p.12 #1 · Digital Medium Format / 35mm equiv difference
For Foveon fans, you should really take a look at what Contas does with using MF glass on Foveon. While I'm using MF glass on SLR/C (FF), he is using MF glass on an even smaller platform @ APS-C.
Do take notice of the varying lighting conditions as well.
There's more of his that can present the MF on Foveon look (although I think he needs more work on his lighting selection/WB). You might PM him so he can point you to even better representation than I was able to find some links to.
Personally, I'm hoping to see how Canon's work on their tiered Foveon style sensor turns out. I'm inclined to think it will be a breakthrough that will be a game changer at sensor response. This is one that could pull me away from the SLR/C if it comes to market in FF and anywhere near what a mortal can afford.
p.12 #2 · Digital Medium Format / 35mm equiv difference
Mescalamba wrote:
Um, not really.. Foveon in SD1M is 15 mpix true resolution, which is way harder than anything else on lens. That sensor is reason why Sigma made 35/1.4 and 18-35/1.8 .. those lens are made for that at first place (plus its used as test bench for any new Sigma lens).
Foveon isnt as far as resolution goes any different than Leica M-M. Per-pixel resolution, no interpolation of detail. Which among other things is why its so good for BW.
How is it harder on the lens? From what I gather a Foveon 14mp and Bayer 14mp is still 14 million pixels, the only difference is how these pixels are arranged. It's not like Foveon pixels are smaller than Bayer pixels putting stress on the lens to be able to resolve detail. No, pixels for the given sensor area, are same size. Foveon does not interpolate an image, resulting in crisper detailed image, because one pixel can receive red, blue, and green light. The way I understand this (correct me if I'm wrong), for Bayer to be able to get equivalent pixel quality at 100% viewing, requires that an image is captured at higher resolution and then down-sampled to match the Foveon resolution. Or if you take D800's 36mp sensor, which puts a lot of stress on the lenses with its small photo-sites, and compare that to, say a 24mp Foveon sensor, the Foveon should contain more detail even if up-sampled to 36mp.
p.12 #3 · Digital Medium Format / 35mm equiv difference
'Now, if would just work on his WB a bit more.'
It seems to go with the territory for that platform, unfortunately, which is a pity indeed. WB issues are almost non-existent with a broad range of acceptability on Sony cameras, to name just one brand.
p.12 #4 · Digital Medium Format / 35mm equiv difference
"broad range of acceptability" ... not quite sure I follow
Ambient lighting is so incredibly variable with regard to accurate color temp and/or mixed lighting. Combine that with the lack of a known reference neutral and it can require a bit of study to discern how to correct your color casts (which steal from your clarity/sharpness).
For me, getting my color balance set to optimal color contrast (i.e. neutral) first is key, prior to even considering how much sharpening/saturation/contrast/etc. to use. Frequently, I find that with some color correction, you can pull back on the saturation/contrast ... which tends to avoid that "baked" look. But, if you just pull back on the contrast/sat and still have some color cast ... dullsville.
Imo, unless you are using a known reference, "one click" WB isn't really very viable/reliable for getting proper WB in ambient (variable/mixed) ... and a deeper study of your light source/direction/color is pertinent to achieving it. Many times, ambient shooters leave their WB trust to camera/pp software and a cast in their images remains, which can deprive the image from showing all its glory.
Color correction theory is a pita that most folks don't bother to study/learn (still working on mine). That is the primary reason why I skip ACR and go straight to PS where I can get a more refined adjustment/control of the lighlights vs. mids vs. shadows @ "independent" correction, not only in the hue, but also in the value.
p.12 #5 · Digital Medium Format / 35mm equiv difference
Let's assume for a moment that perception of colour is individual or/and subject of manipulation, and as such relative.
How about transition perception?
p.12 #6 · Digital Medium Format / 35mm equiv difference
Transition is going to be rate of change in general ...
Consider (exaggerated) a scene going from yellow to neutral to blue vs. going from barely yellow to slightly blue to too blue because there is a blue cast.
The amount of change/contrast in hue perception will be different and the transition perception is different. Just take a neutral WB image and put a color filter over it ... the reduction in perceived contrast/rate of change is readily apparent. Attention to remove such cast is reciprocal in effect, even if subtle in some instances, obvious in others.
The rate of change / transition @ Hue, Saturation, Luminance all work together (.3 * .3 * .3) to yield our perception of contrast/change. If I can achieve sufficient contrast in my Hue, then I don't need to push my Saturation or Luminace contrast as much to achieve the level of contrast that I desire for the image. Then I can "bake" it less elsewhere in saturation/luminance, i.e. more natural. Conversely, that cinematic toning impacts our perception of transition by reducing it.
Putting your image where you want it to be is a puzzle. Ultimately, the pieces all work together. The contrast level of the light, the resolving/contrast level of the projected image (lens), the response of the sensor and the algorithm/PP.
If we take the PP and subdivide it into sub-components of H,S,L and apply our process methodology, then we are now looking at 6 steps (lighting/projected image/sensor response/H/S/L.
If ad nauseum hasn't already set in, we could take it further, but I think you get the gist.
The salient point is that subtle ^ Nth degree ... isn't quite so subtle any more. Are there differences in the MF glass ... absolutely (which is why I shoot it). Are there differences in a sensor without an AA filter ... absolutely (which is why I shoot it). Is there a difference in large pixels vs. small pixels @ response to the light/projected image or color profile from Kodak vs. Canon vs. Foveon vs. Sony vs. ... absolutely (which is why I shoot it). But individually, these don't make a lot of difference, it is their culmination that makes them yield a difference ... particularly when you apply those differences in conjunction with the subtle (or not so subtle) differences in the other areas (i.e. lighting/pp).
Even with those differences in play, they still are only part of the process encompassed by the projected image/sensor response. Without the other pieces receiving some refinement to accompany them, their aggregate difference is smaller than when accompanied by the others. Typically, those who put forth the $$$$$$$ to garner the diff @ MFD ... ALSO put forth the effort to garner the diff lighting/pp to accompany it ... and the extra attention to good WB is part of that as well. Blue rooftops vs. neutral rooftops, some may like the blue cast owing to that's the way the camera captured it. Personally, I find it stole something from the image ... but that's my opinion/style/preference (i.e. subjective), but it can give a clue @ the diff in perceived transition relative to cast vs. neutral.
Imo, we can effect the greatest amount of change to our images at the lighting and pp stages (if we are already using glass @ drawing style of our choosing). Those two stages have tremendous variability that could range from .1 to .9 (see Annie, etc.) , where the diff @ platform change will have a more narrow range of variation from .4 to .6. So as others have noted ... the platform change alone may not be an autonomous holy grail to achieving the "look".
Having noted the above ... I'm wondering if I should look into Capture One instead of Adobe (given my SLR/C sensor @ not Canon/Sony), hmmmm.
p.12 #7 · Digital Medium Format / 35mm equiv difference
Best colorwise is Kodak own software. Then LR/ACR. Then probably everything else..
Capture One has sometimes good profiles (Panasonic m4/3s have near perfect ones), sometimes horrible profiles (my Fuji S5). Its good software, but if you fight a lot with moiré, then getting latest RawTherapee and learning is better.. much better. Tho it reads Kodak files only as DNG (needs free to use DNG Adobe converter, compatibility setting around 5.3 ACR or so).
Otherwise, Capture One has 60-day trial with full functionality, so easy to try.
p.12 #8 · Digital Medium Format / 35mm equiv difference
+1 @ OEM Kodak software.
I've gotten away from using the Kodak software @ 32bit and dog slow on my older laptop. Looking forward to an all 64bit (PC) platform @ future laptop, I know I'm leaving something on the table ... just wondering if somebody plays well with Kodak color better than Adobe. As the rest of the PC world marches on ... the Kodak OEM software gathers another layer of compatibility issues, destined to be relic (oh wait, it already is).
I'm not for certain, but I think that the SLR/C sensor was made by the same mfr of some MFD sensors ... hence the wonder @ C1 or other MFD software to get just a little more of that "subtle" piece of the puzzle, that would play nice with 64 bit OS.
p.12 #9 · Digital Medium Format / 35mm equiv difference
One thing that really drives me up the wall is when software tries to manage my files for me.
I used original CaptureOnePro, but as soon as they updated it and it became a bit of Big Brother.
It can be nice to tether an MFD to, but it's still iffy...
p.12 #11 · Digital Medium Format / 35mm equiv difference
Not really... Regarding the manufacturing process, CCD is easier - and certainly less sensitive to manufacturing deviation and/or defects. It also often has a higher effective APS - Active Pixel Surface - since there's less metal covering the silicon photodiode. This does give a native insensitivity to angles, if implemented correctly.
But in the commercial sector, CCD hasn't really evolved as fast as the CMOS sensors. They're easily six or seven years after in evolution - which might tell you how good they MIGHT have been with better R&D budgets.
But regarding image quality, there not much between them now. CCD still has some advantages in overall smoothness at lower ISO, since you often get lower PRNU values. The CMOS principle in practice is basically that each pixel has its' own little charge>voltage converter, in CCD each column shares one single charge>voltage converter placed at the edge of the sensor. This gives better uniformity, but also higher charge losses. Slightly more noise, but more uniform / less "pattern" noise.
CMOS does in it self not warrant any real disadvantages. But the current trend towards video-optimizations, live-view, 120fps readouts to do contrast-detection focus and so on has really taken CMOS down a notch. A CMOS sensor built ONLY for still images, no fast readouts, no section-cropping readouts and no live-view could easily be half a stop better in ALL relevant performance metrics (than what we have now). Sensors like that exist in the military/surveillance/medical sector.
I'm often called over-conservative for saying it, but it's actually a quite simple truth. Video and Live-view hurts still image quality. Newer sensors have more transistors per cell, higher pixel structures, worse efficiency - than what they could have had.
I do solemnly believe that video and LV is necessary for sales, so I'm not protesting against manufacturers implementing it. It has saved a lot of companies a bit further by actively generating sales. But since i don't use it very often myself in a few of my cameras, I would be very happy to see a performance optimized model without all the extras. But I doubt the model would sell enough to survive to another generation of models.
p.12 #12 · Digital Medium Format / 35mm equiv difference
Not suprised, I noticed IQ going down in certain aspects since dawn of video/LV and most important since ISO war started. Latest Canon creations are living proof..
Would prefer color/DR/pure resolution wars instead..
p.12 #13 · Digital Medium Format / 35mm equiv difference
One of the reason I asked was a class action against Fujitsu.
The controllers for their hard disks were made by Cirrus, who in turn bought chips from a manufacturer unknown to me.
A large batch of the chips had a cheaper compound which actually accumulated humidity during contraction/expansion which in turn lead to hard disks failing nearly exactly 1 year of usage.
Made me think about production costs and volumes of CCD v CMOS and how cost of material might affect the product...
p.12 #16 · Digital Medium Format / 35mm equiv difference
Carsten: Cheers. I seem to fail to re-create this kind of stuff. And this is completely unprocessed shot (except slight contrast adjustment and sharpening/resizing).
Move to 5DM3 increased detail, but I rather dislike the colour.
p.12 #17 · Digital Medium Format / 35mm equiv difference
Interesting observation about the colour. I think, and this is probably very noncontroversial, that as other factors rise above certain thresholds (DR, resolution, ISO performance...), that colour is becoming more and more important again, in the minds of photographers. I am also not 100% happy with the colour of my D800. It is quite good in general, and specifically for landscapes and urbex, but skin tones are still a bit ruddy, for my taste. It would be great to have Fujifilm working with Nikon to get the old S3/S4/S5 colours (and highlight behaviour) into modern Nikon DSLRs, but even though these two companies no longer really compete much with each other, I guess we will never see this happen.
p.12 #18 · Digital Medium Format / 35mm equiv difference
theSuede wrote:
CMOS does in it self not warrant any real disadvantages. But the current trend towards video-optimizations, live-view, 120fps readouts to do contrast-detection focus and so on has really taken CMOS down a notch. A CMOS sensor built ONLY for still images, no fast readouts, no section-cropping readouts and no live-view could easily be half a stop better in ALL relevant performance metrics (than what we have now).
Another degree of subtle...
theSuede wrote:
I do solemnly believe that video and LV is necessary for sales, so I'm not protesting against manufacturers implementing it. It has saved a lot of companies a bit further by actively generating sales. But since i don't use it very often myself in a few of my cameras, I would be very happy to see a performance optimized model without all the extras. But I doubt the model would sell enough to survive to another generation of models.
Agreed, I much prefer a camera optimized for image quality, without the multiple subtle compromises that have crept in over the years. The market as a whole values a jack of all trades over a master of one.
p.12 #19 · Digital Medium Format / 35mm equiv difference
carstenw wrote:
It would be great to have Fujifilm working with Nikon to get the old S3/S4/S5 colours (and highlight behaviour) into modern Nikon DSLRs, but even though these two companies no longer really compete much with each other, I guess we will never see this happen.
... or a phoenix resurrection of Kodak color into DSLR @ Canon/Nikon/etc..
Funny how Fuji and Kodak (i.e. history of film making) seem to have a better understanding of color than consumer electronic gadget makers ... I wonder why that would be? Although with Sony's background, they would seem to be a candidate for making great color too.
While the advances/trends of DSLR was essentially the death of Kodak (at the hands of the drive for higher ISO, more speed vs. better color), MF may have embraced the teachings of Kodak a bit more.
I guess this is in part of why I consider my SLR/C to be a smaller version of MF. Maybe it's the other way around ... maybe MF is a larger (and now more technologically advanced) version of Kodak's SLR/C.
p.12 #20 · Digital Medium Format / 35mm equiv difference
RustyBug wrote:
I guess this is in part of why I consider my SLR/C to be a smaller version of MF. Maybe it's the other way around ... maybe MF is a larger (and now more technologically advanced) version of Kodak's SLR/C.
Same with the DMR with its Kodak sensor and all that.
