Luminous Landscape Forum
Equipment & Techniques => Digital Cameras & Shooting Techniques => Topic started by: button on February 12, 2009, 06:56:26 pm
-
http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm (http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm)
Worth a look. What do you think?
-
Absolutely.
Eduardo
http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm (http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm)
Worth a look. What do you think?
-
http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm (http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm)
Worth a look. What do you think?
All the pictures are identical! You are just imagining that they look different. But they must be identical! Because, as is well known, "your camera doesn't matter."
-
Maybe I'm misunderstanding how he cropped, but the 5DII was cropped at 108% while the Nikon was 100%. Why not compare crops from 100% on both cameras? If he is increasing the magnification of the 5DII to 108% and then cropping, that's going to produce a crappier crop than one taken from an image at 100%.
As for ergonomics, I have only handled one Nikon which was a pro level, but I can't remember which one. I did like the feel of it much better than my 1DS3. It just felt more solid and easier to handle. It just felt better to me.
All of the other stuff I could care less about. Any camera you buy today from the 5D1 on up is going to give you super good results for most every application you need, just like Rockwell admits. I'm just less and less interested in quality of digital anymore. We're at a time now where things are more than good enough using mid to top level digital FF cameras.
-
no comments about the manner in which ken rockwell conducts his tests, but from a cost analysis perspective (the d3x retails 3.3x more than the 5d2) the improved quality makes logical sense.
Maybe I'm misunderstanding how he cropped, but the 5DII was cropped at 108% while the Nikon was 100%. Why not compare crops from 100% on both cameras? If he is increasing the magnification of the 5DII to 108% and then cropping, that's going to produce a crappier crop than one taken from an image at 100%.
As for ergonomics, I have only handled one Nikon which was a pro level, but I can't remember which one. I did like the feel of it much better than my 1DS3. It just felt more solid and easier to handle. It just felt better to me.
All of the other stuff I could care less about. Any camera you buy today from the 5D1 on up is going to give you super good results for most every application you need, just like Rockwell admits. I'm just less and less interested in quality of digital anymore. We're at a time now where things are more than good enough using mid to top level digital FF cameras.
-
Because, as is well known, "your camera doesn't matter."
And so often, neither does Ken <g>
-
http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm (http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm)
Worth a look. What do you think?
Gee thanks... you sent me to a review comparing top of the line slr cameras with this quote at the beginning of the review "The wild part is that if I go through the trouble of shooting CR2..." In other words if you want to buy a $3000 camera lug it out into the wilderness you sure shouldn't be shooting RAW because jpgs are easier
Thank you so much for posting the link
-
You mean that the JPEGs from the various cameras are NOT identical in terms of rendering ? Boy, you can learn something (stupid) from Ken every day.
-
I have a D3x. It spits out some pretty good files. There is a very good Raw==>Jpeg converter in the camera. It can do some magic like local shadow-recovery.
Edmund
-
Thank you so much for posting the link
And thank YOU for your response. I found the addition of the scanned Velvia 50 samples from the Leica/Pentax cameras the most interesting part of the test- I guess I should have mentioned that initially.
John
-
The simple truth:
All of the cameras will look identical when you print them out due to dithering, ink bleeding, resizing and scaling and so on during the print process. At least at say, under 8*12. Or under 2-3ft at normal painting type viewing distances.(affordable printing technology at this point lags tremendously behind DSLR pixel counts)
What I would look at is the color rendition and accuracy. Many of them look quite clean but washed out and/or soft. The Pentax 645 is actually closer to the real colors, though some of that can be compensated for in the DSLRs via Lightroom or other software that post-processes the raw files.
I don't print exhibit-sized prints. Nor do I print 3-4ft wide/tall giant posters and painting sized prints, either. My machine goes up to maybe 9*15 if I futz with it(8.5 full bleed and whatever length I can cram into it). But let's say you actually have a machine that can handle 10 inch wide rolls. And the largest aspect ratio for most cameras is 2:3, so figure 14-15 inches long per print. 300lpi(or equivalent to that on a dye sub printer)
10x300=3000
15x300=4500
Add both together and you get: 13.5MP More Pixels than this are literally thrown away or require either a magnifying glass or a much larger printer. And that's essentially 1:1 quality. Most people can't tell decently printed 100lpi versus 300 at more than 2-3 feet.
Unless you are a pro, forget about MP beyond about 12-16MP. It's well into the "enough already!" category, much like cars are today(can't remember when I actually needed to go 0-60 in 6 seconds...).
But let's say you did splurge on a printer that can do 16 inch wide rolls. The A900 is 6048 x 4032, or 252lpi to stretch to 16 inches. That gives us 24" wide to keep the aspect ratio correct. 25MP is a huge resulting print. Could you use a few more pixels to get the lpi count up? Sure - about 35MP would be perfect with a magnifying glass. But 250vs 300? Try it sometime. You'll be amazed at how close you have to get to tell a difference.(hint - it's in inches)
-
The simple truth:
All of the cameras will look identical when you print them out due to dithering, ink bleeding, resizing and scaling and so on during the print process. At least at say, under 8*12. Or under 2-3ft at normal painting type viewing distances.(affordable printing technology at this point lags tremendously behind DSLR pixel counts)
This other gentleman (not Ken) who has had a careful look at Raw converters seems to differ (http://naturewindows.com/articles/article090203.html)
To quote:
The in camera JPEGs from my 5DMKII, on the other hand, will likely stay in the realm of web work and very small prints. I'm quite disappointed in Canon's approach to low ISO JPEG processing and quite pleased with how much detail is available in the raw files. Preliminary results from the new Nikon D3x show that in camera JPEGs can do a lot more.
Edmund
-
http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm (http://kenrockwell.com/nikon/d3x/sharpness-comparison.htm)
Worth a look. What do you think?
hmmm ... Ken Rockwell? Should I click?
guess not. Think I'll pass this time.
-
This other gentleman (not Ken) who has had a careful look at Raw converters
Software converters(especially using JPEGs ) are also a problem.
The cameras themselves raw output is well beyond the average person's printing capability or need. Of course, this makes it less of a choice of the body and more about the lenses and software/processing. Kind of like how it was with film?
-
I have to say that my experience when meeting fellow photographers when travelling, is that most of them seem to be satisfied with their camera's jpeg output. Those of us who are obsessed with extracting the greatest highlight and shadow detail from our images, including myself, are in the minority.
-
Having only this week had dental surgery, stepped barefooted on broken glass, dislocated my big toe kicking a ball to my terrier Trix and had my stomach pumped after ignoring the "Discard After" date on a package of Muenster found in the back of my refrigerator, I'll pass on visiting Ken's House of Happiness.
Just this once.
-
Should I click?
guess not. Think I'll pass this time.
+1.
-
Ok let's put this differently: Back in the days of film, slides had nice colorful imagery on them. Some very high-end people did differential processing, clipping off pieces of film and testing them, but mostly photographers used the color and exposure as shot and were happy with that. On the other hand, negative film had wide lattitude, and was usually shot with a view to a choice of exposure and filtration at print time.
Now, Adobe wants us to believe that what comes out of a digital camera is necessarily unusable because it's like negative film, and we *need* to get Lightroom or Photoshop and twiddle sliders like crazy to get a decent print from a Raw. The camera makers on the other hand seem to realize that many users, incluidng pros, want or need immediately usable previews and immediately printable imagery, right out of the camera, like slide film: Raw here is just for saving or tuning the shot.
Of course, it's clear that Fuji, Nikon and Canon and Sony have the ability to process Raw files, in spite of what Adobe wuld like us to believe. In fact as the originators of camera technology FNCS have the ability to deliver results better than any one-size fits all solution. And with in-camera video coming, I believe they now have a greater incentive to deliver good renderings. Jpegs and extended Jpegs will improve soon.
Assessing Jpeg quality is a perfectly reasonable subjective test of a camera. I'm astonished no reviewer did the KenRockwell test before. But then, nobody in the press never really asked whther there were really "real" WMDs in Irak, an equally reasonable albeit subjective assessment, which had consequences considerably more far-reaching.
Edmund
-
Adobe wants us to believe that what comes out of a digital camera is necessarily unusable because it's like negative film, and we *need* to get Lightroom or Photoshop and twiddle sliders like crazy to get a decent print from a Raw.
It's worse than that. The truth is that Knoll, Schewe and the entire team of Adobe devs are well known as undercover agents for Apple. Have you ever seen Knoll and Steve Jobs in the same room at the same time? Of course not, because they're one in the same person. Laugh if you must, snicker if you will, but my own sister knows a guy who has a photo of Schewe before he glues on that beard in the morning, and he's clearly Steve Wozniak.
-
Now, Adobe wants us to believe that what comes out of a digital camera is necessarily unusable because it's like negative film, and we *need* to get Lightroom or Photoshop and twiddle sliders like crazy to get a decent print from a Raw.
Horseshit! Look, if you want to go down that Adobe rabbit hole again, please do it elsewhere. So ALL the other Raw processors, all the cameras that produce access to the Raw are doing so because Adobe brainwashed photographers?
Assessing Jpeg quality is a perfectly reasonable subjective test of a camera.
Yes it is, if you shoot JPEGs. I suspect the vast majority of users here are not, and its not because they are drinking some Adobe flavored Koolaid.
-
Horseshit! Look, if you want to go down that Adobe rabbit hole again, please do it elsewhere. So ALL the other Raw processors, all the cameras that produce access to the Raw are doing so because Adobe brainwashed photographers?
Yes it is, if you shoot JPEGs. I suspect the vast majority of users here are not, and its not because they are drinking some Adobe flavored Koolaid.
I fully agree with Ronald here. This need not be an either/or choice. Only JPEG or only RAW. When I shot medium format film-Fuji Provia and Astia and on occasion Velvia-I could plop it down a light table and the out of the box rendering in terms of color, tone and "look" was generally stunning. My objective in printing a chrome was to come as close as possible to the original. If I could achieve that, I was happy. It was hard, and often impossible, even with laborious masking, but at least I had a perfect target to shoot for. I now use a camera that costs over $30K, and start with with RAW files that simply look like crap out of the box compared to well exposed Fuji Provia. Until I start fiddling with white balance, curves, tone, etc., the RAW file does not compare. Yes, I can do more with the RAW ultimately than I ever could with the transparency, but it is neither quick nor easy to just get in the ballpark before I start to fine tune a file. This is crazy. If I could shoot a JPEG simultaneously with the RAW that was 90% of the way there, that would be very valuable. Some cameras allow this, so the quality of the JPEG a camera produces would be important to me. I would love to have a JPEG alongside the RAW that was the equal of a gorgeous chrome. Alternatively, I would love to have a RAW file that provided an out of the box, baseline rendering that was the equal of a gorgeous chrome, but that does not seem to be in the cards. This is what has always struck me as the big lie about how great the RAW workflow is. Can you imagine if Fuji and Kodak told the buyers of its film that they had to fiddle and diddle with the chemistry for each frame to get a really good result?
How long has it been since you shot slide film alongside your RAW files and compared the out of the box renderings? Try it. You may be depressed for a while. I am afraid to try it. [LOL]
-
I fully agree with Ronald here. This need not be an either/or choice. Only JPEG or only RAW. When I shot medium format film-Fuji Provia and Astia and on occasion Velvia-I could plop it down a light table and the out of the box rendering in terms of color, tone and "look" was generally stunning. My objective in printing a chrome was to come as close as possible to the original. If I could achieve that, I was happy. It was hard, and often impossible, even with laborious masking, but at least I had a perfect target to shoot for. I now use a camera that costs over $30K, and start with with RAW files that simply look like crap out of the box compared to well exposed Fuji Provia. Until I start fiddling with white balance, curves, tone, etc., the RAW file does not compare. Yes, I can do more with the RAW ultimately than I ever could with the transparency, but it is neither quick nor easy to just get in the ballpark before I start to fine tune a file. This is crazy. If I could shoot a JPEG simultaneously with the RAW that was 90% of the way there, that would be very valuable. Some cameras allow this, so the quality of the JPEG a camera produces would be important to me. I would love to have a JPEG alongside the RAW that was the equal of a gorgeous chrome. Alternatively, I would love to have a RAW file that provided an out of the box, baseline rendering that was the equal of a gorgeous chrome, but that does not seem to be in the cards. This is what has always struck me as the big lie about how great the RAW workflow is. Can you imagine if Fuji and Kodak told the buyers of its film that they had to fiddle and diddle with the chemistry for each frame to get a really good result?
How long has it been since you shot slide film alongside your RAW files and compared the out of the box renderings? Try it. You may be depressed for a while. I am afraid to try it. [LOL]
not to keep this discussion going any longer than it has to as I feel it's a mute point. However a couple of quick observations, First I never used the Manufacture's quoted ISO, did anyone? Second film was bought in large batches so as to know how one batch (of Velvia) responded to the next, (of Velvia). Everything was tested & or kept tract of. Third chrome film went to "reliable" labs because any alteration in the development would change the output, film tests were done, ( driving time to & from a lab to evaluate this) pushing or pulling the development to achieve different looks or characteristics. And let's not forget decades of trial & error went into the processing of film unlike PhotoShop version 3 to 8 in 1 decade. Even with the canon rebel the way the jpg is created can be altered to the user's personal taste not having to rely on the Manufacturers' default jpg settings
I will never be depressed not having to drive back & forth to the lab to check on a film test, to drop off film because they weren't open last night, to go back because your film hasn't dried yet, or to see that someone left a big fat fingerprint on my 4x5 transparency, but you are right looking at gelatin images that are illuminated from behind is very special
-
All of which is doubtless true. But at least, at the end of the shoot the film went to the lab and the photographer to the bar. Now the photographer heads to the computer room because *he* has to set the color and do the processing image by image.
Decent Jpegs or Tiffs may now arrive to change that. Nikon's D3x excellent in-camera Jpeg renderings bring me hope for fast automatic processing. A step in the right direction.
Edmund
not to keep this discussion going any longer than it has to as I feel it's a mute point. However a couple of quick observations, First I never used the Manufacture's quoted ISO, did anyone? Second film was bought in large batches so as to know how one batch (of Velvia) responded to the next, (of Velvia). Everything was tested & or kept tract of. Third chrome film went to "reliable" labs because any alteration in the development would change the output, film tests were done, ( driving time to & from a lab to evaluate this) pushing or pulling the development to achieve different looks or characteristics. And let's not forget decades of trial & error went into the processing of film unlike PhotoShop version 3 to 8 in 1 decade. Even with the canon rebel the way the jpg is created can be altered to the user's personal taste not having to rely on the Manufacturers' default jpg settings
I will never be depressed not having to drive back & forth to the lab to check on a film test, to drop off film because they weren't open last night, to go back because your film hasn't dried yet, or to see that someone left a big fat fingerprint on my 4x5 transparency, but you are right looking at gelatin images that are illuminated from behind is very special
-
People who frequent this site are a whole different group (mostly) interested in squeezing the last bit (!) out of every image, and I'm comfortable claiming that vast majority of mid- to top-end dSLR shooters shoot JPEG. But as Michael's own tests show, if seasoned pros can't tell a difference between a P&S and top-end MFDB in prints, I'd imagine it's the same story with RAW vs JPEG. Yet I still shoot RAW and am moving to a bigger camera...
In any case, it's a very interesting and valid comparison for JPEG shooters - regardless of what you think about the always controversial but never boring Ken.
Decent Jpegs or Tiffs may now arrive to change that. Nikon's D3x excellent in-camera Jpeg renderings bring me hope for fast automatic processing. A step in the right direction.
And the swiftness is exactly why many (most?) photojournalists and wedding photographers use in-camera JPEGs.
-
But as Michael's own tests show, if seasoned pros can't tell a difference between a P&S and top-end MFDB in prints, I'd imagine it's the same story with RAW vs JPEG. Yet I still shoot RAW and am moving to a bigger camera...
First off, I have nothing against JPEGs. I just don't shoot that format. I see no need in Raw+JPEG in my workflow. In fact, when I got the first camera that did this, I learned early on that those JPEGs were a waste of space on my card and drive. Why? Not because they suck. No one in their right mind would say that when an image is properly captured for JPEG. The camera manufacturers spend huge amounts of R&D, like the film cameras, trying to produce a rendering in camera most people will like. The reason the JPEGs didn't cut for me is, trying to get the Raws to match was just too much work. Show a client a JPEG, they expect that exactly from the Raw. Now the new camera profiles from Adobe are a step in the right direction. They attempt to mimic the "look" of the JPEGs.
So JPEGs are fine. Unless of course you don't like the canned rendering after which you're kind of screwed. There's only so much turd polishing one can conduct on an 8-bit JPEG (plus its slow and very damaging). The reason many of use shoot Raw is obvious. Its about control. Its not because the JPEGs would suck.
Again, IF you shoot JPEG, looking at the rendering quality of differing cameras is important. One has to question all the data, gamut, potential loss of DR from doing this, assuming you don't also shoot Raw.
Yes I agree, on the print, one would be hard pressed to tell the difference. The question has to be directed not at the viewer but at the image creator. 99 times out of 100, they prefer the JPEG to what they could accomplish from Raw? I really doubt the figures are anywhere near that level.
This is a bit like the old days when you had photographers that shoot film and also did their own printing. Many were too well paid to spend anytime doing anything but shooting. Can't argue with that. Not every photographer HAS to process their own Raws. Plenty have digital assistance to do most of the heavy lifting. Jay Maisel wouldn't know a Raw converter if it hit him on the head. He's not the least bit interested in processing images. He shoots RAW+JPEG and (I think to some degree, unfortunately) dismisses any JPEG he doesn't like and has his assistants process the Raws to match. He's not having to process hundreds let alone dozens of images at a time anyway. His staff can spend as much time as necessary messing with the Raws. He can go out and shoot. Others find rendering their Raws as important as shooting. There's no right or wrong answers here.
The idea however that Adobe, or anyone is brainwashing users into Raw is silly. It doesn't take very long to show most users the upside of Raw.
-
Hmm, Andrew is undoubtedly is well aware of the issues involved and is providing a cogent description of what the situation is like now.
The interesting novelty is that Nikon seem to be providing some really good canned renderings. A corollary is that if you can get the camera to give you good canned renderings, you can offload the printing quickly to a lab, even if you are a pro. I would expect quick photo batch printers (laser-printer like speed) to become a breakthrough product for serious amateurs and corner labs if the Jpeg renderings really work. At the moment batch printing is pretty useless as a feature for the dSLR user because the Raws are so dull looking, and work with the Raws takes time so the printer guys save money and make slow printers. Then Joe User doesn't print, because printing more than two images is a total pain.
I've been amazed at the good response I have been getting from my Jpeg postcards when I hand them out to people I've imaged.
Edmund
First off, I have nothing against JPEGs. I just don't shoot that format. I see no need in Raw+JPEG in my workflow. In fact, when I got the first camera that did this, I learned early on that those JPEGs were a waste of space on my card and drive. Why? Not because they suck. No one in their right mind would say that when an image is properly captured for JPEG. The camera manufacturers spend huge amounts of R&D, like the film cameras, trying to produce a rendering in camera most people will like. The reason the JPEGs didn't cut for me is, trying to get the Raws to match was just too much work. Show a client a JPEG, they expect that exactly from the Raw. Now the new camera profiles from Adobe are a step in the right direction. They attempt to mimic the "look" of the JPEGs.
So JPEGs are fine. Unless of course you don't like the canned rendering after which you're kind of screwed. There's only so much turd polishing one can conduct on an 8-bit JPEG (plus its slow and very damaging). The reason many of use shoot Raw is obvious. Its about control. Its not because the JPEGs would suck.
Again, IF you shoot JPEG, looking at the rendering quality of differing cameras is important. One has to question all the data, gamut, potential loss of DR from doing this, assuming you don't also shoot Raw.
Yes I agree, on the print, one would be hard pressed to tell the difference. The question has to be directed not at the viewer but at the image creator. 99 times out of 100, they prefer the JPEG to what they could accomplish from Raw? I really doubt the figures are anywhere near that level.
This is a bit like the old days when you had photographers that shoot film and also did their own printing. Many were too well paid to spend anytime doing anything but shooting. Can't argue with that. Not every photographer HAS to process their own Raws. Plenty have digital assistance to do most of the heavy lifting. Jay Maisel wouldn't know a Raw converter if it hit him on the head. He's not the least bit interested in processing images. He shoots RAW+JPEG and (I think to some degree, unfortunately) dismisses any JPEG he doesn't like and has his assistants process the Raws to match. He's not having to process hundreds let alone dozens of images at a time anyway. His staff can spend as much time as necessary messing with the Raws. He can go out and shoot. Others find rendering their Raws as important as shooting. There's no right or wrong answers here.
The idea however that Adobe, or anyone is brainwashing users into Raw is silly. It doesn't take very long to show most users the upside of Raw.
-
First off, I have nothing against JPEGs. I just don't shoot that format. I see no need in Raw+JPEG in my workflow. In fact, when I got the first camera that did this, I learned early on that those JPEGs were a waste of space on my card and drive. Why? Not because they suck. No one in their right mind would say that when an image is properly captured for JPEG.
Horseshit. They suck.
Okay okay here's the caveat.
I'm really only interested in printing, and looking at files from a 40D (shot raw+jpeg) printed A4 on an Epson multifunction, you could be half blind and still see the jpeg print had lost a lot of detail compared to the raw file rendered to a .tif with just basic editing. The colours were passable. Haven't tried a comparison on a 3800, but I see no reason it would be any different.
What would be really interesting is knowing if this is just the camera. Edmund, how much detail do you think is lost from the Nikon files with the in-camera rendering? At what point would it become noticeable? Cheers, David
-
First off, I have nothing against JPEGs. I just don't shoot that format. I see no need in Raw+JPEG in my workflow. In fact, when I got the first camera that did this, I learned early on that those JPEGs were a waste of space on my card and drive. Why? Not because they suck. No one in their right mind would say that when an image is properly captured for JPEG. The camera manufacturers spend huge amounts of R&D, like the film cameras, trying to produce a rendering in camera most people will like. The reason the JPEGs didn't cut for me is, trying to get the Raws to match was just too much work. Show a client a JPEG, they expect that exactly from the Raw. Now the new camera profiles from Adobe are a step in the right direction. They attempt to mimic the "look" of the JPEGs.
So JPEGs are fine. Unless of course you don't like the canned rendering after which you're kind of screwed. There's only so much turd polishing one can conduct on an 8-bit JPEG (plus its slow and very damaging). The reason many of use shoot Raw is obvious. Its about control. Its not because the JPEGs would suck.
Again, IF you shoot JPEG, looking at the rendering quality of differing cameras is important. One has to question all the data, gamut, potential loss of DR from doing this, assuming you don't also shoot Raw.
Yes I agree, on the print, one would be hard pressed to tell the difference. The question has to be directed not at the viewer but at the image creator. 99 times out of 100, they prefer the JPEG to what they could accomplish from Raw? I really doubt the figures are anywhere near that level.
You really do seem to always want to frame the argument as an either/or choice between JPEGs and RAWs. My point is that's a false choice. Why would anyone NOT want a beautifully rendered JPEG as good as a Fuji Astia/Provia/Velvia chrome(give me the choice of look) to go WITH a RAW fie, not INSTEAD of the RAW. Even better would be a Raw Converter that gives you an out-of-the-box rendering as good as a really great JPEG, but that does not seem to be in the cards.
-
You really do seem to always want to frame the argument as an either/or choice between JPEGs and RAWs. My point is that's a false choice. Why would anyone NOT want a beautifully rendered JPEG as good as a Fuji Astia/Provia/Velvia chrome(give me the choice of look) to go WITH a RAW fie, not INSTEAD of the RAW. Even better would be a Raw Converter that gives you an out-of-the-box rendering as good as a really great JPEG, but that does not seem to be in the cards.
It is an either or choice for many, and I explained why JPEGs are far less then desirable for many.
If you want to funnel high bit, wide gamut, full DR capable data into an 8-bit sRGB JPEG and throw away a heck of a lot of data, let alone accept a fixed rendering, by all means do so. Its your image and data.
I don't want a beautiful rendered 8-bit sRGB (and really worse, 8-bit Adobe RGB (1998)) image.
I don't know what out of the box means. Which rendering preset? The idea that some default Raw setting is ideal for all images is as silly as accepting that the last filter pack you had in your color enlarger is ideal for every color net you'll ever print with. Its as equally silly when reviewers talk about the quality of a Raw converter by comparing their default settings.
-
Horseshit. They suck.
Okay okay here's the caveat.
I'm really only interested in printing, and looking at files from a 40D (shot raw+jpeg) printed A4 on an Epson multifunction, you could be half blind and still see the jpeg print had lost a lot of detail compared to the raw file rendered to a .tif with just basic editing. The colours were passable. Haven't tried a comparison on a 3800, but I see no reason it would be any different.
What would be really interesting is knowing if this is just the camera. Edmund, how much detail do you think is lost from the Nikon files with the in-camera rendering? At what point would it become noticeable? Cheers, David
I agree that some testing would be necessary to see which cameras and which settings produce what results. After looking at the Ken Rockwell site, my feeling is that the high-end cameras can be set up to render all the detail to file, of course the sharpening *in the file* is then just capture sharpening and not print-specific sharpening. I don't know what sort of sharpening you'd get addiitonally if you drive the printer from the camera directly via PictBridge, or insert a card in the printer - on a computer print sharpening is *not* applied by default when printing from Photoshop.
Edmund
-
You really do seem to always want to frame the argument as an either/or choice between JPEGs and RAWs. My point is that's a false choice. Why would anyone NOT want a beautifully rendered JPEG as good as a Fuji Astia/Provia/Velvia chrome(give me the choice of look) to go WITH a RAW fie, not INSTEAD of the RAW. Even better would be a Raw Converter that gives you an out-of-the-box rendering as good as a really great JPEG, but that does not seem to be in the cards.
As I understand, the choice between jpeg and RAW is largely a matter of dynamic range potential. When DPreview test a camera's performance, the vast majority of images they use for comparison with similar cameras, are out-of-camera jpegs. However, when testing dynamic range by photographing a Stouffer Step Wedge, Dpreivew now provide results for the DR of both the jpeg as well as the RAW image. The difference is huge.
Check out the DR results for the D3 at http://www.dpreview.com/reviews/NikonD3/page20.asp (http://www.dpreview.com/reviews/NikonD3/page20.asp)
The D3 jpeg is described as having a usable DR of 8.6EV; the ACR converted RAW a full 12 stops, not with ACR defaults, but ACR 'auto' setting. To quote:
As usual the default Adobe Camera RAW conversion delivers less dynamic range than JPEG from the camera (a more contrasty tone curve and very little noise reduction in shadows). Simply switching to 'Auto' in the ACR conversion dialog reaps huge rewards (we measured the result to have exactly 12 stops of dynamic range), and in our tests with real world shots produced superb results with images that seemed to be over exposed beyond redemption.
This is the main reason for shooting RAW in preference to jpeg. If you don't care about DR. If 8.6EV is plenty for your purposes, then fine. Shoot jpeg.
-
Horseshit! Look, if you want to go down that Adobe rabbit hole again, please do it elsewhere. So ALL the other Raw processors, all the cameras that produce access to the Raw are doing so because Adobe brainwashed photographers?
Adobe is obviously not responsible for the inability of some cameras to output decent jpgs, but - speaking of conspiracies - it is pretty clear that Canon marketing has been doing a great job at polarizing the assessement of cameras along metrics they are/were good at, namely high ISO noise and sensor size.
Other circles have undergone less Canon influence, the photogrpahic community as a whole in Japan for instance, and jpg quality is a hot topic around here.
This doesn't mean that people don't shoot raw, but it means that the general consensus is that cameras should be outputting a usable file without the need for a computer based conversion first.
As I mentioned several times already, Nikon's goal in digital has always been to make the experience of a DSLR as seamless as possible even for people that are not tech savvy, this is the reason why they have invested money in a decent jpg engine, have including very advanced processing options in camera including a pretty decent D-Lighting capability. Fuji is following the same philosophy.
Why is Capture NX2 not free even with a 7000 US$ camera? Because they are many people inside Nikon who feel that raw shooting is an optional activity, not a mandatory one and their stance is that people who shoot jpg would complain if they had to pay for a software they don't need...
Cheers,
Bernard
-
I agree that some testing would be necessary to see which cameras and which settings produce what results. After looking at the Ken Rockwell site, my feeling is that the high-end cameras can be set up to render all the detail to file, of course the sharpening *in the file* is then just capture sharpening and not print-specific sharpening. I don't know what sort of sharpening you'd get addiitonally if you drive the printer from the camera directly via PictBridge, or insert a card in the printer - on a computer print sharpening is *not* applied by default when printing from Photoshop.
Edmund
I don't know how you'd set up a test that would compare different manufacturers, but it would be helpful if it could be done. I just had a look at the photos from when I was comparing a tif and jpeg. Both images were sent to the printer via Qimage, so the print sharpening was the same for both. But I can see on the jpeg file that you can't sharpen what isn't there. It's most noticeable in the shadows, but is apparent almost right across the board. It sounds from what you've been saying that Nikon have made considerable progress here. Or that the high end cameras will do a better job in the rendering. David
-
I don't know how you'd set up a test that would compare different manufacturers, but it would be helpful if it could be done. I just had a look at the photos from when I was comparing a tif and jpeg. Both images were sent to the printer via Qimage, so the print sharpening was the same for both. But I can see on the jpeg file that you can't sharpen what isn't there. It's most noticeable in the shadows, but is apparent almost right across the board. It sounds from what you've been saying that Nikon have made considerable progress here. Or that the high end cameras will do a better job in the rendering. David
I think we need to distinguish the encoding (8 bits and various color tricks), compression, and the actual image being encoded. Jpeg can be pretty good as an encoding goes, and 8 bits are ok if you're not going to move the curves around. However, where does the 8-bit Jpeg come from? From a rendering engine in the camera, which determines white balance, applies clipping, uses a tone curve, corrects for vignetting, applies Dlight correction etc. In other words from a renderer. If this renderer is ok you will get good imagery out of the camera as Jpeg files. This won't replace Raw if you need to move the curves around heavily or modify the color balance extensively, but could be more than good enough for immediate printing.
Edmund
-
Now although the out of camera JPEG rendering may look pretty good for most people, it surprising how flat it can suddenly look after you've tweaked the RAW file, even if just using a preset. Obviously you could tweak the JPEG to look better, but that's getting away from the point of them for most people. I think the main reasons some people think the JPEGs look so good is they simply look better than a flat undeveloped RAW file. And the correct camera profiles help with the JPEg colours. Adobe's calibration tweaks can be good, but not always.
One reason I also shoot JPEGs is that they process differently from RAW files and I don't necessarily care about say losing a smidgen of detail, as the look of the image is way more important than maximum theoretical perfection. And if I can get alook easier via a JPEG, why not do it.
-
One reason I also shoot JPEGs is that they process differently from RAW files and I don't necessarily care about say losing a smidgen of detail, as the look of the image is way more important than maximum theoretical perfection. And if I can get alook easier via a JPEG, why not do it.
This thought posits an excellent argument for shooting RAW+jpeg. If you take the time to get to know your camera's jpeg settings, you might happen upon a "look" that's not readily easy to create from scratch during RAW conversion. If you shoot both, then that jpeg might create a launching point for your vision. Who knows- maybe that particular shot's jpeg rendered a RAW conversion that you wouldn't have considered, whether it be contrast, tonal spread, or even something more mundane like sharpening. I can see having the RAW next to the jpeg in ACR, for example, as a sort of "soft proofing" for your creativity. I have shot strictly RAW up to this point, but I'm going to give RAW+jpeg a try.
John
-
I have been shooting JPG+Raw for a while with my Nikon. At one stage I just shot the Raws because if you have your JPG settings correct you can just process them through NX and get the same results as if it has been done with in-camera processing. It is a pity NX2 (or NX) is such a bad piece of software
The profiles in Adobe are a step in the right direction even if they still cannot generate the same kind of quality as NX.
I don't have to hand over my files immediately so I see no reason to shoot JPG. Running your raws through software to generate files that are identical to those that come out of you camera as JPG's directly doesn't cost you much time if you have setup your workflow properly.
Shooting Raw+JPG is a waste of card space if you know how to process your stuff and you don't need to hand over your images right after shooting them.
-
... The camera manufacturers spend huge amounts of R&D, like the film cameras, trying to produce a rendering in camera most people will like. The reason the JPEGs didn't cut for me is, trying to get the Raws to match was just too much work. ...
I’m shooting Raw + JPG, which I both consider to be orginals. Then, the decision is either to take this or that file for further usage. Often enough, I’d prefer to avoid this either/or decision.
From a conceptual point of view, it could still be interesting (at least for me) to have a Raw converter which uses the Raw file as the source to enhance the given JPG, rather than trying to do everything from scratch and on its own. For example but not limited, by blending in the missing highlights information. Comparable to Kodak’s ERI (http://www.kodak.com:80/global/en/professional/products/cameras/erit/erit.jhtml) technology (http://www.kodak.ca/global/plugins/acrobat/en/professional/products/cameras/erit/whitePaper_ecgadr.pdf?id=0.2.26.28.3.18.3.20.3&lc=en).
Peter
--