe, and you can go down to 1/320 with what would be acceptable motion contamination in most circumstances. By 1/4 or 1/2 second, things are generally good again. Here's a series with the 135mm APO-Telyt that shows similar effects.I’d like to emphasize here that the ISO 12233 target is an extremely demanding subject and you’ll see motion effects on it that you wouldn’t see with almost all real-world subjects. Once you find suspicious shutter speeds with the ISO 12233 target, you should try them on real subjects to see if you can see any effect there. The amount of acceptable shutter shock will also depend on the use you intend to make of the images. I bought the a7R thinking it could replace the D800E for some very demanding work and give me better live view to boot. I've since gone back to the D800E for that work. The high frequency component of the a7R shutter shock is so strong that with lenses of 55mm to 135mm (at least, and maybe others) a 1/1300 second (t.1) strobe can’t stop all the motion if you use first curtain synch. All of the above is for the camera in landscape orientation. In portrait mode, things get worse. If you want to ask me questions about specific lens/plate/head/tripod combinations, I’ll do my best, but I probably can’t reproduce your setup. I encourage you to do your own testing. Download and print out the ISO 12233 target, put a good variable ND filter in front of your lens, set ISO and f-stop to a constant, and make a series of exposures at 1/3 stop increments of shutter speed. Use a distance that gets the inside of the outer black square to about 700 pixels. You don’t have to use my fancy 5500K continuous lighting setup; put the target in direct sunlight – this should allow you to set the ISO to 100. Take your time. These series are laborious and prone to error. The reason I haven’t posted the continuous/strobe 5500K a7R/Leica 135mm APO-Telyt images is that I blew it when I made the series yesterday. I made a 1/3 stop series with no lens support, then a 1/3 stop series with a lens support, then removed the lens support, turned out the continuous light, set the shutter speed to 4 seconds, and made a trailing-curtain synched image with the Einstein set to 2.5 watt seconds, which gives a 1/13000 t.1. The strobe picture was not as sharp as the first 1/250 second continuous lighting picture. Something must have slipped. I’ll have to repeat the whole series. Jim Link to post Share on other sites More sharing options...
JimKasson Posted January 31, 2014 Share #4256 Posted January 31, 2014 Here's a link to the (old) ISO 12233 target: ISO 12233 Test Chart Jim Link to post Share on other sites More sharing options...
chris_livsey Posted January 31, 2014 Share #4257 Posted January 31, 2014 Chris, I wish there were a simple answer to your question. Jim How's that for a snip ? Indeed if I had engaged my brain I would have worked out some of those variables. A problem I see is that it will be possible to use a combination of good legs suitable floor good head meticulous set up and find a sweet spot speed for a particular lens but will that translate to real life? What I mean is the dampening etc in the "best" set up may be just the right weight, frequency etc and that hand held at that speed will not be as good perhaps as a different speed. No way does this negate the work done as the more precise the measurements the easier the theory is to develop and point to remedial action. With these lenses and sensor resolution hand held is never going to "cut it". With my meagre P20 16MP on the 'blad mirror lock up is best (better) hand held, which makes for interesting framing. On the Gitzo/Manfrotto geared head all locked down it can be spectacular. (vote for leaf shutters now). Thanks for the target link, printed now, better than my sheet of graph paper Keep up the good work. BTW at work our (Hospital) patient records and e-prescribing software system ran an update and all the Windows 7 machines, and only those, needed a patch, only after deployment of course, testing is tedious but essential. BTW not sure if this was intended, if it was brilliant, if not it should have been!! "Zony 55mm lens"and DxO certainly think it is sharp they must have mounted it well Link to post Share on other sites More sharing options...
JimKasson Posted January 31, 2014 Share #4258 Posted January 31, 2014 (vote for leaf shutters now). I'll say. I'm dusting off my old H2D39. But there are a lot of downsides: Firewire, Phocus, weight, etc. I don't know if I have the patience and stamina any more. testing is tedious but essential. That's what I keep telling myself. I'd much rather be making real pictures. BTW not sure if this was intended, if it was brilliant, if not it should have been!! "Zony 55mm lens"and DxO certainly think it is sharp they must have mounted it well Yes, intended. I didn't think of it, I just copied someone else. Joe Holmes, I think. Jim Link to post Share on other sites More sharing options...
jankap Posted February 1, 2014 Share #4259 Posted February 1, 2014 For leaf shutters one needs a new set of lenses, because this type of shutter should be inside the lens (near the diaphragm). Otherwise vignetting would be the next problem to solve. Would a wooden tripod make sense? The problem is, that a metal tripod tends to swing. I think, metal destroys movement more slowly than wood. Meaningful in this respect is, that the portrait and landscape modes are so different (rotating versus nodding). But how to measure? A wood construction with the same weight as a metal one? Jan Link to post Share on other sites More sharing options...
k-hawinkler Posted February 1, 2014 Share #4260 Posted February 1, 2014 CF tripod. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.