carstenw Posted August 12, 2007 Author Share #21 Posted August 12, 2007 Advertisement (gone after registration) But overall, the frame lines aren't as badly off as I had anticipated. Maybe Leica's idea of setting them for closest focus isn't all that bad after all? I would disagree with both of those statements. Removing a strip of 10% all around will leave a 6MP envisioned photo, and most of the framelines are not far from that. Welcome back R-D1. This isn't an issue for everyone, but it is for some. At the moment, I try to be flexible about where I frame, and chimp to make sure that it is acceptable, but tight frames would be a godsend. Quote Link to post Share on other sites More sharing options...
Advertisement Posted August 12, 2007 Posted August 12, 2007 Hi carstenw, Take a look here Frameline Accuracy Comparison. I'm sure you'll find what you were looking for!
carstenw Posted August 12, 2007 Author Share #22 Posted August 12, 2007 By the way, when I was taking these shots, especially the 28s, I noticed for the first time how much distortion there is in the viewfinder, and how unsharp it gets towards the corners. Has anyone else noticed this? I would love that to be improved as well. Quote Link to post Share on other sites More sharing options...
Jamie Roberts Posted August 12, 2007 Share #23 Posted August 12, 2007 My own experience is that they're OK (though could be better) at 24 - 35 and 90. 50 is not so great and 75 is way off. Cheers, Sean This is exactly my feeling as well. The difference, especially, between the 90 (dead on almost close up) and the 75 (very conservative) is a bit disconcerting. Quote Link to post Share on other sites More sharing options...
carstenw Posted August 12, 2007 Author Share #24 Posted August 12, 2007 It occurs to me that if we wanted to know how bad the framing would be with the 1m being perfect, all we need to do is mentally subtract the 1m error from the errors at various distances. Is there anyone here who is handy enough with Photoshop to put up some examples? Quote Link to post Share on other sites More sharing options...
sean_reid Posted August 12, 2007 Share #25 Posted August 12, 2007 Just a reminder, BTW, that Carsten's tests (thanks again Carsten) relate to the specific lenses he tested. The results will vary somewhat depending on what specific 24, 28, 35, 50, 75 or 90 mm lens one is using. Sometimes the differences are quite noticeable because a group of lenses of the same nominal focal length may well have different actual focal lengths. Cheers, Sean Quote Link to post Share on other sites More sharing options...
Guest tummydoc Posted August 12, 2007 Share #26 Posted August 12, 2007 it would be great to have better frames so we could just forget about memorisation and go back to taking photos with a clear mind. Perhaps the discrepancy is a bit less than the M8 but my MP and M7 also require memorisation. Unless the frameline mechanics are completely redesigned so that they expand and contract correctly for each focal length fitted as the lens is focussed, the lines will always be accurate at only one subject distance, becoming less accurate at others. Whilst a (eg) 2m calibration might delight some users (myself included) it would no doubt infuriate others, and in any case would not obviate the need for memorisation. Quote Link to post Share on other sites More sharing options...
marknorton Posted August 12, 2007 Share #27 Posted August 12, 2007 Advertisement (gone after registration) True, but the existing frames seem pessimistic for all focal lengths and subject distances. With Sean's suggestion of a 1m standard, the inaccuracies might be better balanced for near and distant. The real solution is to adjust the frames electronically based on the coded lens mounted and the focussing distance. Currently, the frame is adjuste for parallax with image distance but not size. Quote Link to post Share on other sites More sharing options...
hankg Posted August 12, 2007 Share #28 Posted August 12, 2007 True, but the existing frames seem pessimistic for all focal lengths and subject distances. With Sean's suggestion of a 1m standard, the inaccuracies might be better balanced for near and distant. The real solution is to adjust the frames electronically based on the coded lens mounted and the focussing distance. Currently, the frame is adjuste for parallax with image distance but not size. Differences in actual focal length can be accounted for with coded lenses as well. So if one 50 is a 50mm while another is a 52mm frame size will be accurate for both. It will add another value added benefit to having Leica coded lenses. Quote Link to post Share on other sites More sharing options...
JGW Posted August 12, 2007 Share #29 Posted August 12, 2007 True, but the existing frames seem pessimistic for all focal lengths and subject distances. With Sean's suggestion of a 1m standard, the inaccuracies might be better balanced for near and distant. The real solution is to adjust the frames electronically based on the coded lens mounted and the focussing distance. Currently, the frame is adjuste for parallax with image distance but not size. Differences in actual focal length can be accounted for with coded lenses as well. So if one 50 is a 50mm while another is a 52mm frame size will be accurate for both. It will add another value added benefit to having Leica coded lenses. Hank Graber This would be fantastic. Presumably this would require a hardware change (or M9) or could this be done via firmware? Perhaps the frameline select lever would be redundant, but that wouldn't matter too much. JGW Quote Link to post Share on other sites More sharing options...
marknorton Posted August 12, 2007 Share #30 Posted August 12, 2007 Differences in actual focal length can be accounted for with coded lenses as well. So if one 50 is a 50mm while another is a 52mm frame size will be accurate for both. It will add another value added benefit to having Leica coded lenses. Indeed, having electronic framelines which match the particular lens in use makes a lot of sense. Quote Link to post Share on other sites More sharing options...
carstenw Posted August 12, 2007 Author Share #31 Posted August 12, 2007 I outlined a suggestion of mirroring in LCD frames which take into acount both focal length and focusing distance, but without changing the lens mount and all our lenses, probably impossible for the latter, there is no way to go any further than that. I did outline the idea to the Leica mechanic who visited Leica Camera Berlin a couple of months ago, and he passed it to Stefan Daniel, so who knows, maybe there will be something coming in the future. Quote Link to post Share on other sites More sharing options...
marknorton Posted August 12, 2007 Share #32 Posted August 12, 2007 This would be fantastic. Presumably this would require a hardware change (or M9) or could this be done via firmware? Perhaps the frameline select lever would be redundant, but that wouldn't matter too much. Sadly, that is for the future. The current framelines are created using two metal screens with slots cut in them and as one moves over the other, so the different framelines appear. That's why they are incomplete as well. If they were complete rectangles, there would be nothing to hold the centre of the screen. An electronic version would consist of an LCD screen, normally black where pixels are made clear to allow the frame illumination to pass through. Depending on the actual lens and the focussing distance, measured by the displacement of the rangefinder roller, the frames would bet set to match the lens and move/adjust with focussing distance. There would still need to be default frame sets for uncoded lenses based on the frame selection lever (allowing preview of them as well) and if I was implementing it, I'd let the user choose "classic" frame pairs or single frames to match the lens. Quote Link to post Share on other sites More sharing options...
gwelland Posted August 12, 2007 Share #33 Posted August 12, 2007 So is the colour shift in each of these images just due to AWB or poor white balance selection in post processing? Or are all these taken with no filters, similar filters or a mix? Quote Link to post Share on other sites More sharing options...
Guest guy_mancuso Posted August 12, 2007 Share #34 Posted August 12, 2007 Every Leica lens has a code on it, you can build the code into the electronics to bring up the exact lens attached to the camera and in turn shows the exact frame lines for it in the finder. This is not very hard to do Quote Link to post Share on other sites More sharing options...
marknorton Posted August 12, 2007 Share #35 Posted August 12, 2007 The theory is easy to state but I wouldn't say it's "not very hard" to actually implement. You might start with a 1024 * 768 LCD and worry about whether the "black" is black enough under brilliant sunshine and the "clear" is clear enough to illuminate the frame lines in low light. You'd also worry about having framelines when the camera is off, so you wouldn't be able to look through the finder to check framing with the camera off. Your next problem would be to measure the displacement of the focussing roller over its 3mm movement range to at least 8 bit accuracy and provide a way of calibrating it. Add to that the tiny space it all has to fit in and it's quite a challenge. Quote Link to post Share on other sites More sharing options...
sean_reid Posted August 12, 2007 Share #36 Posted August 12, 2007 Differences in actual focal length can be accounted for with coded lenses as well. So if one 50 is a 50mm while another is a 52mm frame size will be accurate for both. It will add another value added benefit to having Leica coded lenses. Hank Graber This would be fantastic. Presumably this would require a hardware change (or M9) or could this be done via firmware? Perhaps the frameline select lever would be redundant, but that wouldn't matter too much. JGW It couldn't be firmware. It would need to be a whole new system within the viewfinder,etc.. I have mixed feelings about the idea although I can appreciate the theoretical advantages. Cheers, Sean Quote Link to post Share on other sites More sharing options...
hankg Posted August 13, 2007 Share #37 Posted August 13, 2007 It couldn't be firmware. It would need to be a whole new system within the viewfinder,etc.. I have mixed feelings about the idea although I can appreciate the theoretical advantages. Right now Leica has got it's plate full with the M8 and the launch of the presumed to be coming R10. Messing with the viewfinder is a very delicate undertaking as it is at the heart of the M's 'Mness'. I think it makes sense to do and the technology exists to do it right, but screwing it up would be a disaster. So there would probably be some extensive usability tests to make sure it preserves the look and feel and everything that is good in the present setup. So if it gets done at all it is likely something for an M9. Quote Link to post Share on other sites More sharing options...
hankg Posted August 13, 2007 Share #38 Posted August 13, 2007 I'd add one more caveat. Leica really has no competition for it's RF finder. It's still the best out there. So they might be reluctant to invest resources in an area where they have no competition. Even if a another brand digital RF was to be released it would be in the sensor or price where they would probably be facing competition not in the finder which like every other competitors finder since the M3 will likely be a notch below what Leica has. Now if a new design reduced manufacturing costs while increasing reliability and accuracy that might be worth doing even without competition. Quote Link to post Share on other sites More sharing options...
jaapv Posted August 13, 2007 Share #39 Posted August 13, 2007 So is the colour shift in each of these images just due to AWB or poor white balance selection in post processing? Or are all these taken with no filters, similar filters or a mix? Given the purpose of the test, I would venture that these questions were not on the OP's list of priorities. Quote Link to post Share on other sites More sharing options...
carstenw Posted August 13, 2007 Author Share #40 Posted August 13, 2007 Graham, I don't know exactly what the problem was. The AWB was a little off in a couple of shots, but I selected all the shots in each series and used the best white balance I could find for the series as a unit. For some reason some of the shots show white balances which look wrong after that, and I don't know the reason. I was outside, and it was in the daytime, so a simple daylight setting should have sufficed, and I should not have had to do even the minor fiddling that I did. I did not want to white balance each shot individually, since this was about framing and there were 40 shots, and I was just trying to get it posted. Quote Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.