alandamp said:
Well, I don't know if this will satisfy you, but I trusted this source:
http://www.hometheaterblog.com/hometheater/2006/04/mailbag_hd_movi.html
Go down to the section titled, "The resolution of the movies shown in theaters." This references a study by the International Telecommunications Union which has a link in that portion of the article.
Let us know if you have any issues with their conclusions.
Thank you! This is useful. I bypassed the author of the link, and reviewed the published research data directly. The author of the link incorrectly interprets the results of the test data. I reviewed the MTF plots, and for example, where the blog author claims that 2400 lp/ph are possible with the original negative, this is in fact not true in a practical sense. If you look at the graphs in the research, the MTF=6% at 2400 lp/ph. This means that the contrast between the black test chart lines and white background is 6% linear difference. Less then MTF=50 is quickly becoming difficult to perceive in actual photographic images that are not of solid black lines on solid white backgrounds. I'll be generous, and give them 30% as a lower limit of usefulness in real photographic scenes, because with proper post processing, you can raise the MTF of this lower limit to a perceptively higher value, to an extent. So, reviewing the 30 percent limit for the original negative, it extends to approximately 1500 lp/ph. I am not reviewing the end-result theatre prints, because they are much lower, and not relevant, since you would not make the HD scan from these. The HD scan would be made from the original negative, and the scanner will have higher resolution than the negative, so nearly all resolution should be retrievable with almost no loss. I am assuming the scanners used in the motion film industry, as well as the image processing steps, are at least on par with still photography, but this may not be the case due to available processing time for the high number of frames in motion pictures. The study does not address the factors of ISO sensitivity and the resulting grain. The grain structure further masks the resolution, and it is not a subject of this study. An additional study needs to be performed to evaluate this issue. Assuming the grain was not an issue (but it is), then 1500 lp/ph would exceed the resolution capability of the 1080P HD format. Note: Jedi2016 brings up other relevant factors. Perfect focus and depth of field are real issues, and are rarely going to be perfect when shooting moving objects.
Because of lack of data on the factor of noise(grain), I can not come to any solid conclusions, except that under theoretically perfect conditions(perfect focus, still object, perfect light), that 1080P would not be able to contain the resolution of a 35mm original film negative. When I look at HD sources of HD scanned film, even on just a 720P display, the grain is easily noticeable in areas of lower brightness, especially in broad same or similar color areas. This is why I must question the relevance of noise on this subject.
Note: Still photography 35mm film negatives have considerably higher resolution, at least with low ISO types. Low ISO is an option for still photography, as you have available bright flashes for lighting, and also, longer shutter times are possible. A motion picture has a maximum of 1/24 of a second available. Because even this would result in high motion blur, I suspect that considerably higher shutter speeds are used, requiring much higher ISO sensitivity film.
-Chris
