jeremy_clay wrote:
Or shoot thousands of frames a week with both and are attuned enough to their images and style to see the difference.
Since the primary context of this thread seems to be about which digital cameras exhibit a more film-like look, I guess the obvious question is, how many thousands of rolls of film have you shot, noobie?
pjbishop wrote:
RD, from Canon's White Paper on the 5D II, see below:
This is going to sound dumb but I'll ask anyway, if we put film under microscope, do they have similar texture in blocks(or grain) like those pixels from digital sensor? If "Organic" characteristic is related to pixel size, well...the 5D2 is 6.4µm, which is identical to 1Ds3, how come nobody complaint about the 1Ds3 being too digital and not as organic as the 5D?
I had the 1Dmk2n with Digi2 sensor ( 8.2µm) for awhile, identical sensor as the 5D, looking at all the photos I took with this camera in the past, none of them have the film-like characteristic by the way.... I'm using a calibrated 37" 1080P monitor powered by a GTX 285 video card, this setup should be decent enough to display the differences if there's any.....
-Just because a quality cannot be quantified, doesn't mean that it doesn't exist.
And just because it can be proven that a "difference" exists, it does not necessarily follow that a) the difference is significant or b) that one choice is the "best" one.
yauyi wrote:
This is going to sound dumb but I'll ask anyway, if we put film under microscope, do they have similar texture in blocks(or grain) like those pixels from digital sensor? If "Organic" characteristic is related to pixel size, well...the 5D2 is 6.4µm, which is identical to 1Ds3, how come nobody complaint about the 1Ds3 being too digital and not as organic as the 5D?
Basically, if you put film "under a microscope" you'll see various sizes and shapes of "particles" of varying density arranged in essentially random patterns and orientations. The result is, on the microscopic level, basically nothing like what you get from your DSLR. No current DSLR (5D, 5D2 or any other) produces a file that contains anything remotely like the film grain structure that holds the image data on film.
Now, I suppose one could take the position that a pseudo-random pattern of "grains" produced by a chemical interaction with light and then revealed through a post-processing workflow involving a bunch of scary chemicals is more or less organic than a process that roughly-speaking records the luminosity of various colors of light in small areas on an electronic sensor as binary data.
But, if so, it seems like this would shed no light (pun intended) on the question of which digital machine used to record light values at a sensor is more "organic" or "film-like" than the other. On this level, at least, it would be hard to argue that either is at all "film-like."
gdanmitchell wrote:
And just because it can be proven that a "difference" exists, it does not necessarily follow that a) the difference is significant or b) that one choice is the "best" one.
Dan
Indeed. You may note in my profile that for various reasons, I currently own a 5D2, despite the fact that I consider it to be one of the cameras producing more conspicuously "digital" looking files. How significant the difference may or may not be depends on the application and usage, as I said in my original post. I don't consider the difference to be meaningless, but nor do I consider it to overwhelm the other various other concerns that factor into the purchase and utilization of a camera for my business. Personally, I prefer the look of the files produced by my 1Ds3 to those from my 5D2 under most circumstances (excepting extremely high ISO), for precisely this reason (organic qualities). However, the 5D2 is definitely Canon's current high-ISO full-frame king, and thus has a place in my camera lineup.
RDkirk: I think that calling digital "film-like" is misleading in many cases. I would say that I prefer "less-conspicuously the result of digital processes." Personally, I've reached the conclusion that the best way to achieve film-like results is... shooting film.
RD, you might want to read the entire paragraph from the 5DII White Paper (following), which discusses the 'photo diode area' and 'pixel' size, and look at Canon's own diagram showing relative sizes of the pixel/photo diode area assembly for both cameras, which I included in my previous post. Yes, Canon did compensate a great deal for the smaller pixel size on the 5D II's sensor by re-engineering of the photo diode area, but I was speaking strictly of pixel size. (It's also been claimed that the individual sharpness of the 5D's individual pixels is very slightly superior to the II's.) Sayeth Canon in the White Paper:
"Compared to the original EOS 5D, the gap between the on-chip microlenses on the EOS 5D Mark II has been narrowed while the photodiode area ratio (photodiode area divided by the pixel size) has been increased. The distance between the microlens and photodiode has been further reduced to improve light gathering efficiency, and new color filter materials have been added to increase light transmission while retaining accurate color reproduction. These optical
advances and technology improvements to the sensor, plus countermeasures that help suppress noise from other camera circuits, are responsible for the sensor’s high signal to noise ratio,enabling higher ISO speeds and improved dynamic range at low ISOs.Those are impressive sensor improvements, considering that the original EOS 5D had larger 8.2μm pixels compared to 6.4μm pixels on the EOS 5D Mark II, and that pixel noise usually increases as size decreases."
RDKirk wrote:
Look up a few more of those white papers where Canon discusses in more detail that what we casually call a "pixel" is really what they call the "photo diode area"--the actual light-collection area of the total "pixel"--and that the "photo diode area" is the same size in both the 5D2 and the 5D.
EEGADS! Is this an example of how digital you can make digital look?
>>This is a crop from the 500D (said to have a digital look a while ago), interpolated to 500% with genuine fractals. Yes. Five hundred percent. And added grain of various sizes and types.
Well, silver is a *naturally* accuring substance and not biodegradable...
The distribution of film grains tend to form in natural patterns... like different colored sand grains on a beach etc... we respond to this as more natural.
I have long contended that for digital to be successful it had to be transparent, and still feel this way. If the digital process is transparent you can produce a beautiful image, when you are aware of the digital process it becomes unnatural and unpleasant (to me).
When you are aware of some film grain in some images the effect is pleasing (to me).
There is more to images than the sensor, they are all heavily processed and this adds its own character.
It has been shown that humans respond to frequency content that should be below human perception, we perceive thing at different levels other just resolving lines and dots as lines and dots.
RDKirk wrote:
I guess we can debate some other time about whether silver halide is particularly "natural" (if so, why is it not biodegradable?) and whether a digital sensor isn't actually a closer analog to the human retina than silver halide film (thus, actually more "organic"), and whether a grainless image isn't really more "natural" than a grainy image--given nobody actually sees "grain" in one's natural eyes.
But how would a lesser resolving sensor be "more organic" than a greater resolving sensor, since the smaller pixels would start approximating "reality?"
Well, if you alone have a full understanding of God and the Universe you are either crazy or a messiah... I am happy to accept that I do not... and cannot understand the universe... and will continue enjoying my moments spent here.
>>The link is just a fun, slightly related link, and I never tried to convince you to believe in God. There's no need to believe in God because the Universe and God are identical. Creation myths are exactly that; myths. You see, the word "God" is completely misunderstood by the vast majority of people, both Christians and Atheists alike. That's the source of the discord.
gdanmitchell wrote:
Yes. And everyone who knows, knows that God only shoots Nikon. And only full frame. Or maybe not. On the other hand, have you seen his gallery? I want to see side by side comparisons. On my monitor.
Since God's photography is only manifested through the myriad photographers of the Universe, it could be said that God does in fact shoot with every brand of lens on the planet, and every "brand" of lens used by any alien races on any of the billions of planets like ours in the Universe.
I still think there is no God... just the vast potential of the universe... and the universe could care less about photographs (or anything for that matter)... so yes, we aren't so special, but Leica is kind of special to me.
cogitech wrote:
Since God's photography is only manifested through the myriad photographers of the Universe, it could be said that God does in fact shoot with every brand of lens on the planet, and every "brand" of lens used by any alien races on any of the billions of planets like ours in the Universe.
I've been waiting for someone to bring up Kodachrome. No grain at all - yet a revered film - legendary... So lets just drop the whole grain thing - not applicable to this conversation.
cogitech wrote:
Regarding "the film thing"; it isn't that 5D images are more film-like, just that they are not so obviously digital as some other cameras are.
Darn. I really wish you lived near me. It would wonderful to watch how you distinguish prints made from photographs captured on the 5D and the 5D2. I make them, and I sure cannot see a consistent difference of significance corresponding to which looks more like film. (Which is, to repeat, not a claim that there are no differences between the cameras.)
(And, yes, I can already see the follow-up post: "Well, if you really knew how to print..." )
by the way.... I'm using a calibrated 37" 1080P monitor powered by a GTX 285 video card, this setup should be decent enough to display the differences if there's any.....
