noise

Discussion in 'Nikon' started by suetye_photography, Nov 27, 2010.

  1. Hi, i have just purchased a Nikon d300s, an upgrade from the d80. I have not long had it so don't know my way around it yet but have notices that there is a lot more noise (artifacts to be precise, i'm told) in images from the d300s and wondered if i needed to adjust my settings. I checked out raw file settings on the camera and see that currently lossy is selected, am i right in thinking that this is a form of compression and that I'd do better to select non compression?
    Also I am getting a kind of ghosting on some images. I am in the habit of shoot my macros the same way, in low light, daylight, with long exposure. Always on a tripod but usually with a remote. I haven't figured that one out yet on the new camera. Would this make so much difference? Thanks, Sue
     
  2. Your sample picture had quite bad compression artifacts, but I do not know if that happened in the conversion to png or did it come out as such from the camera. Have you tried shooting raw? That one should be free of compression artifacts and you can decide in the post processing how much compression is tolerable for you.
     
  3. Set to RAW compression to lossless compressed. In my view, this is your best bet as it should not introduce any artifacts (and in my experience, it doesn't), without the rather excessive file size you get from uncompressed. The example you show, PNG is a compressed format too, so it's impossible to tell where the problem originates.
    As for the remote, assuming you had the Nikon ML-L3 (if I recall well) IR remote for the D80 - that one won't work on the D300. Remote control for the D300 will unfortunately typically cost a lot more.
    Could it be the difference - in my view, without seeing an example, yes. In macroshots, any camera movement is hyper-visible, and squeezing the shutter on a tripod mounted camera often still makes the camera vibrate a bit. Alternative for what you're doing would be to set the self-timer to 10 seconds, worth the shot at least.
     
  4. Hi Juhani, no the artifacts are there on the tiff which i exported from raw file in light-room @
    400 %. They can also be seen in lightroom on the raw file viewed at 4.1. I've cropped the image and saved as jpg to see if this shows any different for you. I understand jpg is also compressed but I can tell you they are as i am viewing them now. Actually not sure i'll get the opportunity to add an image again.
    00Xkqq-306125584.jpg
     
  5. Hi Wouter
    Thanks for your response. The camera is set to lossless compression.
    RE the remote, I can believe it, i know I had this problem with the d80 so always used a remote with this kind of work. My remote is a generic cheepo so I guess i'll be buying another one! Self timer, yes good alternative, thank you.
     
  6. Hi. Are you using mirror lockup for macros? This will almost certainly help. Also, try converting RAW files in either View NX or Capture NX. I personally find these make a better job of it than camera raw.
     
  7. Hi Simon
    No haven't tried that, thank you, I will. I guess the d300 mirror action is heavier than the d80. I'm using light room to convert but I could use the nikon software which i believe is good.
     
  8. Sue when you convert raw to tiff you already process a lot using settings in lightroom.
    It would be good to either list all settings or better make a raw file available.
    It is generally helpful to post an example but in this case it may not be sufficient and I may nor be the only one who needs a little help in pointing out what exactly the artifacts are (I have no clue what the image should look like without artifacts).
    Are the posted images 400% crops? I know it can help to increase magnification but just going from 100% to 400% will generate some artifacts. Whether these are the ones we are supposed to identify or not is a question in itself .-)
     
  9. I'll try a crop of the raw file. Thanks
     
  10. Hi Sue,
    The D300(s) are of factory set to Auto ISO ( mine was..), did you set iso to a fixed value before shooting ? Best performance, noise-wyse, is at iso 200 for the D300 cams, but when shutter times get above 15 sec's it is better to move to iso 400 or 800
    Also for longer times you can set noise compression on and active d-lighting to off..
     
  11. Hi thanks for your response. I will look at the iso settings. Noise compression on is good advise too. If you set iso to 400 or 800 for long exposure when you want clear macro shots won't that defeat the object and make the images soft? I used to find it okay with my d80. I've had some good tips here today though so will try them all. Many thanks.
     
  12. Hi thanks for your response. I will look at the iso settings. Noise compression on is good advise too. If you set iso to 400 or 800 for long exposure when you want clear macro shots won't that defeat the object and make the images soft? I used to find it okay with my d80. I've had some good tips here today though so will try them all. Many thanks.
     
  13. Hi thanks for your response. I will look at the iso settings. Noise compression on is good advise too. If you set iso to 400 or 800 for long exposure when you want clear macro shots won't that defeat the object and make the images soft? I used to find it okay with my d80. I've had some good tips here today though so will try them all. Many thanks.
     
  14. Hi thanks for your response. I will look at the iso settings. Noise compression on is good advise too. If you set iso to 400 or 800 for long exposure when you want clear macro shots won't that defeat the object and make the images soft? I used to find it okay with my d80. I've had some good tips here today though so will try them all. Many thanks.
     
  15. Sue, I also saw from your example that you have set your aperture to F32 ?
    This will cause diffraction on digital slr's, so maybe you want to try wider apertures too, this allows for shorter shuttertimes, and this again helps reduce noise.. ( just play with this, I find optimum aperture for 105mm at around f8 till f16 for makro...)
     
  16. CPM is right, anything smaller than f/11 will result in diffraction. this will be the case on a 10mp ccd D80 as well.
     
  17. Ok thank you. Perhaps i've been mistaken in setting the aperture as small as i can thinking that there will be more detail but maybe as you suggest this is not right for the lens. That said with the d80 I was satisfied with the outcome. I will play around with this and see what comes up.
    Just out of interest the guy in our local shop suggests that it may be to do with lightroom as this seems to convert raws to dng when it exports original. I sent him a copy of the raw also and he was expecting it to be an nef as is standard for Nikon he tells me.
     
  18. Okay Eric, thank you for your thoughts. I'll try that.
     
  19. Actually i've just checked out an image shot at 200 iso @ 2.8 on the 105 and the artifacts are the same. Jagged. I've also tried lossless compression and no compression and it is the same more or less.
     
  20. Sue, when you import image files into Lightroom, there are four options at the top of the Import screen.
    1. Copy as DNG - this is what you are doing but probably not what you want.
    2. Copy - This puts a copy of each raw file in another location.
    3. Move - This moves each raw file to a new location and removes them from the original location.
    4. Add - This leaves the files where they were without moving or copying them. This is the option that I use, because I have already copied my raw files to the location where I want to store them, and I have already renamed them.
    I would suggest that you review your files in View NX before looking at them in Lightroom. View NX will give you the most "native" view of the files, i.e. as the camera's settings intended.
    Regarding ghosting with macros, are you certain that your subject isn't moving, i.e. as when a flower is blown around by a breeze?
    When you post files here for others to review, JPEG would be a more suitable format. You can export them as JPEG files from View NX or Lightroom. Lightroom can also resize them. 600 or 700 pixels on the long side should be plenty big enough for online display.
    Make sure that the S, C, M switching on the front of the camera is set to S or C, not to M. Make sure that the lens is set to M/A rather than M if it has such a switch.
    If you can fix these problems first, we'll have a better chance of figuring out what else might be going wrong. Further, why don't you take some "easy" pictures first, such as the side of your house in daylight from ten feet away to determine if the camera is working properly? I can't make anything out in the macro examples that you have posted thus far.
     
  21. i dont seem to have a problem with high iso with my d300 with winter weddings i normally shoot above 1600 iso and sometimes 3200 & 6400 i have noise reduction set to high in camera and use topaz denoise or noise nin[​IMG][​IMG]ja to tweak
    00XkzI-306261584.jpg
     
  22. take to lightroom 3 problem solved
    good luck
     
  23. Dan, thank you so much, i believe this may have solved the problem. The nef's are noticeably sharper, however at 4.1 magnification in lightroom the images still seem noisier than i would have expected. I will do some more experiments today.
    RE ghosting, no i'm shooting still life indoors so no movement other than maybe a slight bounce from the mirror perhaps? I will try today with the mirror up and timer. The tripod is on a wooden floor so maybe that could be the culprit. Many thanks again. Sue
     
  24. Sue, WRT the remote release - Don't waste your money on the expensive Nikon version. I've been happily and successfully using both wired and wireless "Yung No" 3rd party releases for well over a year. These were bought from a well-known online source for about one-fifth the cost of Nikon's simple release cable for the pair of them. There is no way that the make of release can affect camera shake, unless you pull the cable while releasing it; but then you could do that with the genuine Nikon article as well.
    Personally, I think it's high time that Nikon stopped overpricing its accessories the way it does. A cabled remote release is a simple pair of switches in a cheap plastic case: Manufacturing cost about $3 US.
    Try lighting your macro shots with a small hand-held flash unit. This will eliminate any camera shake and give you a clearer idea of where the problem lies.
     
  25. Hi Martin,
    Yep your image is as clear as a bell. Is it me or is the d300s actually better at shooting in hight iso? I have taking some really nice images on high iso. The problem for me is probably a combination of things which i must get to grips with. Dan's advice was helpful so will go from here with it. Other advise about the optimum f stop for a 105 may well be helpful though right now i'm trying to do like for like experiments under the same circumstances as other images shot on the d80 which previously i have been pleased with.
    Thanks for writing, Sue
     
  26. Hi Lui, interesting that you say take to light room three. I noticed looking back on old files that light room imported them as nef's but only since i've had lightroom 2 have they imported as dng's. It's only now that i've made a connection. Worth the extra money then do you think? The images certainly look better since i imported them as nef's as Dan suggested.
    Thanks for writing. Sue
     
  27. Thanks Rodeo Joe, I'll check them out. You're right, mega expensive the Nikon stuff. My problem was that my old one for the d80 didn't appear to work with the d300s so will be buying a new one. Such an expensive business this photography.... it's all money out right now.
     
  28. Sue: I didn't see a response to your ISO vs. exposure duration question, so I'll answer it.
    Unlike film, where grain is inherent in the negative and is thus (relatively) constant - over and under exposures aside - on a digital camera noise is produced by heat on the sensor. Higher ISOs require more electricity and thus produce more heat, and a longer exposure produces heat longer. With digital cameras, longer exposures always produce more grain.
    As a rule of thumb, cameras like the D300s that have good high ISO performance should always have the ISO bumped up if it will keep you out of the 1 second or slower exposure range. The small amount of noise added by going to the next ISO is marginal compared to the larger amount of doubling your exposure time. The only time you DON'T want to boost your ISO (again, just speaking as a general rule of thumb here) is when it would take you into the boosted settings (those are the ones that say 'HI' instead of using a number), if if your exposure is going to be several seconds no matter what.
    Also, look at your tripod. With a bigger, heavier camera the strength of the tripod matters much more. I had a Manfrotto CX190X tripod - not a cheap tripod by any means. It's their 'standard' carbon fibre tripod, and costs about $350-$400. It's plenty strong enough for my DSLRs, but when I got a Hasselblad I learned that the lightweight tripod just doesn't have enough mass to anchor the heavy Hassy for long exposures. It would easily support either camera for a few second exposure or faster, but even using a cable release, mirror lockup, and hanging weight off the centre column, it just wasn't stable enough to give me crisp images on the Hasselblad over a minute, even though it did that just fine for my Nikons.
    If (and when) you need a new tripod, I recommend the Manfrotto stuff, or Promaster is really good too. Most of Promaster's high-end tripods like their carbon fibres (I have one of those too and love it) are literally just cheaper, rebranded Giottos, and those products are pretty good. I recommend their supercompact 5-section carbon fibre ($250-$300) for travel as it supports a D300s and a VERY heavy lens easily if you're not doing exposures over a minute. My only beef with it is that it does portraits terribly, since its not tall enough to reach eye level.
    If you want to do long exposures, I like Manfrotto's 55 series tripods. The plain old aluminum one weighs a good amount, but it will support (within reason) anything you want to put on it, at any height, for any length of exposure, and it's reasonably priced.
     
  29. Hi Zack
    The image I was looking at was shot at 200 iso f32 @ 30.0 secs. Thank you for this insight re grain etc. I was more concerned about artifacts than just grain, though I have said I felt the images coming from this d300s are more grainy/noisy in general which came as somewhat of a surprise. I have always shot my macros in the same way and can't say that I was ever disappointed with the results on my d80. I take your point that the longer the exposure the grainier the image, i didn't know that. As i said in an earlier post i was coming to the conclusion that the d300s operates better with higher iso as i've taken some quite pleasing shots in the higher iso bracket. Your post confirms this.
    I have purchased a manfrotto 055xprob if that makes any sense to you. I am pleased with it, worth the extra money. You are right in what you say though, re heavier camera. I did originally come home with a German make just under half the price but wasn't happy with it at all.
    I am posting here, as suggested earlier a jpg of one of these images in re the artifacts. Just to recap, I am convinced now that lightroom's conversion of the file from nef to dng had something to do with the results i was getting. But... if anyone has anything else to add having seen this image, i'll be grateful.
    Thank you for taking the trouble to write in . Sue
    00XlHR-306563584.jpg
     
  30. Hi Sue,
    I took the dng and opened it in Picasso viewer and thought that it was beautiful. Picasso allows magnification past 700%. I wonder why you want to go to 400%? I use a D3 and never go above 200% and then only looking for a small detail. I can not open the dng in Capture NX2 so I can not comment on it. I only shoot raw and seldom convert to jpg but those artifacts don't seem to be the jpg type of artifacts I have seen. I looked at the right side of the petals and there is a slightly lighter spot but that seems to be just more light on that spot. Why don't you post the nef raw file so we can see it.
     
  31. Hi James
    Thank you for looking. Why do I like to look at 400%, well I guess i just like to check that there is no shake, no ghosting and general clarity. What alerted me here was that I felt for such a huge outlay I should be seeing something of a higher quality than the d80 - i was concerned about this so compared the two. I agree it was on the right hand side of the image, in the darker area particularly I thought.
    Unfortunately I can't post the neff as light room converted it to a dng. Thanks to Dan i've altered that now. I will now start again, in better light, with my newly discovered knowledge and see what comes up. I'll post again if I feel it's still an issue. Thanks again. Sue
     

Share This Page