![]() I like the ability to get the statistics over the entire image or just in a small section of the image. The histograms use this range of values as well. I like that the pixel values go from 0-16384 reflecting the 14 bits per color channel in the raw file. I like the ability to see each individual color channel R,G,B,G2 and the min, max, avg and std dev for each channel as well as over and under exposure percentage by channel. I really like the EXIF window, which is phenomenally complete, right down to the focal point distance, DOF, LV, and shutter count. ![]() The GH2 is also unique (among all of the G-Series cameras released before and after the GH2) in having "Constant Prevue" functionality (in Manual shooting mode only), which makes the Live Histogram believable - without the irritating acrobatics using Preview/Display buttons after every half-press of the Shutter Button - and resulting in a Live Histogram which may (still) lack accuracy.I picked up Rawdigger research edition after reading this post to see if it can help improve my PP kills. 's test show that G3 playback RGBL histograms appear to have a 1.53 EV to 1.59 EV per-division scale-factor, I am quite sure from my testing that simliar playback RGBL histograms on GH2 (for the first division) have an around 0.5 EV per-division scale-factor (relative to RAW levels). I have previously reported about my GH2 preview Live Histogram system - which displays luminance image-data of some form, and auto-scales the 64 "dot" height Y-axis of the Live Histogram to (in some circumstances) be able to annunciate the occurance of as few as around 10 illuminated photo-sites at RAW ADU levels in the critical upper portions of the Live Histogram display (a much better indicator of RAW-channel clipping). May find with his G1 (or what other Panasonic G-Series users may find with their cameras). Thus, this problem is independent of the particualr WB mode utilized While the amounts of RAW-channel clipping shown in the (above) log histogram displays is not a very large amount, it would be enough to cause me to seek a modified approach of some sort.įurther, it is quite unclear to me how the use of AutoWB (or other WB modes) would serve to do anything whatsoever to solve the above-described problems with playback RGBL histograms. ![]() ![]() The extent (the degree in percentage of reading) of histogram data "truncation" is unclear, and I get the feeling that the G3 (anyway) playback RGBL histograms may be a somewhat lacking in quality indicator for "chimping" test-shots (whether one is using UniWB, AutoWB, or any other WB mode, for that matter), perhaps leaving the G3 user with only playback blinkies - the threshold of which likely assume some amount of Red/Blue channel boosting - as is not the case when using a UniWB configuration. It seems pretty clear that the G3 (and, I imagine your G1) playback RGBL histograms (with a X-axis per-division scale-factor of something in the range of 1.53 EV to 1.59 EV, in the case of the G3, anyway) do correspond rather well withĭata (where it comes to peaks of RAW-channel data), but appear to be rather lacking in their vertical Y-axis resolution - which is not a big surprise considering that they appear to be scaled so as not to "clip" any of the individual histograms, and they are made up by a fairly limited number of display-screen "dots". (in a manner consistent with your request for logarithmic Histograms in cases where they were not posted by I'm preparing a couple of entries for this experiment and I'm realizing that it's all about examining the top 1/2 EV in detail. You will see some fascinating stuff as you bracket exposures. Look at your RawDigger histograms using a This is what I meant when I wrote that the vertical axis does not have enough resolution to judge the clipping. It's like Splenda claiming that it has 0 calories per package, where as there's actually just under 0.5 calorie in each one: they are allowed to round off to the closest whole calorie number. ![]() Most likely the in-camera histogram will simply round them off to zero and not show at all. In the space between the right side of the peak and the edge of the histogram, where in-camera histogram showed no pixels at all, Raw Digger shows 50-100 pixels per brightness level, all the way to the edge. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |