Specks on photos


Sgdevilzz

Senior Member
May 16, 2010
1,623
1
38
Good day fellow Csers. Need a little help here! Recently i noticed my photos has these dots/lines. It's not dust, more like something just poked the sensor, leaving a mark. I can only see the dots when zoomed in 100% and slightly on 50%. I'm not a pixel peeper but it gets really frustrating when these things happens. Any possible explanation for this?

6774732404_8be66ff1b0_b.jpg

[Normal print screen capture, just a reference]

6920848275_95a1972f5d_b.jpg

[zoomed 25%]

6774733146_0463eacfe6_b.jpg

[zoomed 100%]

6774733620_15a06de456_b.jpg

[Different photo, same spot]

6920849325_793c0ce0aa_b.jpg

[I took a photo of the playback screen of my camera, seems clearer here, spots are obvious when zoomed in 100%. Ignore the dirt on the top left, it's just the screen]


I appreciate all the help given. (P.S i went to NSC to clean the censor and they said there are minor scratches on the sensor, is this the mark they are talking about?)
 

Last edited:
Deathlines

perhaps that is what your camera can see?

heh.

jokes aside, if the lines are in the same area, it might be due to the scratches i guess
 

While it could be dead pixels, it seems unlikely if there is a halo around the pixels. Could you share a RAW image somewhere online for us to download for better assessment ? I don't quite understand how a scratch will present itself as a dot on the image either. Do you have another lens to try with and also gives the same result ?

Should the result is really due to scratches in front of your sensor or on the AA filter, then perhaps you could request if the NSC can treat them as dead pixels and remap them.
 

Last edited:
Deathlines

perhaps that is what your camera can see?

heh.

jokes aside, if the lines are in the same area, it might be due to the scratches i guess

Now that is what i'm afraid of, but if it's scratches it wouldn't be presented in dots like David said.

While it could be dead pixels, it seems unlikely if there is a halo around the pixels. Could you share a RAW image somewhere online for us to download for better assessment ? I don't quite understand how a scratch will present itself as a dot on the image either. Do you have another lens to try with and also gives the same result ?

Should the result is really due to scratches in front of your sensor or on the AA filter, then perhaps you could request if the NSC can treat them as dead pixels and remap them.

Previewing them from the camera itself and from the computer is different. On the camera it shows the halo but on the computer it shows black dots. It could be dead pixels, but i really spent alot these days and can't really affort NSC to charge me for it's price again.

uploading the photo on mediafire now. Takes awhile
 

Here's the full untouched NEF file, take note the spots are kinda everywhere but really really difficult to spot. DON_6488.NEF
 

Here's the full untouched NEF file, take note the spots are kinda everywhere but really really difficult to spot. DON_6488.NEF

I have did an assessment of the raw image. It doesn't look like dead pixels, but I would like you to do another trial to help yourself.
First use a long exposure of around 10s, face your lens towards a plain wall and keep moving wild around the wall's plane after you release the shutter as long as the shutter is still open. Please do not use lossy compression for your RAW when you do this test. Use no compression for your RAW, then upload the RAW image again.

I can't be too certain this is dead pixel because I see that mild compression is there when I open up using Camera Raw in CS4. I can only see 1 very obvious 3x black pixels in horizontal placements. That is very much like dead pixels or perhaps due to the scratches.

In any case, is NSC going to charge you for dead pixels remapping ? Please find out with NSC as I'm not sure.
 

I have did an assessment of the raw image. It doesn't look like dead pixels, but I would like you to do another trial to help yourself.
First use a long exposure of around 10s, face your lens towards a plain wall and keep moving wild around the wall's plane after you release the shutter as long as the shutter is still open. Please do not use lossy compression for your RAW when you do this test. Use no compression for your RAW, then upload the RAW image again.

I can't be too certain this is dead pixel because I see that mild compression is there when I open up using Camera Raw in CS4. I can only see 1 very obvious 3x black pixels in horizontal placements. That is very much like dead pixels or perhaps due to the scratches.

In any case, is NSC going to charge you for dead pixels remapping ? Please find out with NSC as I'm not sure.

thank you very much david! I will do the test tomorrow. I used Uncompressed 14bit. I'm very sure the file you downloaded has the same RAW recording.
 

Don't look like scratches to me. Look more like dead pixels. bring it into Nikon for a re-map.
 

Tested on white paper. Various ISO range and apertures. Rest are on the rar folder. Ignore the watermark. forget to remove as i went batch process.


1.80sec f.16 HI2 by DonParadox, on Flickr


0.6sec f16 iso 500 (Shake) by DonParadox, on Flickr


(Test 2.rar)


There were no dead pixels shown in the photos but still visible in the LCD screen on my camera. It's weird. Any ideas why?



called NSC about the dead pixels, they quote me $160. That is the price for full maintenance, don't know why they charge me that. I asked for just the remap and they said i must go down and let them have a look first.
 

Last edited:
Tested on white paper. Various ISO range and apertures. Rest are on the rar folder. Ignore the watermark. forget to remove as i went batch process.


1.80sec f.16 HI2 by DonParadox, on Flickr


0.6sec f16 iso 500 (Shake) by DonParadox, on Flickr


(Test 2.rar)


There were no dead pixels shown in the photos but still visible in the LCD screen on my camera. It's weird. Any ideas why?



called NSC about the dead pixels, they quote me $160. That is the price for full maintenance, don't know why they charge me that. I asked for just the remap and they said i must go down and let them have a look first.

My greatest suspicion is it has nothing to do with your sensor. It is your lens. Have you tried another lens ?
If it is sensor issue, it will show up regardless with aperture you are using. I notice those speck of dirts and a blurred horizontal streak of seeming deadpixels only show up when you are taking a f/16, while at f/2.8 none shows. Do you have another lens to test with to isolate the possibility of the sensor issue ?

Also please do not send us jpg for assessment of such issues, JPG files will blur out any possible pixel size issues. Try PNG with lossless compression at 24bits.
 

Last edited:
David Kwok said:
My greatest suspicion is it has nothing to do with your sensor. It is your lens. Have you tried another lens ?
If it is sensor issue, it will show up regardless with aperture you are using. I notice those speck of dirts and a blurred horizontal streak of seeming deadpixels only show up when you are taking a f/16, while at f/2.8 none shows. Do you have another lens to test with to isolate the possibility of the sensor issue ?

Also please do not send us jpg for assessment of such issues, JPG files will blur out any possible pixel size issues. Try PNG with lossless compression at 24bits.

The first set was taken with a 14-24 while the 2nd set was taken with 50mm. Using the different lenses and testing it out still shows the dots with halo when 100% zoomed in, on the camera LCD. Sorry on the part for the jpegs, I will upload the nef asap.
 

The first set was taken with a 14-24 while the 2nd set was taken with 50mm. Using the different lenses and testing it out still shows the dots with halo when 100% zoomed in, on the camera LCD. Sorry on the part for the jpegs, I will upload the nef asap.

Lets just concentrate on final output from the NEF. If it's sensor related, it wouldn't matters where you observe the issue. We will ignore the halo issue and just concentrate on visible dead pixels. On Raw, dead pixels are very easy to detect and almost constant.

Perform the following across multiple lens and apertures all on ISO 100, shaking it across a plain wall

Set A (f/8, lens A)
Set B (f/8, lens B)
Set C (f/2.8, lens A)
Set D (f/2.8, lens B)

Do it in RAW and upload the images.
 

David Kwok said:
Lets just concentrate on final output from the NEF. If it's sensor related, it wouldn't matters where you observe the issue. We will ignore the halo issue and just concentrate on visible dead pixels. On Raw, dead pixels are very easy to detect and almost constant.

Perform the following across multiple lens and apertures all on ISO 100, shaking it across a plain wall

Set A (f/8, lens A)
Set B (f/8, lens B)
Set C (f/2.8, lens A)
Set D (f/2.8, lens B)

Do it in RAW and upload the images.

Thanks David,!I will do the test tonight.:)
 

I note you seem to be using a D3.

D3 oil spots on sensor. Is it problem solved by now? : Nikon D3 - D1 / D700 Forum: Digital Photography Review

Even Thom Hogan has mentioned it:
New D3 bodies have a tendency to throw oil on the sensor, which is a pain to clean that first time. But dust bunnies like the D3 sensor more than any other recent Nikon. You'll be cleaning more often.

http://bythom.com/nikond3review.htm

And I thought it was a Pentax K5 issue only.

Then again, Thom Hogan also says that:

The D3 series is notorious for the shutter throwing oil, some of which eventually ends up on the filter over the sensor. When you stop all the way down oil tends to be a larger undefined gray area rather than the smaller and more defined blackness of hairs, grit and other things sitting on the filter. If you do a wet cleaning and it still looks like you have those undefined gray blobs, it's probably oil. Unfortunately, they don't come off easily. You probably should have Nikon do the cleaning, as the amount of pressure and cleaning you have to use to remove it puts your filter at risk. Soapy cleaning liquids (e.g. Visible Dust's) seem to work better on oil than alcohols (Eclipse).

http://www.bythom.com/cleaning.htm

Anyways, just a possibility. Cheers.
 

Last edited:
Thanks for the examples edutilos. The halo is visible only when seen on the camera LCD. When transferred to the PC, it shows Hard black dots. And the halo appears in different colour yellow and cyan. Then again i have to test it again tomorrow as i'm busy for tonight. Thank you for the help people! :D
 

Uploaded! Names of lens and settings are inside. Thanks all! Test 3.rar