I have been playing around with my Nikon cool pics 5700 and am still getting a lot of digital noise on some shots. Dave Williams seems to think it's the digital zoom? Any suggestions? Here's an example from 7th Heaven on Blackcomb this weekend. Notice the lines? <BR> <BR>Please advise.... <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95328.jpg" alt="">

But it's not in this one? <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95331.jpg" alt="">

I agree with Dave. Digital zoom is bad, don't use it! <BR> <BR>

digital zoom is crap. turn it off. i dont know if that is exactly the reason you are having problems, but nothing good can ever come from digital zoom.

email me the original pic of these two images at my other email account: <a href="mailto:bill@ryno-gear.com">bill@ryno-gear.com</a> <BR> <BR>I'll play with them in Photoshop and see what I can figure out. That camera definitely shouldn't be producing noisy images. <BR>

You can get noise for a lot of reasons besides the digital zoom. In a scene with strong light and dark areas like the upper shot the cam may be accounting for the brighter areas exposure wise and the chip doesn't have the latitude to correctly expose the darker areas resulting in noise as the chip tries to interpolate the darker part of the image. Are you shooting those on full auto?

I was in full auto. Should I have been correcting for appature or shutter speed?

Digital zoom can cause the noise almost every time. My photo printer software has a digital zoom eliminator feature when printing pics. Some photo editing software also has this feature.

One other thing to consider, everytime you save it as a jpeg you are taking more and more data out of it. Which would overtime, result it more quality being lost. Just something to know if you dont already.<img src="http://www.wakeworld.com/MB/Discus/clipart/wink.gif" border=0>

It could very well be the digital zoom. But it doesn't appear to be throughout the picture to me which I would expect from that. If the camera is raising the ISO way up trying to stay within a certain range of speed and aperature you could get some noise too. Next time try it without the zoom and on manual and see if you can eliminate either factor. <BR> <BR>RD

There is an auto focus mode for single shot or continuous. I think the continuous has something to do with it as well.

Like when I'm moving, it's re-focusing or something? <BR>

Steven, I'm convinced that your camera is defective. The banding, or vertical stripes, should not be there. Even with digital zoom you are getting a ton of noise, way more than you should for that camera when shooting outdoor shots with adequate lighting. <BR> <BR>Here's a crop from the original - you can really see how much noise is in the pic. The banding is harder to see but it's there, and it's also in the pic of Steven where you can notice it in the jacket. <BR> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95487.jpg" alt="">

i would agree - the banding is very discernable in both shots. almost like slats on a house. i would have that checked out steven.

I messed around with the camera today. It has a noise reduction feature as well as a single shot focus selector but I couldn't figure out how to turn off the digital zoom. Here are a few examples. Bill I will e-mail you the original shots. <BR> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95502.jpg" alt=""> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95503.jpg" alt=""> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95504.jpg" alt="">

Well I'm not seeing much noise in those. Snow scenes can create problems when shooting with auto exposure. You're still getting some banding though.

Obviously those pics are MUCH better than the others - maybe the camera has a problem with cold? The sample images from that same camera posted on dpreview.com are outstanding, so the camera SHOULD be able to delivery quality results. I'm still leaning towards something being wrong with that one. The banding is a real issue and shouldn't show up even on a cheap camera never mind a Nikon. <BR>

I have also noticed that I get more digital noise in low light conditions with my camera.

Rich is right. Snow scenes wreak havoc on weekend photographers and digital cameras. White balance on something light gray. It may help. <BR> <BR>Andrew, you are dead on as well. When my production crew films early in the morning, before sun up, the digital noise is ever present. Its because the iris is set open to the max. Digital cameras cant handle it. We are shooting with $50,000 cameras as well. Our new cameras are handling it better though. <BR> <BR>A wise man once told me: <BR>"digital = poop" <BR>Take that to heart.

"We are shooting with $50,000 cameras as well." <BR> <BR>Lucky dog! <BR> <BR>A wise man once told me: <BR>"digital = poop" <BR> <BR>I think <BR>"digital = cheap(er)" <BR>might be more ap-pro-po! <BR> <BR>

Could you explain how a wide open iris contributes to digital noise? From everything I've read noise is only introduced if the sensitivity (ISO) is boosted, or by long exposures.

sorry for the F' up Rich. <BR> <BR>"Digital Zoom = Poop" .........much better. <BR> <BR>my Sensay (sp) is gonna kill me for misquoting him. I better break out the turtle wax. <BR>

A digital camera records images by measuring the voltage difference in individual pixels, which are affected by being hit with photons, compared to the baseline voltage. In a perfect world, this baseline would be constant. In reality, it is not. That's where you get noise. <BR> <BR>Now if the signal, i.e. the rate at which photons are hitting the sensors, is very low, it becomes less distinguishable from the background noise. <BR> <BR>Increasing the ISO equivalence basically just increases the baseline and the significance of the noise. It doesn't actually make the camera more sensitive. <BR> <BR>The bottom line is: Less light (per unit time) reaching the sensors = more noise and higher ISO equivalence = more noise. You will always get less noise with the widest aperture. <BR> <BR>Incidentally, the reason that digital SLRs have less noise than their point-and-shoot counterparts is their larger CCDs. Each pixel's sensor is larger, thus gathering more light, producing a better signal, and hiding the baseline noise. <BR> <BR>As for Mr. Hahn's problem, I think the noise could be a function of the camera settings. The lines, however, are definitely not normal. I'd check the manual for the operating temperature range for the camera though before sending it in. I noticed that I can only see the lines in the green channel of the images. Interesting. <BR> <BR>BTW, it looks like you can probably turn off the digital zoom by going to record mode, hitting the menu button, going to the 'S' tab, and changing the "Zoom Options". Also, I didn't find how to do it in the online camera review I was looking at but, if you can, I'll try to force the camera to always use the lowest ISO setting. Bumping up the ISO on these cameras should usually be your last resort.

I reset the factory defaults and took some shots of Mt. Ranier today. When I get home I'll post them. It seems to have solved the problem for now. I stopped by the camera store and spoke with the manager. He said it could also be a sensor or processor issue and may need service.

Here are the raw pic's just resized. <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95758.jpg" alt=""> <BR>And a couple from the boat show too: <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95759.jpg" alt=""> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95760.jpg" alt=""> <BR>And one from after the boat show <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/95761.jpg" alt="">

Okay, problem solved. The first pics were taken with the ISO set to 800. That will cause noise every time... <BR> <BR>The bottom pics obviously look a lot better, those were shot after Steven reset the camera to factory defaults and the ISO for those shots was 100. Nice and clean pics...

Reduce noise: <BR> <BR>- Gaussian Blur 4.0 <BR>- Edit> Fade Blur > (Mode = "color" from drop down menu) <BR>- Problem solved. <BR> <BR><img src="http://www.wakeworld.com/MB/Discus/messages/87667/98321.jpg" alt="">

Nick, we solved the problem by putting the ISO back to where it should have been - 100 instead of 800! <BR>

Now im no digital camera buff but a ccd sensor is a whole bunch of photodiodes that convert light energy (photons) into electrical charge which is collected in a potential well. In the case of an overexposed photo these wells become full and the response to additional incident light becomes non-linear. <BR> <BR>Anyway the charge stored in this matrix of potential wells is transferred via a vast array of serial and parallel shift registers to an output amplifier that in turn produces an analogue raster scan of the photo-generated charge across the entire array of photodiodes. The banding in my opinion would correspond to the location of these parrallel shift registers as the size of the potential well in the silicon is would be less than that for the rest of the ccd. Thus they can hold fewer electrons and the bands are darker than that of the rest of the image. The solution as you have found is to lower the iso setting. As for digital zoom, its basically cropping and resizing the photo effectively reducing the resolution, to some of you reducing the resolution is unacceptable but if you're only going to view it on a monitor or print out a small print (or have a particularly high megapixel count) it can be acceptable. As for the quality of digital SLR cameras compared to that of compact cameras the size of the CCD does have some impact as does the quality of the lens. Feel free to pick this to shreds!!!

I'm not quite sure what you're saying here but it doesn't sound right to me. The amount of exposure is determined by 3 things: the aperature, the shutter speed and the ISO setting of the chip. Higher ISO's crank up the gain factor the CCD is working under. If a camera setting is fixed at a high ISO the CCD is in a high gain state and thus more sensative to the amount of light hitting the chip. The auto exposure modes in the cam account for this by increasing the shutter speed, decreasing the aperature size or both so the image isn't blown out. In the case of Steve's picture the camera accounted for the brightness of the sky fine but underexposed the mountain part of the shot. Underexposure with the CCD set at a high ISO creates pixel noise and banding in turn is caused by the D to A converter being unable to resolve the noisey areas properly. If this is what you're saying than fine but it sounds like you're saying that there are physical differences in these shift registers of CCD itself which are causing this.

Thankyou for your reply, and for not being rude when pointing out errors. <BR> <BR>>I'm not quite sure what you're saying here but it doesn't sound right to me. The amount of exposure is determined by 3 things: the aperature, the shutter speed and the ISO setting of the chip. <BR> <BR>Ok, i understand this. <BR> <BR>> Higher ISO's crank up the gain factor the CCD is working under. <BR> <BR>Ok this is where my knowledge is a bit hazy, you mean that changing the ISO setting on a digital camera changes the gain on the output amplifier?? I always thought it changed the shutter speed (integration time) and thus varied the amount of light incident on the photodiodes that way. <BR> <BR>>Underexposure with the CCD set at a high ISO creates pixel noise <BR> <BR>That makes sense, however it would also be the case if a change in iso setting means a change in integration time as the pixel noise you are talking about is due to a low signal to noise ratio. That is the noise from the amplifier and sweeping of electrons etc etc is significant compared to the magnitude of the signal. <BR> <BR>>banding in turn is caused by the D to A converter being unable to resolve the noisey areas properly. <BR> <BR>I dont think this is correct, do you mean A/D converter? If this was the case the banding would not be in regular intervals. <BR> <BR>Now for some clarification on what i was saying, it is only a guess after all. The parrallel shift registers require space on the silicon substrate and thus the potential wells adjacent to these registers cannot be as big as those in the rest of the sensor, thus these cannot store the same amount of charge and you get darker bands corresponding to these smaller due to lower number of electrons. <BR> <BR>

The CCD is is just a substitute for a film frame in a digital SLR so the shutter and aperature are definitely independent of the film speed or ISO. This may be different in a video camera where I don't believe there is a physical shutter. Now how they make the CCD more sensitive to simulate "faster film" isn't something I'm down with as far as all the technical details go but I do know high ISO makes the underexposed areas prone to noise. I pointed this out to Steve above when everyone else was talking about the digital zoom/defective camera. You're right about the convertor, I think the signal is analog coming off the CCD so I got that backwards. The Canon 1d exhibited banding at high ISOs when it was first released which I believe was later fixed with a firmware upgrade. So I figured banding must be something in they deal with via software. Thanks for your explanation although I'm not sure I understand the exact relationship between what you're saying to how the camera actually works! (I seem to only have room in my brain these days for "how it works" and not "why it works")