sean_reid Posted October 18, 2006 Share #21 Posted October 18, 2006 Advertisement (gone after registration) I should be able to confirm the following further over the next few days but here is what I *believe* to be true: 1. The RAW file truly is RAW and is not affected in-camera by the 6-bit processing. It is tagged with the 6-bit info. which can potentially be used by a RAW converter. 2. C1 does not currently correct vignetting, even though the data may be tagged in the file. 3. C1 4.0 will offer vignetting compensation that may well make use of the Leica 6-bit data. To be sure, the current version of C1 does support the M8. Lest people be too concerned about this, my experience so far suggests to me that many lenses on the M8 will not vignette enough to need any compensation so very often this will be a moot issue. Cheers, Sean Link to post Share on other sites More sharing options...
Advertisement Posted October 18, 2006 Posted October 18, 2006 Hi sean_reid, Take a look here Some very interesting news about C1 .... I'm sure you'll find what you were looking for!
scott kirkpatrick Posted October 18, 2006 Share #22 Posted October 18, 2006 While you're at it, can you share any information about how the reduction to 8 bits per pixel in the .dng is achieved? There are several (? maybe only one) posters claiming that some nonstandard, and probably lossy code is employed. There are enough inconsistencies in the posts that i suspect a red herring, but would like to know more. Another suggestion, while I've got the floor. In your LL and RR articles comparing the 21 and 24/5 Leica and Zeiss lenses you've tended to use standard exposure practices, and then say that one may blow highlights a bit, while the other pushes the shadows into the basement. That was with the RD-1. With the M8 and its Kodak sensor, I would expect exceptional dynamic range. Can you get a handle on this by using both Zeiss and Leica lenses, backing down the exposure on the brighter one so that both place the highlights at the same place, where they can be rendered nicely, and then compare shadow detail? cheers, scott Link to post Share on other sites More sharing options...
Jamie Roberts Posted October 19, 2006 Share #23 Posted October 19, 2006 {snipped}1. The RAW file truly is RAW and is not affected in-camera by the 6-bit processing. It is tagged with the 6-bit info. which can potentially be used by a RAW converter. 2. C1 does not currently correct vignetting, even though the data may be tagged in the file. {snipped} Sean Sean, this is what I was trying to get at... thanks for explaining it so clearly. Adding metadata in the DNG file in the "private space" would not mean the RAW file was not RAW, but that a converter could (not "must") take advantage of the information. Heck--it could be tagged in the EXIF information--or even the file header--and this would not affect the capture RAW image data. It's metadata. This is why you don't *have* to code your lenses. So I'm sorry if I gave the impression that some funny processing was going on inside the RAW itself. You would never do this, precisely for the reason JR was mentioning. But that doesn't mean you can't send a little message to the RAW program saying, basically, "we think you should interpret this edge data this way, relative to the rest of the file; please feel free to ignore our suggestions". What I was really curious about was "is the metadata in the RAW file the actual correction data, or only the 6 bit encoding?" In other words, would the raw converter need some correction lookup table from Leica, or is the lookup actually carried in the file itself? I'd rather it go with the file, in some ways. And even RAW converters with no idea of the special tagging would simply ignore it, but you could conceivably get much better results based on what Leica knows about the edge performance with certain lenses. This is why DNG has a "private space" for individual manufacturer's "secret sauce." With 16 bits per channel of sensor data, you can make any number of subtle suggestions for the RAW converter to follow without adversely affecting anything at all! Let's not forget--RAW files are always interpreted... the "image" in them is not absolute, and different converters do different things. I want to clarify too that what we're talking about here is is NOT normal lens vignetting (and certainly nothing to do with barrel or pincushion distortion) but, according to Leica, is the result of building a digital rangefinder. The lens is that much closer to the sensor plane than SLRs, so the light fall-off is significantly different--and different than it was with film, too. But the effect--whatever it is--is sure to be subtle in any case. And it doesn't matter to me, really, if C1 supports this now or not, though I'll be happy to see C1 V4. Link to post Share on other sites More sharing options...
Riley Posted October 19, 2006 Share #24 Posted October 19, 2006 hmmm coded lenses, an ability to relate that code in camera Oly are going this road too corrections for minor lens distortions applied in exact form ? this could get interesting Riley Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 19, 2006 Share #25 Posted October 19, 2006 I want to clarify too that what we're talking about here is is NOT normal lens vignetting (and certainly nothing to do with barrel or pincushion distortion) but, according to Leica, is the result of building a digital rangefinder. The lens is that much closer to the sensor plane than SLRs, so the light fall-off is significantly different--and different than it was with film, too. Jamie, The beauty of this is that since the vignetting corrections are determined from a particular body/lens combo tests rather than lens vignetting information/graphs such as contained in the Leica brochures, the RAW developer that would correct the vignetting from all "sources" of this vignetting (due to lens or sensor). Link to post Share on other sites More sharing options...
Guest guy_mancuso Posted October 19, 2006 Share #26 Posted October 19, 2006 It may have the correction data but right now it is worthless since no raw converter can see it or use it. So if it is there or not is a moot point at this time. Just as long as there not doing anything to the raws , i really don't care . i doubt very much we will have big vingetting issues anyway. My only concern and this maybe for another thread is the need to turn off the lens coding in camera. I am certainly not going to go back and forth changing that when i change a lens. I need to also read what they say about this. I just glanced at the manual. Link to post Share on other sites More sharing options...
sean_reid Posted October 19, 2006 Share #27 Posted October 19, 2006 Advertisement (gone after registration) It may have the correction data but right now it is worthless since no raw converter can see it or use it. So if it is there or not is a moot point at this time. Just as long as there not doing anything to the raws , i really don't care . i doubt very much we will have big vingetting issues anyway. My only concern and this maybe for another thread is the need to turn off the lens coding in camera. I am certainly not going to go back and forth changing that when i change a lens. I need to also read what they say about this. I just glanced at the manual. Hi Guy, I can confirm in a couple of days but I don't think you'll need to bother when shooting RAW because the camera won't be processing the file with that info. any way. Perhaps the EXIF would be off, we'll see. The Leica lenses arriving tomorrow are coded. Cheers, Sean Link to post Share on other sites More sharing options...
Guest guy_mancuso Posted October 19, 2006 Share #28 Posted October 19, 2006 Sean maybe try turning off the coding in camera and just read what is missing from the EXIF data Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 19, 2006 Share #29 Posted October 19, 2006 It may have the correction data but right now it is worthless since no raw converter can see it or use it. So if it is there or not is a moot point at this time. BTW, you can do like me an use 3rd party plug-ins and applications and use the 6-bit coding info (aka EXIF). I do this successfully with distortion for my R lenses (50mm and below and zooms) (I don't care much about vignetting. Link to post Share on other sites More sharing options...
osera Posted October 20, 2006 Share #30 Posted October 20, 2006 BTW, you can do like me an use 3rd party plug-ins and applications and use the 6-bit coding info (aka EXIF). I do this successfully with distortion for my R lenses (50mm and below and zooms) (I don't care much about vignetting. JR- A bit OT here, but what applications are you using for R lens distortion? I presume you will try the same on the M8 files. -Allen Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 20, 2006 Share #31 Posted October 20, 2006 JR- A bit OT here, but what applications are you using for R lens distortion? I presume you will try the same on the M8 files. -Allen Allen, I calibrated all my lenses and now have my own calibration of the a, b, c, d parameters for Panotools. I have virtually no distortion left in my images. Link to post Share on other sites More sharing options...
sean_reid Posted October 20, 2006 Share #32 Posted October 20, 2006 Sean maybe try turning off the coding in camera and just read what is missing from the EXIF data Hi Guy, Yes, I'll play with various things to get a sense of this. Cheers, Sean Link to post Share on other sites More sharing options...
Jamie Roberts Posted October 20, 2006 Share #33 Posted October 20, 2006 Allen, I calibrated all my lenses and now have my own calibration of the a, b, c, d parameters for Panotools. I have virtually no distortion left in my images. JR--I completely agree with what you're saying. I use Panotools for the R lenses as well, though I just picked up the 19 and 28, so I'll need to try to figure that out. The current 50 lux doesn't really need correction, do you think? I don't really care about vignetting with the current R lenses anyway, because they don't really vignette much (well, maybe the 80 lux wide open, but I like that). But the older lenses (older summilux 50, for example) wow does that ever vignette! Anyway, back on topic--the vignette here isn't part of the lens design / trade offs; I don't think Panotools would help. I thought the coding was all part of trying to ensure compatiblity with the lens system itself, given the built-in vignette you will get with the shorter film plane. It will get worse, if I understand the M8 issues, the wider you go. Sean--so you think the whole coding thing is for an in-camera-JPEG-only adjustment!? That'll be interesting! Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 20, 2006 Share #34 Posted October 20, 2006 JR--I completely agree with what you're saying. I use Panotools for the R lenses as well, though I just picked up the 19 and 28, so I'll need to try to figure that out. The current 50 lux doesn't really need correction, do you think? I don't really care about vignetting with the current R lenses anyway, because they don't really vignette much (well, maybe the 80 lux wide open, but I like that). But the older lenses (older summilux 50, for example) wow does that ever vignette! Anyway, back on topic--the vignette here isn't part of the lens design / trade offs; I don't think Panotools would help. I thought the coding was all part of trying to ensure compatiblity with the lens system itself, given the built-in vignette you will get with the shorter film plane. It will get worse, if I understand the M8 issues, the wider you go. Sean--so you think the whole coding thing is for an in-camera-JPEG-only adjustment!? That'll be interesting! Jamie, The distortion on the 50mm f/1.4R (latest) is minimal. Follow the link to see a before/after sample I prepared for you. Leica 50mm f/1.4 Distortion JR Link to post Share on other sites More sharing options...
jlm Posted October 20, 2006 Share #35 Posted October 20, 2006 my guess is the vignetting effect is more an issue of the sensor than the lens per se, given that the M8 is using the center out of the lens image circle. What the lens will still contribute is the angle of incidence at the sensor; worse with wide angle lenses and deep rear elements Link to post Share on other sites More sharing options...
rvaubel Posted October 20, 2006 Share #36 Posted October 20, 2006 Jamie, The distortion on the 50mm f/1.4R (latest) is minimal. Follow the link to see a before/after sample I prepared for you. Leica 50mm f/1.4 Distortion JR JR Wow! that demonstration was very cool. Could you show a sample of a lens that does distort, say the new 35mm lux or the 28mm cron ? I'm impressed with the software. Rex Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 20, 2006 Share #37 Posted October 20, 2006 JR Wow! that demonstration was very cool. Could you show a sample of a lens that does distort, say the new 35mm lux or the 28mm cron ? I'm impressed with the software. Rex Here is the 15mm f2.8R before/after distortion correction Leica 15mm f/2.8 R Distortion Sample - JR Geoffrion Link to post Share on other sites More sharing options...
Jamie Roberts Posted October 20, 2006 Share #38 Posted October 20, 2006 Here is the 15mm f2.8R before/after distortion correction Leica 15mm f/2.8 R Distortion Sample - JR Geoffrion JR--nice demonstrations! And exactly what I suspected on the 50 as well. That 15 is absolutely astoundingly good. My great heavens! I know it's cropped and in the center, but holy heck man that's a well corrected wide. It looks even better than the Elmarit 19, or my Oly 21, both of which are really well corrected! Now the big question--did you actually do the testing with Pano tools to get the right corrections for each lens from scratch, or did you use one of the starter sets and tweak? Understand, I'm not asking for your corrections (though I'd happily discuss buying them from you!) but I'd like to know your method. I've tried doing this from scratch with limited success... PS--can you tell we're killing some time till the M8 is here Link to post Share on other sites More sharing options...
jrgeoffrion Posted October 20, 2006 Share #39 Posted October 20, 2006 Now the big question--did you actually do the testing with Pano tools to get the right corrections for each lens from scratch, or did you use one of the starter sets and tweak? Understand, I'm not asking for your corrections (though I'd happily discuss buying them from you!) but I'd like to know your method. I've tried doing this from scratch with limited success... PS--can you tell we're killing some time till the M8 is here Although some of the lenses could be well corrected (but not "perfectly") with the b parameter (the most flexible) or the a-b combination, I've determined the a, b, and c parameters from scratch for all my lenses for better results. The problem of determining 3 or 4 variables is that you need one more equation than the number of unknowns to be able to solve the equations. I get the equations by using up to 9 images and matching point combinations. Just for fun, in a couple of weeks, I'll do a comparison of my method vs. the PTLens 'stock' calibrations (which mostly use 1 or 2 parameters) and see if I'm wasting my time. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.