Soft images from a D90, what could be the problem?

Discussion in 'Nikon' started by j_w|13, May 17, 2010.

  1. I have been having problems getting sharp images from my Nikon D90. Image after image comes out soft no matter what, whether I stop down or use a tripod does not make a difference. I reseted the custom settings. But didn't solve the problem. For the longest time I was blaming the lens. But after testing a variety of lens, it's definitely the camera.
    I on purposely darkened the rest of the image, while highlighting the blown up area. This image was shot with a D90 with a 24mm AIS manual focus lens. Lens set at infinity. Shutter speed 1/1000 s.
    [​IMG]
    This image shot with a D300 with the same 24mm AIS lens:
    [​IMG]
    Another example. This with a 70-300 mm autofocus lens on the D90:
    [​IMG]
    And on the D300:
    [​IMG]
     
  2. How is sharpening set on both cameras? Also, the exposures seem very different, and I don't know how much of it is that.
    On the first pair of images, btw, the close-up doesn't seem to match the lighting on the "full frame". Perhaps they are from different frames.
    Also, on the last two imates, I notice that the sharpening on the D90 image is set to "normal" and on the D300 it's "hard". That could have something to do with it. And... the exposure is a little different, too.
     
  3. How is sharpening set on both cameras? Also, the exposures seem very different, and I don't know how much of it is that.
    On the first pair of images, btw, the close-up doesn't seem to match the lighting on the "full frame". Perhaps they are from different frames.
    Also, on the last two imates, I notice that the sharpening on the D90 image is set to "normal" and on the D300 it's "hard". That could have something to do with it. And... the exposure is a little different, too.
     
  4. That doesn't look like blurry, it looks like less contrasty due to exposure or raw processing or active d-lighting or something.
     
  5. To tell you the truth, I didn't think much about the sharpening. I always leaves it on default. The default is fairly low (3rd bar from the left). I am going have to reshoot some of those images tomorrow at a higher sharpening.
    I on purposely darkened the picture post processing. I think I overdid it. Will get some more images tomorrow.
     
  6. did you test the 24mm on the D300s? why did you choose 1/1000s?
    how about 1/500s or 1/250s?
     
  7. Nikon takes the blame and/or the credit. The D90 and D300 have the same size sensor, but they are not identical in performance. You may never get the same results with each camera, same lighting, same lens...but you may try to do so. The same situation goes for the Nikon D3 and the Nikon D700 bodies...both are FX camera bodies, but the results are not equal.
     
  8. Check out the review on dpreview.com of the D90. Essentially the results of their testing showed it really didn't gain much resolution over the D80, and despite having the same sensor as the D300, it does not have the same internal processing, so it does not have identical image quality to the D300. A lot of people think the D90 is just a D300 with features stripped away, and this is not true. If the image quality of the D300 is what you are looking for you'll have to buy one to get it, plain and simple.
     
  9. @ Jerry, "The same situation goes for the Nikon D3 and the Nikon D700 bodies...both are FX camera bodies, but the results are not equal."
    never heard of that on any discussions here before. Doubt it very much.
     
  10. "Image after image comes out soft no matter what, whether I stop down or use a tripod does not make a difference"
    Have you tried applying a little extra sharpening?
    What program are you opening your images with? If you are using NX, in-camera settings are applied to your images and the default sharpening/contrast differences between the two could account for the differences in your images. If you use a program like Photoshop and shoot RAW, none of the in-camera settings are taken into account when the RAW files are opened and you can perhaps get more accurate comparison shots.
    It also appears you took the shots at vastly different times (as the contents of your crop show differences in the position of the garage door height and one of the window treatments), perhaps a different shooting location,, different lighting (the shadow areas vary) and a different camera angle. For testing purposes, you should try to keep everything equal.
    In any case, it appears the only real differences between the two could be 'corrected' with a bit of extra sharpening to your D90 shot.
    I find this site very reliable for comparison photos:

    http://www.imaging-resource.com/IMCOMP/COMPS01.HTM
    Checking the D300 vs the D90 there, it appears there are virtually no differences.
    Adding a bit of sharpening to your D90 shot (crop) in Photoshop seems to level the playing field per the corrected sample below. I don't believe your camera is defective is that is your thought. But if you have any doubts, why not send your camera in for a checkup.
    00WUDV-244989584.jpg
     
  11. "Cameras with same sensors, but with different performance... "
    As far as I know, the anti-aliasing filters are very important in current camera performance. There could be a noticeable difference here. I have read anywhere that it also highly affects camera costs.
     
  12. Just a guess, but i would say you have two different "picture control settings" the top looks like "standard" and the bottom image looks like "vivid" look at the yellow poles at the bottom of each pic, the d300 is more vibrant yellow, while the d90 is kinda dull.
    If you are using the same settings, then check your WB, maybe you could use a grey card to set both cameras, then try your comparison.
    Sharp or not sharp, there is no way the two cameras are that far apart in colour rendition.
     
  13. The D90 does indeed have sharpening set kinda low by default compared to other cameras.
    J W, after a little tweaking and re-testing, I'd love to know if you can get them closer. That said, the pixel-peeping differences that are shown here will not, I am guessing, make a huge difference in final printed output. We should also keep that thought in mind.
     
  14. Reading through the Magic Lantern D90 Guide, It recommends that if you are using the JPEG files to print directly, then you should slightly increase the in-camera sharpening. If you are shooting primarily in NEF format with intent to post-process, then they recommend reducing the in-camera sharpening to 0. This will give a soft shot in jpeg, but will allow for shapening to be done in post-process without the risk of doubling up.
    These settings can be adjusted in the Shooting Menu, under Picture Control, by modifying one of the existing settings, ie: Standard, Neutral ect. These have different sharpening settings by default, as well as contrast, brightness, saturation, and hue.
    Hope this helps you to get the best out of your camera,
    cheers :)
     
  15. Both the D90 and D300 were shot in their standard picture mode (not vivid). I have shot both Raw and Jpeg and it doesn't seem to make a difference in picture quality. I have used Capture NX, Micrografx (an old program), and Lightroom to view the pictures. They all looked the same to me. I will test out the D90 with higher sharpening and see what happens. I took thousands of pictures with the D90 on a trip a couple months ago. The softness is similar to what camera shake or misfocus would look like. But I know that is not the case.
     
  16. J W,
    I have the same issue right now in figuring out how best to set this camera (although I'm not comparing it to anything, just trying to figure out how best to set the sharpening).
    As far as picture modes, right now I'm using Vivid and sharpening just a tad more than vivid calls for. I'm very interested in finding out what you come up with, and may do my own tests here shortly. Obviously, the defaults on these two cameras produce wildly different images (okay, maybe not wildly, but clearly different), but I suspect that you will be able to get the two cameras to take very very similar photos, as the reviews I've read state clearly that this is possible.
     
  17. "...I took thousands of pictures with the D90 on a trip a couple months ago." And now for the rest of the story... Have you had (a guess?) your D90 in for a service check at Nikon? With thousands of exposures, the camera *in the consumer class* may be a bit worn out already. It does not have the shutter life cycles that a Nikon D700 or D3 has.
     
  18. Might be worth checking with Nikon if you can't figure out the soft image problem but the D90 is rated to 100,000 shutter cycles right from Nikon: "Durable, high precision shutter: Testing to over 100,000 cycles assures shutter life and accuracy." I can't see it being a wear issue. My D80 have 10,000 cycles when I gave it to my Dad and it still works flawlessly (except the matrix metering tending to overexposure in certain situations which is well documented.)
    I have my sharpness set higher than default with my D90 and still sharpen a little more in Photoshop.
     
  19. ShunCheung

    ShunCheung Administrator

    Your image from the D90 is underexposed. See how dark those buildings are? I would test again with more even exposure between the two images.
     
  20. j w-i would check some items. 1. make sure that EVERY SETTING in the d90 and d300 are set the same. there is an obvious coloration diffewrence between the test shots. that should be accounted for. 2. as a suggestion, and a big one, set the sharpening to slightly above the neutral or middle position. THEN, AND MOST IMPORTANT, plan and do some sharpening in pping. ALL dslrs are meant to have their images pped to some extent, and this includes sharpening. what i would try if you have pse or csx is to simply use auto sharpen in the image menus. this would adjust the sharpening to the standard amount that the auto setting is set for. and if one dslr takes more than the other then that dslr would get it. then do a comparison. 3. when shooting make sure the lens iso shutter speed and fstop are the same with no EC adjustment in either camera. 4. i personally would not expect 2 different models from the same maker to give identical results. even if the sensors are the same size. there is no guernetee that the incamera processing is the same. in fact, they very are not. if they were the same then what would be the sense of making 2 different models? the 2 dslr internally would be too close to end up with 2 legitemently different models.
     
  21. When I read the Dpreview review of the D90 it was mentioned that the D90 was softer than the D300.
     
  22. Stuart,
    Did it say that or did it say that the default setups in the D90 were softer?
     
  23. You'd really need to take the two raw files and put a few minutes into making them the same to judge if there are real differences. Like I said, I wouldn't have said one of these is sharper than the other unless you'd told me that was true, but the exposure isn't the same and the color and contrast treatment are different. You're just seeing a lot of speculation now on why a D90 would be less sharp when there's nothing here to suggest that's true. How fine can you split two these hairs when one of your knives is dull? Or something to that effect.
     
  24. Here is one page that mentions the difference between the D300 and D90
    http://www.dpreview.com/reviews/nikond90/page29.asp
    And here is the other page that suggests why there could be such differences.
    http://www.dpreview.com/reviews/nikond90/page33.asp
     
  25. The D90 has 16k actuations. Should not be anywhere close to worn out.
     
  26. ShunCheung

    ShunCheung Administrator

    J W, as long as your two images have such a big difference in exposure, the rest of the discussion is moot. I would suggest redo your testing so that both images have the same exposure, as close as you can get.
     
  27. Stuart, those dpreview comparisons are very helpful, however I wonder if the subtle differences they show will show up in any kind of print. Even if our OP's softness issues are different (exposure differences, setup differences, etc...), for us to think that a D90 can ever give exactly the same result as a D300? Nope. Can they produce functionally identical 8 x 10 or 11 x 14 prints? I bet.
    (Interesting: on the "High ISO" page of that review, the D90 actually looks better than the D300. Of course, the newer D300s gets even better. I love progress!)
     
  28. I don't doubt that that the D90 can give prints that look just as good as the D300 in many situations. When the only way to see the difference with static shots is to shoot identical images and then pixel peep them at 100% then the differences can't really be that great. The D300 will likely do better in situations that allow one to take advantage of the D300 AF system.
     
  29. Not sure if I'm missing something that has already been noted, but checking the EXIF data fort he first few images.
    (EDITED: sorry realised Peter had already pointed this out in his first reply.)
    The D300 shot with the 70-300 - had in-camera sharpening set to Hard (+2) and saturation set to +2.
    The comparative D90 shot - was Sharpening 0, Saturation 0.
    Hardly suprising the D300 looks more saturated and sharper.
    Martin
     
  30. This time both cameras had the same setting. All were taken on a tripod. Small images are clickable.

    D90, with the 100-300 lens, f/8 1/30s, JPEG small:

    [​IMG]
    [​IMG]
    D300 with the same setup:

    [​IMG]
    [​IMG]
    These were cropped from RAW images. D90 with 100-300 lens, 100mm 1/30s f/8:

    [​IMG]

    D300
    1/40s f/8:
    [​IMG]
     
  31. Those differences are pretty incidental I'll bet, in a final print.
    Why in the second set of two above is the exposure not exactly the same? Just curious.
     
  32. These from my 35mm AIS manual lens. Again all images taken on a tripod. Focus set at infinity. The EXIF data says 24mm for the D300, but I accidentally forgot to change the settings on the camera.
    D90, 1/40s f/8:
    [​IMG]
    D300:
    [​IMG]
     
  33. ShunCheung

    ShunCheung Administrator

    Concerning the images posted at 7:50pm, I think it is mainly a sharpening issue. The D300 image has more sharpening, including the noise. Pay attention to the building with a large flat wall right behind the SUV. The D300 version looks more noisy (the noise is sharper) and the overall image looks sharper.
     
  34. Perhaps the lens that you are using focuses slightly past infinity, causing the lack of sharpness.
    Try stopping down and use the depth of field scale on the lens to place your infinity mark. Give yourself some leeway. Say stop down to F16 & place infinity at F11.
     
  35. In the shots with the cars, and the little building on the bottom right with the slanted roof, I can see some difference in the white railing, between the two cameras. The latest pictures are much closer in quality than the original. Both have the same exposure. Other than that, the difference is quite small. I wonder how big the print would need to be for me to see that railing as I do in the examples ?
     
  36. If you look at these pictures from a distance, the differences become more noticeable though.
     
  37. In the last two they are pretty close. There are slight differences but even two D90 bodies would likely show very slight differences. The D300 and D90 are different cameras and may only share a very similar sensor.
     
  38. These images are the same, plus or minus a margin of error that can be accounted for in the raw processor. I'll say again, if you must split these hairs (I don't think there's much point here, but if you must) shoot raw (without d-lighting or any fancy stuff like that) and use your raw processor to try to make them the same.
     
  39. Let`s apply some philosophy here:
    Premises---
    1. Good Anti Aliasing filters are very expensive, highly affecting manufacturing costs.
    2. Software variations are relatively cheap, hence a minor manufacturing cost issue.
    Conclusions---
    1. Same sensor + elaborated AA filter = higher priced camera (=higher performance product, =higher end target).
    2. Same sensor + economic AA filter = "average" priced camera (=consumer performance, =consumer end camera).
    I bet Nikon don`t use the very same parts on such different products like the D300 and D90; probably, they use software to maximize performance.
     
  40. The cameras use the same sensor. D90 may actually give a slightly better image in some respects since the image processor in camera is a later development. In a very real sense you have the same camera twice.
    I do not think this is an equipment issue in the least. Both cameras have myriad settings, especially the D300 which allows four selectable, separate setups of virtually every setting on the camera to coexist, essentially four different identities of the camera. The D300 is softer default on sharpening, in jpeg settings, to allow greater freedom in PP.
    If it were me, I'd spend more time with more comprehensive courses/books in the vagaries of sharpening, contrast, color spaces, and processing, both in camera and in post processing. You might consider getting Thom Hogan's guide for the D300 and/or the D90. These are both a lot of camera, and they both demand a lot more of the photog compared to the consumer DSLRs in order to produce quality images.
     
  41. I recently started shooting with the D90 myself so Im relatively new to the system. My first batch of shots were indeed soft as well until I started playing with some settings. Im shooting handheld but pretty pleased with the sharpness I get perhaps 80% of the time. Firstly, I turned on the shutter release delay. I suspected some of my blurring was due to mirror shock when the mirror flipped. This helped, it delays the shutter for about 1 second until the shake from the mirror settles. I also of habit usually shoot on as low an ISO as possible (depending on situation). I personally prefer to as much as possible, avoid post-capture sharpening but thats just my preference. Also watch your aperture, shallow DoF can sacrifice some sharpness in various parts of your composition.
     
  42. dear to JW . in this case I had the same problem in CANON , one of my friend reffer me to do TEST CHART . let go to .
    http://FocusTestChart.com
    good luck & having fun .
     
  43. Jordan, with respect, there's a place for test charts, this ain't it. I can see, especially by the last two images that J W posted, that there is no issue with his camera. Shooting test charts to "solve this problem" is, again with respect, a waste of time. Again, I love test charts, they have their place.
    There is nothing "wrong" with either of his cameras, and I have to add, shooting digital without mastering Post-processing is simply setting up a system designed to fail and get sub-optimal results. All these issues can be solved by mastering camera settings (as J W is doing, evidenced by the much closer results in his most recent images than the first set) and post-processing, which some want to avoid, much to their detriment.
    We are so far beyond basic nit-picking here it's almost not funny. Both cameras are fine. Both images (in the last example) are great.
     
  44. Don't use such test charts. They simply cannot be relied upon. I note the author on the referenced site tries to defend his own test by basically claiming all the detractors are ignoramuses. Very convincing strategy, indeed.
    To ascertain whether two cameras really behave differently, rigorous testing is necessary and all spurious influences have to be removed. You need to work under stable light conditions, for example, midday sunshine on a clear day, or under heavily overcast skies. Intermittent cloud cover will cause a lot of trouble for later side-by-side comparisons and should be avoided.
    Use the cameras set to exactly the same settings, which should be: RAW, no tone curve adjustment, identical colour space, no in-camera shapening, fixed white balance, etc. Switch off AF and VR (on lens). Turn off any noise reduction. Use mirror lock-up and a reliable cable or remote release. Hopefully at least one of the bodies have Liveview (LV) and that feature should be used for critical focusing. Also, use a lens with its own tripod mount. Put the lens on a sturdy tripod and focus the lens using LV. Then, tape down the focus collar so it won't move, recheck the focus to ensure it still is perfect, and remove the camera. Take a full series with each camera, running though the entire aperture range. Allow at least 3 seconds betwwen each time the camera is activated. Never ever try to push the release button on the camera for tests like this.
    Process the RAW files in a good conversion program at default (zeroed) settings, then do A/B comparisons on *unsharpened* output files at 100%.
    Since the electronics and AA filters are unlikely to be identical, one should expect small differences between the two test series. A huge difference is unlikely considered the general high quality level now offered even on lower-priced gear.
     
  45. We dont see much difference anyway. Even though the exposure settings and infinity focusing are the same. The difference of the inaccuracies of the two cameras create the differences in exposure, white balance, and focusing
     
  46. sorry, I looked but cannot see where you give the working lens size in the first group of shots.
     
  47. SIGH. I could be a Smart a** and say this is a great argument for shooting film...you know whether the shot is on focus or not! No reason to blame the camera.
    But I will resist and say that:
    • I see very little difference between the 2 camera samples.
    • Any difference I detect could be attributed to the monitor I'm viewing on
    • You need to make some large (16 x 20 Inch) prints to really know if your camera or lens is malfunctioning
    • These differences are negligible and you should concentrate on making better photos by choice of subject and composition.
    Not to make light of your concerns. Nikon does occasionally screw up and churn out a bad camera. My original D70 had two nasty problems; backfocus and "Blinking GReen Light of Death syndrome". Nikon USA fixed both of these....and BGLOD was out of warranty repair which they did NOT charge me for.
    So bottom line is if you really think it's a bad camera send it off to Nikon with as much documentation as possible. If there is a problem they will fix it. If you're in the USA send it to the New York Repair facility, whether you're in the Western Region or not! New York is way faster and better than the LA LA repair facility.
     
  48. I'd like to send it off, but the cost to do this would be prohibitive. I am at least the second owner of the camera.
    I like the film days too. Focusing was never an issue with any film SLRs that I've owned.
     

Share This Page