Red, blue, and (a few) green pixels in long exposure shot

[wkteoh]

Hello fellow CS-ers,

Found this weird problem last night as I was fooling with long exposures last night with my D90. Below is a 100% crop of the problematic picture.

Taken at ISO 100 (LO 0.3) and exposure was 750+ seconds. Long exposure noise reduction was turned on, but the battery died as it was still doing the dark frame noise reduction (probably 1-2 minutes from completion though)

From what I have seen elsewhere, these seem to be hot pixels. I was wondering if any of the below is causing it:
a) I've been taking 30s to 5 mins exposures prior to this for probably for 1 hr and 20 mins almost consecutively (all with long exposure NR turned ON). Is this why the pixels got warmed up?
b) The incomplete noise reduction is causing this?
c) Sensor is screwed up?
d) UFO?

Anyway I tested it again this morning. Lens cap on, ISO 3200 1/4s; then ISO 100 for ~300s. No noticeable dead/hot pixels seems to be apparent from camera screen.

What gives?

 

[Rashkae]

It's just ISO noise from the long exposure.

 

[wkteoh]

I see. Is there any way to remove them other than shooting raw?

 

[Rashkae]

I see. Is there any way to remove them other than shooting raw?

RAW will show them too. You would need to let the cam finish it's NR shot, or use PP to clean them afterward.

 

[night86mare]

use pp to clean, not an issue.

 

[wkteoh]

Thanks guys.

May I ask what's the process involved in removing these noises? Noise reduction (using NeatImage) doesn't seem to get rid of it - or maybe I'm just not proficient enough with the program?

I talked to a girl yesterday and she told me she took a 10-hour exposure with D70 without having any of the R/G/B pixels showing up in the picture - then how come it is happening with D90??

 

[chanjyj]

Thanks guys.

May I ask what's the process involved in removing these noises? Noise reduction (using NeatImage) doesn't seem to get rid of it - or maybe I'm just not proficient enough with the program?

I talked to a girl yesterday and she told me she took a 10-hour exposure with D70 without having any of the R/G/B pixels showing up in the picture - then how come it is happening with D90??

Manually do it. I don't expect a noise reduction program to be able to cope with something that doesn't really look like noise.

 

[wkteoh]

Too much work methinks, there are probably thousands of those pixels in the full picture.

By the way I mentioned shooting in raw because there seems to be a program which can remove the 'hot pixels' so to speak. And sometimes I don't use LENR because it takes way too long and eats battery.

 

[night86mare]

Thanks guys.

May I ask what's the process involved in removing these noises? Noise reduction (using NeatImage) doesn't seem to get rid of it - or maybe I'm just not proficient enough with the program?

I talked to a girl yesterday and she told me she took a 10-hour exposure with D70 without having any of the R/G/B pixels showing up in the picture - then how come it is happening with D90??

i highly doubt that the girl didn't experience hot pixels for 10 hour exposures:

http://www.imaging-resource.com/ARTS/HOT/HOT.HTM

the longer the exposure, the more likely you are to experience hot pixels:

In fact, with a long enough exposure to darkness, a disturbing pattern of exposure will appear from any CCD, because CCD elements tend to leak current. They may not all produce a bright white spot, but they are all -- with a long enough exposure -- capable of reporting exposure in darkness.

 

[night86mare]

btw, what are you trying to do, shoot star trails?

 

[ahbian]

How about opening up the aperture and reducing the exposure time?

Or the "shortest cut", convert to B&W, then your pixels will become stars. :bsmilie:

 

[alphabeta93]

How about opening up the aperture and reducing the exposure time?

Or the "shortest cut", convert to B&W, then your pixels will become stars. :bsmilie:

hahaha there's always a way out eh?

 

[wkteoh]

ah bian | not a bad idea. Converting B&W does fix a lot of problems. Regarding opening up the aperture, I need the longer shutter time to record trails.

night86mare | yes, trying to shoot star trails but a short one, not hours long one (which is for camping). How would you PP out the hot pixels? By the way are these hot pixels at all, usually (even in the link you showed me) there are only a few but not all over the picture.

 

[Ghostfit]

I've read something about this awhile back, it seems on long exposure shoots (1" and above) , the sensors can kinda leaks tiny amount of electricty into the CCD/Cmos matrix which causes these colored spots.

Usually the system will snap another normal shot to compare and correct (NR) these artifacts, you are seeing this as your camera have not finished processing the picture b4 it ran out of juice.

You can also try simulating this by enabling the bracketing mode and shooting long exposure shots with the lens cap on.

 

[wildcat]

Taken at ISO 100 (LO 0.3) and exposure was 750+ seconds. Long exposure noise reduction was turned on, but the battery died as it was still doing the dark frame noise reduction (probably 1-2 minutes from completion though)

From what I have seen elsewhere, these seem to be hot pixels. I was wondering if any of the below is causing it:
a) I've been taking 30s to 5 mins exposures prior to this for probably for 1 hr and 20 mins almost consecutively (all with long exposure NR turned ON). Is this why the pixels got warmed up?
b) The incomplete noise reduction is causing this?
c) Sensor is screwed up?
d) UFO?

Anyway I tested it again this morning. Lens cap on, ISO 3200 1/4s; then ISO 100 for ~300s. No noticeable dead/hot pixels seems to be apparent from camera screen.

What gives?

The LE noise reduction didn't run for that picture. I assume that you normally don't get that problem with LE noise reduction turned on, based on the other info you provided, correct?

i highly doubt that the girl didn't experience hot pixels for 10 hour exposures:

http://www.imaging-resource.com/ARTS/HOT/HOT.HTM

the longer the exposure, the more likely you are to experience hot pixels:

She prob had her LENR turned on also, thus didn't show?

 

[night86mare]

20 hours of shooting?

i'm not sure if any camera out there has sufficient juice in one battery to do that.

 

[wkteoh]

wildcat | My point is, even with LENR turned off, it shouldn't be all over the place - as in the girl's 10-hour exposure, that would mean that she needs 20 hours in total which she mentioned she did not use.

 

[wildcat]

wildcat | My point is, even with LENR turned off, it shouldn't be all over the place - as in the girl's 10-hour exposure, that would mean that she needs 20 hours in total which she mentioned she did not use.

Your point is? Whether you should switch to the same camera as the girl?

As night86mare mentioned previously, it is highly doubtful that the girl didn't experience hot pixels for 10 hour exposure, and since you say that she didn't, then the only solution is to change to the same camera as her, correct? Otherwise LENR seem to be working fine for your camera other than that time when it didn't work because you ran out of battery just before it completed, and you're wondering why you're getting so many hot pixels without LENR. I mean who knows, maybe one is using a D3000 and another is using D3s so based on the limited information provided, I'm trying to figure out what your point is.

Or is it that you want me to tell you that your camera is spoilt?

 

[ovaltinemilo]

I talked to a girl yesterday and she told me she took a 10-hour exposure with D70 without having any of the R/G/B pixels showing up in the picture - then how come it is happening with D90??

is it 10 minutes or 10 hours?
十分钟还是十个钟?

Sorry, I just can't believe D70(no hot pixel and batt can last so long) is capable of that. Maybe I can try out for experimental purpose.

 

[liveevil]

What is the girl shooting?:dunno: 10 hours? star trail become sunrise already.