t024484 Posted June 13, 2011 Share #161 Posted June 13, 2011 Advertisement (gone after registration) Lindolfi, Yes, not 100% of the pixel surface is available for receiving light. Therefore I support your theory that this could explain different sensitivities for light coming from different directions. But when all pixels have the same physical orientation, this effect will be the same for R,G and B unless the shading effect is different for different wavelengths. Hans Quote Link to post Share on other sites More sharing options...
Advertisement Posted June 13, 2011 Posted June 13, 2011 Hi t024484, Take a look here Towards an explanation of the Italian Flag Phenomenon. I'm sure you'll find what you were looking for!
Lindolfi Posted June 13, 2011 Author Share #162 Posted June 13, 2011 Hans, that is exactly what I meant with "This can only be an explanation if the different colour wells (for red green and blue) act differently in their directionality response to incoming light rays.". So I agree. Quote Link to post Share on other sites More sharing options...
sydkugelmass Posted June 14, 2011 Share #163 Posted June 14, 2011 (edited) Some comment on this? leica.overgaard.dk - Thorsten Overgaard's Leica Pages - Leica M9 Digital Rangefinder Camera - Page 13: "Learning the simplicity of the Leica M9" Section : "What does the lens coding do?" Edited June 14, 2011 by sydkugelmass Quote Link to post Share on other sites More sharing options...
sandymc Posted June 14, 2011 Share #164 Posted June 14, 2011 3] Part of the IFP occurs in the conversion of the raw data in the DNG to the RGB values.Part of it occurs during capture by the sensor. This needs to be further explored by doing a hand-coded demosaic on the raw data of the DNG. (In fact you can explain the final IFP by performing a left-to-right "correction" on top of the top-to-bottom discolouration from the sensor: Where does that left-to-right "correction" happen in the loading process of the raw data from the DNG?) Not clear to me why you suggest any role for demosaicing/raw conversion in this. But you can test easily: 1. Adobe's dng_validate program will do four color output if you want to test without the effects of demosaicing/raw conversion. 2. Dcraw is configurable to do either four color output or a variety of different types of demosaicing - including a very simple linear interpolation (q = 0). But I very much doubt you'll see a difference. Sandy Quote Link to post Share on other sites More sharing options...
t024484 Posted June 14, 2011 Share #165 Posted June 14, 2011 Lindolfi, I am Trying to answer one of the earlier questions wether there is a large spread between cameras. For that reason I used your image pinholeto28nocode and mine from posting #159 taken also with a pinhole at 28mm from the sensor. First I have increased the exposure short before clipping within Camera Raw. As a next step I have set the whitepoint of the image in the centre with Photoshop's levels command.. Then I glued two halves together, yours is the one at the bottom. It is quite obvious from this that your image has more vignetting and also more colorshift, although the IFP exists in both images. So if we did everything in a similair way, it seems as if the sensors are not that equal. Hans 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! Quote 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/153464-towards-an-explanation-of-the-italian-flag-phenomenon/?do=findComment&comment=1701797'>More sharing options...
Lindolfi Posted June 14, 2011 Author Share #166 Posted June 14, 2011 Some comment on this? leica.overgaard.dk - Thorsten Overgaard's Leica Pages - Leica M9 Digital Rangefinder Camera - Page 13: "Learning the simplicity of the Leica M9" Section : "What does the lens coding do?" [1] "blue/purple fringing by hightlights" is not handled with the coding of lenses. [2] The figure that Thorsten used that I have drawn for this thread on Italian Flag Phenomenon (IFP) is used to explain about microlenses, while the figure is drawn to demonstrate the effect of chromatic aberration of the microlenses in a theoretical consideration. The caption under the figure reads "How the light rays arrive to the sensors micro-lenses. The wider the lens, the less straight they arrive, and that is one of the things the 6-bit coding try to fix.". What is meant here is that not the less straight property of the light rays is corrected, but their subsequent effect on the discolouration (IFP). All in all I am glad that Thorsten addresses the issue in his fine set of pages on the Leica M9. Quote Link to post Share on other sites More sharing options...
Lindolfi Posted June 14, 2011 Author Share #167 Posted June 14, 2011 Advertisement (gone after registration) Not clear to me why you suggest any role for demosaicing/raw conversion in this Sandy Thanks Sandy. Indeed I will look into the process of demosaicing and I'll do that because I found a difference in pattern in the image coming from the LightRoom RAW converter and my own reading of the three colour components from the raw part of the DNG file (see the image I posted here and compare with the images from the RAW converter elsewhere in this thread). I'll report later what I found. But I agree that there should be no difference. On the other hand it may be that there is some correction code used by the RAW converter that is part of the DNG file that we are not aware of. Quote Link to post Share on other sites More sharing options...
Lindolfi Posted June 14, 2011 Author Share #168 Posted June 14, 2011 Lindolfi,It is quite obvious from this that your image has more vignetting and also more colorshift, although the IFP exists in both images. So if we did everything in a similair way, it seems as if the sensors are not that equal. Hans Thanks Hans for the comparison. I'm not sure that there is such a big difference if not all steps are equal and only the M9 bodies are different. Everything needs to be standardized, before we can make a firm conclusion. For instance the fact that I have more vignetting may suggest that the conditions were not quite the same (size of the pinhole, light falling on the plastic cover etc). You already repeated making your own image since the first one was so noisy and got slightly less IFP simply by changing the evenness of the ambient light. This shows how important standardization is. Quote Link to post Share on other sites More sharing options...
otto.f Posted June 14, 2011 Share #169 Posted June 14, 2011 (edited) I would think that the size of the pinhole has as big consequences as shown here by Hans, because we know that aperture has and we can produce this amount of difference with an uncoded lenses (not) stopped down. The wider, the more IFP. I'm even beginning to think that the amount of light (EV) could have influence on an equally wide pinhole Edited June 14, 2011 by otto.f Quote Link to post Share on other sites More sharing options...
pop Posted June 14, 2011 Share #170 Posted June 14, 2011 ... On the other hand it may be that there is some correction code used by the RAW converter that is part of the DNG file that we are not aware of. The DNG speficications do mention a number of tags pertaining to the processing of the CFA. However, those specifications are a fair bit above my head, so I do not understand how to apply them. Apologies if that's already known. 1 Quote Link to post Share on other sites More sharing options...
Lindolfi Posted June 14, 2011 Author Share #171 Posted June 14, 2011 I'm even beginning to think that the amount of light (EV) could have influence on an equal wide pinhole This statement has been tested: there is no clear difference in IFP over an EV range of 5 stops. 1 Quote Link to post Share on other sites More sharing options...
sandymc Posted June 15, 2011 Share #172 Posted June 15, 2011 Thanks Sandy. Indeed I will look into the process of demosaicing and I'll do that because I found a difference in pattern in the image coming from the LightRoom RAW converter and my own reading of the three colour components from the raw part of the DNG file (see the image I posted here and compare with the images from the RAW converter elsewhere in this thread). I'll report later what I found. But I agree that there should be no difference. On the other hand it may be that there is some correction code used by the RAW converter that is part of the DNG file that we are not aware of. There will be a difference between the raw data and processed data - the red/green/blue in the DNG file is not the same red/green/blue as in a processed image. E.g., because the bayer filters are imperfect, "red" in the raw data is actually red plus some green and some blue. This is (approximately) reversed by the conversion "matrix". CornerFix completely corrects for Italian Flag only by analyzing raw data, regardless of raw processor, so I would say that the evidence is that raw processing has no practical effect. Sandy Quote Link to post Share on other sites More sharing options...
Lindolfi Posted June 15, 2011 Author Share #173 Posted June 15, 2011 Sandy, I just finished writing the bayer demosaic, gamma correction and whitebalance code in Matlab R2011a on the M9 uncompressed DNG file. More results later. Quote Link to post Share on other sites More sharing options...
pop Posted June 15, 2011 Share #174 Posted June 15, 2011 The new firmware is out: http://www.l-camera-forum.com/leica-forum/news/186818-leica-m9-firmware-1-162-released.html#post1735947 Improvements in the M9 Firmware Version 1.162 1. Camera Performance: Improvement of the sensor homogeneity for wide-angle lensesThere is an additional colour calibration for all wide-angle lenses, that minimizes chromatic deviations between image centre and image corner, especially in critical shooting situations. Quote Link to post Share on other sites More sharing options...
otto.f Posted June 15, 2011 Share #175 Posted June 15, 2011 The new firmware is out: http://www.l-camera-forum.com/leica-forum/news/186818-leica-m9-firmware-1-162-released.html#post1735947 Improvements in the M9 Firmware Version 1.162 1. Camera Performance: Improvement of the sensor homogeneity for wide-angle lensesThere is an additional colour calibration for all wide-angle lenses, that minimizes chromatic deviations between image centre and image corner, especially in critical shooting situations. Yes! that suggests even for the CV 15 . . Beats me! Quote Link to post Share on other sites More sharing options...
jaapv Posted June 15, 2011 Share #176 Posted June 15, 2011 I think you read too much into this. For third-party lenses don't expect anything. Whatever improvement there may be is a bonus. Quote Link to post Share on other sites More sharing options...
otto.f Posted June 15, 2011 Share #177 Posted June 15, 2011 I think you read too much into this. For third-party lenses don't expect anything. Whatever improvement there may be is a bonus. I wouldn't understand it anyhow Quote Link to post Share on other sites More sharing options...
jaapv Posted June 15, 2011 Share #178 Posted June 15, 2011 (edited) My CV 15 was not fully corrected by 1.160. Not by a long shot. I doubt that 1.162 will add much. Edited June 15, 2011 by jaapv Quote Link to post Share on other sites More sharing options...
t024484 Posted June 15, 2011 Share #179 Posted June 15, 2011 I will take a pinhole picture later today to see if anything has changed with the new software. Hans Quote Link to post Share on other sites More sharing options...
Lindolfi Posted June 15, 2011 Author Share #180 Posted June 15, 2011 (edited) All of a sudden this thread is about the new firmware, while all sensible tests in the quest for the explanation of the Italian Flag Phenomenon should be done without using the correction processing by the firmware (so with lens coding off). Anyway, it is of course interesting to see what the new firmware does, so I will eagerly run some tests. What I like most about the file "Improvements in the Firmware Version 1_1.62_en.pdf" is the sentence "Improvements in Italian translations" Now that of course is right on topic in this Italian Flag Phenomenon thread! Edited June 15, 2011 by Lindolfi 2 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.