Many of you have read & weighed in on my recent posts concerning problems with soft images. Something posted by Troup Nightingale in answer to one of those post's got me thinking. I posted most of what follows in responce to troup, but It occured to me that it might be a subject worthy of a seperate post for your consideration.
This was Troups comment,..
...I have had images from my D200 that were soft for no apparent reason and the next image of the same subject taken within seconds was sharp. Most are sharp and I figured "I" was the problem on the soft ones...maybe not now!!
This is most of what I wrote in reply,..
...As for your comment about your D200. Honestly,.. before purchasing mine. I did a number of searches in the discussion & image forums using D200 to see what kind of quality, (or problems) the D200 had and to be frank,.. I remember seeing a fair number of softer images being posted from the D200 shooters. Images that (at least to my eye) just weren't as clean & crisp as a lot of the stuff posted from many Canon shooters. Especially in the avian forum. (please,.. Please,.. PLEASE,.. I'm not trying to start a "Canon, Nikon" debate) Lets not go there!!
Lots of Nikon images were tack sharp or I wouldn't have bought my D200. I decided to chalk those soft images I saw up to a probable difference in varying degrees of post processing expertise from the photographers.
BUT,.. what if that wasn't (isn't) the case? The more I think about this,.. the more it seems to beg the question,..
What if a lot,.. maybe quite a lot of digital shooters (Nikon or Canon alike) are keeping images that are actually "Too Soft"!!?? If so, is this "softness" resulting from some as yet unidentified in camera hardware/software defect. Maybe these images are saved & worked with because we've ALL read "RAW files all look pretty soft initially and you have to process them to get them too look right"!!???
That statement,.. "RAW's all look pretty soft initially" has never struck me as a precise enough description to count on in making my own dertminations for acceptible image sharpness in a digital capture. When working with a slides it was usually pretty simple. Either it WAS or WASN'T sharp. There wasn't a lot of wiggle room for error due to differences in individual taste or perception. Not so with digital & RAW apparently!!
In both of these recent posts of mine, as you can plainly read,.. Many folks have focused on the "Need" for me to recognize that RAW files "need to be processed". That various post processing proceedures need to be preformed on them before they look their "sharpest".
Now, I have always been aware of that "need", but it occurs to me that the question of "How sharp is sharp?," for a RAW file, in most cases has a pretty darned subjective answer to it!!
For example,..
Last year I sent a couple of sample NEF files to a "Nature Pro" acquaintance of mine, and he agreed that the one I thought was "sharp", was in fact sharp but needed addititonal post process sharpening. He agreed that the one I felt was too soft,.. Was too soft. No amount of sharpening was going to fix it. He thought that softness was a result of "camera shake. I didn't believe it was. I thought my shutter speed should have been sufficient at the time to eliminate that possability. Especially since I had other sharp images at slower speeds & vise versa.
My belabored point is that even when He & I were in agreement, we disagreed on cause. The determination of what caused the softness in that image would seem to me to be a much less subjective standard to reach agreement on. Yet we did disagree!! How can one objectively determine from a statement like "all RAW's look a little soft" what a "little soft" is??
This year I'm getting tack sharp images from the D70/200-400 combo at 600mm equivalent. Shooting as slow as 1/125th of a sec and unlike my D200, the D70 doesn't have a cable release. I'm shooting "hands on" the camera. I'm on softer, spongier ground in this swamp than I was last year. Camera shake SHOULD be more of a potential problem under my present shooting conditions,.. but ISN'T!! Also, as I mentioned,.. In the few images that aren't sharp, the reason is now obvious.
I can now tell right away when I inspect them what the trouble is. Camera shake, subject movement, or missed focus etc.!! Those problems are much more readily apparent, and occur only in those images where the camera settings are such that you would expect to see them. Again,.. not so with my D200 using manual or auto focus, (or in the past for that matter, using my D70 & manual focus lenses)
I don't know why the D70's working fine now except that I've only used it on auto focus with this lens. I'm not going to even try it on a manual focus lens until after I've sent the rental back,.. (Not going to take any chance of screwing it up until I'm done with the sandhill cranes.)
Again,.. My point is that most folks have tried to help me narrow down the problem asking about my technique, my equipment, my eyes, (all good and reasonable things to suspect BTW) but when my answers would seem to argue that those couldn't,.. shouldn't be the problem, we always came back to "RAW files looking a bit soft to begin with", and as I've stated what does "a little soft" look like??
Until I saw what was possible in terms of (consistant, reliable) sharpness from a digital capture, like I now seem to be getting from the D70 & 200-400 f4 lens. I had to accept that just like folks were telling me,.. I was probably being "Too Picky" on my initial impressions of an image's sharpness when inspected in PS/CS or Bridge.
I no longer believe that. I have a completely NEW, objective, side by side comparision of what is or isn't sharp in a RAW file. The difference in the quality of the sharpness of my D200's Pre & post re-set RAW files is unmistakable. I became absolutely certain after the re-set "fix" didn't stick & the camera again reverted to the Pre re-set level of softness in mid shoot. You can see the change from one batch of images to the other.
So,.. for your consideration, I ask,.. Is there some mysterious "Softness" defect lurking out there in camera's other than my own??!! Is there some more prevelant, commonly occuring camera/software problem that hasn't yet been confirmed or identified. Have others been struggling with this issue because theye have been convinced & accepted the premise, as I did,.. that maybe they're being "too picky" on judging their inital impressions of their RAW files from folks who's cameras/lenses/software are working properly ??
Is there any place out there where one can go to view & compare a completely objective standard of what these "a little soft but ultimately sharp" RAW files SHOULD look like?
Am I barking up the wrong tree or what??? (this whole maddening 3 year episode may have left me a bit on the paranoid side!!)
Craig, there is no real answer to your question because sharpness is not a quality of the raw file. Also, practically every camera has its own raw format.
The raw file is simply a collection of data read from the sensor. How this data is converted into an image is up to the raw converter software and there are practically infinite ways to render an image out of raw data. Each converter has its own defaults for a multitude of settings to give you a good starting point, but no two are exactly alike.
So, there really is no way to compare raw files, only converters. Some apply more sharpening by default, others don't.
I was afraid that might be the case. I'm sure you can understand my frustration.
Until I saw what I should/could be getting in the way of quality from a RAW file using a properly functioning camera/lens outfit, and without having one of you standing side by side with me as I was shooting to be sure my technique wasn't the source of the trouble. The subjective nature of "what is sharp?" Was driving me to fits of depression & despair.
That is the main reason you haven't seen me participating much in the forums. (that & time constraints with work & such) I would just get so fed up, frustrated, depressed & dejected after a shoot. I mean I don't have a lot of time or money (for gas & such) to go out & shoot only to come home with nothing, nothing I want to keep. Not because I didn't have some potentially nice compositions or images but because none of the good ones are sharp enough, they're all too soft to use. I got tired of wasting my time & effort. so I would quit. For a while anyway.
But, because I am practically obsessed with nature photography, (much to the detriment of my finacial health & well being I might add) I kept trying to see if "this time It'll be different". For the past three years it hasn't been.
Ultimately, I'm happy to now have a definitive standard of my own by which to judge my future digital images, I'm glad I stuck it out. I was just wondering if my particular trouble might be have been more common than it appeared to be!
If I could help someone avoid the kinds of agony & frustration I've experienced these past few years trying to solve this it was certainly worth putting out there to see anyway!!
I have to agree with Guy's statement. Keep in mind that camera settings can change the in-camera sharpening settings on a raw file. On my Canon, the default setting for in-camera sharpening when it is set on the "Landscape" setting is 4, when it is on "Normal" it is 3, and when it is on "Portrait" it is 2. So a file taken of the same subject, with the same shutter speed and aperture will look sharper with the landscape setting than the portrait setting. This is all out of the camera, before post-processing of course. You can of course change the default setting and eliminate all in-camera sharpening if you choose.
Scott - Are you sure those settings don't apply only to JPGs? I don't have an SLR, but I was under the impression that most won't do anything to the RAW file.
Anyway, I think "what is sharp" can best be answered by looking at some results in controlled conditions with your particular setup (particularly the lens). Then you have a frame of reference for your camera and that lens. It's possible that with certain lenses, at certain zooms and apertures, you might find all images soft, even if they are as sharp as the lens can produce.
David, the settings don't change the raw file, but the file is tagged with the camera settings which are then applied by default in the raw converter (unless explicitly overridden).
I am wondering if the difference in sharpness between the two cameras is mega pixel related. I would assume that the greater mega pixel D200 would show up sharpness issues well before the comparatively smaller mega pixels of the D70.
I don’t have an issue if my logic needs correcting.
Scott, I didn't know that about the different program modes applying different levels of in camera processing. I'll have to remember that in the future. In terms of my recent situation, I don't (& haven't) used any of the program modes. I shoot RAW, uncompressed, manual exposure & spot almost exclusively. On a few occasions I've changed my metering to "matrix" just to see how it does in some of the more tricky lighting conditions, but that's pretty rare for me.
Mike, I understand what you're getting at, but I would think that with an image file from each camera viewed side by side at 100%. Wouldn't the smaller megapxl D70 be the one most likely to be softer, all other variables being equal? Either way, as I mentioned, I'm not just comparing the D70 to the D200.
The D200's post re-set RAW images were just as sharp & crisp as the D70's. I should also mention that although the reset changed all the cameras settings to their factory defaults, before testing to see if the re-set worked. I went into the menu's and changed the settings to those I normally shoot with (BTW I normally shoot with in camera sharpening turned off completely, per prevailing conventional wisdom, on both camera's,.. and during this mess I've tried shooting with the sharpening set to normal, on both camera's to see if it made a difference. It didn't!) After changing the settings to my normal working ones, the D200 worked fine,.. Nice crisp, sharp RAW images,.. temporarilly! I shouldn't have counted on the "fix" sticking.
Next day, I ended up with more than four gigs of utterly unusable images of an adorable gaggle of goslings peaking out from under mom's breast at the nest. (OHHH!!!!,.. I Nearly blew a gasket on that one!!!)
After that I did another reset I went to the field and shot two, 2gig Sandisk cards of test images. Without changing anything other than possibly the shutter speed or aperture between cards, the first post re-set CF card had sharp images and the second one didn't. Only thing I did was turn off the camera, switch the cards, and turn it back on to start shooting. No more than 20 seconds elapsed time, and complete, softly rendered crap on the second card!!
It will be interesting to see what, if anything other than the camera Nikon will try to blame this on when I send it in for inspection & repair.
Steve, I checked out the site you referred to. On the larger examples they use to show the difference in sharpening,.. The unsharpened RAW image of the "Old scruffy looking Guy" looks quite a bit sharper than the unsharpened, problem RAW's I've been talking about. I'm reveiwing them in bridge, using the loupe mode on the preview, and I believe that I've got ACR's default setting of adding 25% sharpening. I'm not sure if that setting applies sharpening to viewing them in the loupe mode or just when their opened in ACR & CS3, But they look lousy in either!
Discussion
8 repliesThis was Troups comment,..
This is most of what I wrote in reply,..
...As for your comment about your D200. Honestly,.. before purchasing mine. I did a number of searches in the discussion & image forums using D200 to see what kind of quality, (or problems) the D200 had and to be frank,.. I remember seeing a fair number of softer images being posted from the D200 shooters. Images that (at least to my eye) just weren't as clean & crisp as a lot of the stuff posted from many Canon shooters. Especially in the avian forum. (please,.. Please,.. PLEASE,.. I'm not trying to start a "Canon, Nikon" debate) Lets not go there!!
Lots of Nikon images were tack sharp or I wouldn't have bought my D200. I decided to chalk those soft images I saw up to a probable difference in varying degrees of post processing expertise from the photographers.
BUT,.. what if that wasn't (isn't) the case? The more I think about this,.. the more it seems to beg the question,..
What if a lot,.. maybe quite a lot of digital shooters (Nikon or Canon alike) are keeping images that are actually "Too Soft"!!?? If so, is this "softness" resulting from some as yet unidentified in camera hardware/software defect. Maybe these images are saved & worked with because we've ALL read "RAW files all look pretty soft initially and you have to process them to get them too look right"!!???
That statement,.. "RAW's all look pretty soft initially" has never struck me as a precise enough description to count on in making my own dertminations for acceptible image sharpness in a digital capture. When working with a slides it was usually pretty simple. Either it WAS or WASN'T sharp. There wasn't a lot of wiggle room for error due to differences in individual taste or perception. Not so with digital & RAW apparently!!
In both of these recent posts of mine, as you can plainly read,.. Many folks have focused on the "Need" for me to recognize that RAW files "need to be processed". That various post processing proceedures need to be preformed on them before they look their "sharpest".
Now, I have always been aware of that "need", but it occurs to me that the question of "How sharp is sharp?," for a RAW file, in most cases has a pretty darned subjective answer to it!!
For example,..
Last year I sent a couple of sample NEF files to a "Nature Pro" acquaintance of mine, and he agreed that the one I thought was "sharp", was in fact sharp but needed addititonal post process sharpening. He agreed that the one I felt was too soft,.. Was too soft. No amount of sharpening was going to fix it. He thought that softness was a result of "camera shake. I didn't believe it was. I thought my shutter speed should have been sufficient at the time to eliminate that possability. Especially since I had other sharp images at slower speeds & vise versa.
My belabored point is that even when He & I were in agreement, we disagreed on cause. The determination of what caused the softness in that image would seem to me to be a much less subjective standard to reach agreement on. Yet we did disagree!! How can one objectively determine from a statement like "all RAW's look a little soft" what a "little soft" is??
This year I'm getting tack sharp images from the D70/200-400 combo at 600mm equivalent. Shooting as slow as 1/125th of a sec and unlike my D200, the D70 doesn't have a cable release. I'm shooting "hands on" the camera. I'm on softer, spongier ground in this swamp than I was last year. Camera shake SHOULD be more of a potential problem under my present shooting conditions,.. but ISN'T!! Also, as I mentioned,.. In the few images that aren't sharp, the reason is now obvious.
I can now tell right away when I inspect them what the trouble is. Camera shake, subject movement, or missed focus etc.!! Those problems are much more readily apparent, and occur only in those images where the camera settings are such that you would expect to see them. Again,.. not so with my D200 using manual or auto focus, (or in the past for that matter, using my D70 & manual focus lenses)
I don't know why the D70's working fine now except that I've only used it on auto focus with this lens. I'm not going to even try it on a manual focus lens until after I've sent the rental back,.. (Not going to take any chance of screwing it up until I'm done with the sandhill cranes.)
Again,.. My point is that most folks have tried to help me narrow down the problem asking about my technique, my equipment, my eyes, (all good and reasonable things to suspect BTW) but when my answers would seem to argue that those couldn't,.. shouldn't be the problem, we always came back to "RAW files looking a bit soft to begin with", and as I've stated what does "a little soft" look like??
Until I saw what was possible in terms of (consistant, reliable) sharpness from a digital capture, like I now seem to be getting from the D70 & 200-400 f4 lens. I had to accept that just like folks were telling me,.. I was probably being "Too Picky" on my initial impressions of an image's sharpness when inspected in PS/CS or Bridge.
I no longer believe that. I have a completely NEW, objective, side by side comparision of what is or isn't sharp in a RAW file. The difference in the quality of the sharpness of my D200's Pre & post re-set RAW files is unmistakable. I became absolutely certain after the re-set "fix" didn't stick & the camera again reverted to the Pre re-set level of softness in mid shoot. You can see the change from one batch of images to the other.
So,.. for your consideration, I ask,.. Is there some mysterious "Softness" defect lurking out there in camera's other than my own??!! Is there some more prevelant, commonly occuring camera/software problem that hasn't yet been confirmed or identified. Have others been struggling with this issue because theye have been convinced & accepted the premise, as I did,.. that maybe they're being "too picky" on judging their inital impressions of their RAW files from folks who's cameras/lenses/software are working properly ??
Is there any place out there where one can go to view & compare a completely objective standard of what these "a little soft but ultimately sharp" RAW files SHOULD look like?
Am I barking up the wrong tree or what??? (this whole maddening 3 year episode may have left me a bit on the paranoid side!!)
The raw file is simply a collection of data read from the sensor. How this data is converted into an image is up to the raw converter software and there are practically infinite ways to render an image out of raw data. Each converter has its own defaults for a multitude of settings to give you a good starting point, but no two are exactly alike.
So, there really is no way to compare raw files, only converters. Some apply more sharpening by default, others don't.
Guy
I was afraid that might be the case. I'm sure you can understand my frustration.
Until I saw what I should/could be getting in the way of quality from a RAW file using a properly functioning camera/lens outfit, and without having one of you standing side by side with me as I was shooting to be sure my technique wasn't the source of the trouble. The subjective nature of "what is sharp?" Was driving me to fits of depression & despair.
That is the main reason you haven't seen me participating much in the forums. (that & time constraints with work & such) I would just get so fed up, frustrated, depressed & dejected after a shoot. I mean I don't have a lot of time or money (for gas & such) to go out & shoot only to come home with nothing, nothing I want to keep. Not because I didn't have some potentially nice compositions or images but because none of the good ones are sharp enough, they're all too soft to use. I got tired of wasting my time & effort. so I would quit. For a while anyway.
But, because I am practically obsessed with nature photography, (much to the detriment of my finacial health & well being I might add) I kept trying to see if "this time It'll be different". For the past three years it hasn't been.
Ultimately, I'm happy to now have a definitive standard of my own by which to judge my future digital images, I'm glad I stuck it out. I was just wondering if my particular trouble might be have been more common than it appeared to be!
If I could help someone avoid the kinds of agony & frustration I've experienced these past few years trying to solve this it was certainly worth putting out there to see anyway!!
Craig
Anyway, I think "what is sharp" can best be answered by looking at some results in controlled conditions with your particular setup (particularly the lens). Then you have a frame of reference for your camera and that lens. It's possible that with certain lenses, at certain zooms and apertures, you might find all images soft, even if they are as sharp as the lens can produce.
Guy
http://www.cambridgeincolour.com/tutorials/RAW-file-format.htm
The D200 does have a selection of sharpening options, which (as best I can tell) work in the manner Guy stated above.
I am wondering if the difference in sharpness between the two cameras is mega pixel related. I would assume that the greater mega pixel D200 would show up sharpness issues well before the comparatively smaller mega pixels of the D70.
I don’t have an issue if my logic needs correcting.
Thank you
Mike
Scott, I didn't know that about the different program modes applying different levels of in camera processing. I'll have to remember that in the future. In terms of my recent situation, I don't (& haven't) used any of the program modes. I shoot RAW, uncompressed, manual exposure & spot almost exclusively. On a few occasions I've changed my metering to "matrix" just to see how it does in some of the more tricky lighting conditions, but that's pretty rare for me.
Mike, I understand what you're getting at, but I would think that with an image file from each camera viewed side by side at 100%. Wouldn't the smaller megapxl D70 be the one most likely to be softer, all other variables being equal? Either way, as I mentioned, I'm not just comparing the D70 to the D200.
The D200's post re-set RAW images were just as sharp & crisp as the D70's. I should also mention that although the reset changed all the cameras settings to their factory defaults, before testing to see if the re-set worked. I went into the menu's and changed the settings to those I normally shoot with (BTW I normally shoot with in camera sharpening turned off completely, per prevailing conventional wisdom, on both camera's,.. and during this mess I've tried shooting with the sharpening set to normal, on both camera's to see if it made a difference. It didn't!) After changing the settings to my normal working ones, the D200 worked fine,.. Nice crisp, sharp RAW images,.. temporarilly! I shouldn't have counted on the "fix" sticking.
Next day, I ended up with more than four gigs of utterly unusable images of an adorable gaggle of goslings peaking out from under mom's breast at the nest. (OHHH!!!!,.. I Nearly blew a gasket on that one!!!)
After that I did another reset I went to the field and shot two, 2gig Sandisk cards of test images. Without changing anything other than possibly the shutter speed or aperture between cards, the first post re-set CF card had sharp images and the second one didn't. Only thing I did was turn off the camera, switch the cards, and turn it back on to start shooting. No more than 20 seconds elapsed time, and complete, softly rendered crap on the second card!!
It will be interesting to see what, if anything other than the camera Nikon will try to blame this on when I send it in for inspection & repair.
Steve, I checked out the site you referred to. On the larger examples they use to show the difference in sharpening,.. The unsharpened RAW image of the "Old scruffy looking Guy" looks quite a bit sharper than the unsharpened, problem RAW's I've been talking about. I'm reveiwing them in bridge, using the loupe mode on the preview, and I believe that I've got ACR's default setting of adding 25% sharpening. I'm not sure if that setting applies sharpening to viewing them in the loupe mode or just when their opened in ACR & CS3, But they look lousy in either!
Thanks for the interest & help everyone!