excessive noise or pixel peeping?

bugsy603 · 07-06-2010, 06:02 PM

I got a new K7 about 4 months ago, which I've been painstakingly trying to learn & master. I thought it was great until I went on a photo trip last week with a friend of mine who was shooting with his Nikon d70. He uploaded his photos to my laptop, and later I added mine. When I went into Photoshop Elements to look at the various pictures, I noticed two things:

If I change the View to 100% his photos barely get larger & they are so crystal clear with no visible noise at all. My K7 photos get much bigger (a real zoom!) and the noise is visible even at ISO 200 with excellent histograms. I then pulled up some shots I'd taken when I had my K20d, and I didn't notice nearly any noise with similar settings.

In smaller format, or views my photos look great...I wouldn't even know about the noise if I didn't 100% view (or view actual print size)...

I read an article that said that rest assured, print would be fine, and that the added noise was due to the higher MP count. I don't understand why this would be. Can someone explain this to me, or discount this theory?

SpecialK · 07-06-2010, 06:44 PM

Viewing an image 100% and looking for flaws is pixel-peeping. Whether that's good or bad is up to you.

The D70 is a 6MP camera, so images from it will not be as large as your 14.6MP K7 at 100%. More pixels jammed onto the sensor generally increases noise.

We'd have to see a sample of the noise you refer too to make a knowledgeable judgment about it.

Marc Sabatella · 07-07-2010, 10:07 AM

To put the above in other way - if you blew up his images as large as yours, they'd be just as noisy if not noisier. It's somewhat of myth that more pixels means more noise - viewed at the same size, the difference should be insignifant. The main reason the myth got started was because of people viewing images at wildly different sizes, as you did.

There is a *slight* extent to which more pixels could mean more noise even at the same viewing size, as more pixels means the spaces between pixels is probably more significant as a percentage of total area. But sensor makers have done a pretty good job of making that scale well through advancements in technology.

Since you also see a difference between the K-7 and K20D, that's presumably not on account of any difference in viewfing size. The K-7 sensor *is* slightly noisier (one of the penalties of adding video support, I assume). But I'm surprised you can see a difference at ISO 200. We'd have to see unresized crops from the images to say if you're just being unusually picky about tiny differences, or if there is some sort of problem. note the images should be exactly the same on all counts in order to be a valid comparison - there are a *ton* of variables that affect the appearance of noise. So you can't really say anything when comparing a picture of one type of scene taken with one type of lighting and processed one particular way with a picture of another type of scene taken with another type of lighting and processed in a different way.

bugsy603 · 07-09-2010, 05:52 AM

*smacks head* of course the sizes of the photos are different. The Nikon produced a 3008x2000 image, and my K7 produced 4672x3104.

I guess what I still don't quite get though (forgive me if I'm being obtuse) is that when I open my Photoshop Elements 8, and change view to actual print size...why aren't the images the same? Isn't a print size a specific size regardless of the original size of the photo? I know this is probably really a question for a PSE forum... If I enlarge to 100%, I can see that my two photos are still different in size, my K7 being much, much larger so I get that I'd see more noise there.

Its just if I am going to compare my two systems and their image quality...so I can tweak my settings if need be...I want to make sure I'm comparing apples to apples. I've read that adjusting the High/Low Adjustments settings & the Sharpness settings, as well as bringing the Noise Reduction at ISO level down to the 200 would help...I want to be able to accurately examine the photo quality in comparison to other makes/models...

Thanks for your help,

Marc Sabatella · 07-09-2010, 10:15 AM

"Actual print size" means nothing to me when viewing on a monitor. I'd have to know more literally wha it was *actually* doing. But the best way to get an apples-to-apples comparison is to size both pictures to the same dimensions - say, 1200x1800 (the dimensions of a 4x6 print at 300dpi), or even 2000x3000 - then compare both at 100%.

MPrince · 07-11-2010, 04:10 AM

Another good way to compare the images is to actually print them.

Votesh · 07-11-2010, 10:41 AM

Yeah, the best thing to do would be to print the photos at the same size. Viewing on a monitor only goes so far.

dragonfly · 07-11-2010, 11:05 AM

This thread isn't going to help without pics.

bugsy603 · 07-11-2010, 10:33 PM

yes, thank you. I know I can print them to compare (and have)...my question was more geared towards viewing in Photoshop. & thanks to the earlier posters, my question was thankfully answered.

falconeye · 07-11-2010, 11:43 PM

Originally posted by bugsy603

I don't understand why this would be. Can someone explain this to me, or discount this theory?

To understand this in theory is easy.

Think of a sensor as of a photon counter. Now compare this to a vote counter in an election: Wouldn't the vote counter with a few large electoral constituencies have much less variation (aka noise) from constituency to constituency than the vote counter with many small constituencies? Still, the counter with many small constituencies does the better job (collecting more information for later analysis).

Deni · 07-12-2010, 12:50 AM

I've seen the same thing when comparing photos taken with my K100D and the K-X.

The images from the K100D are more pleasing to look at 100%. They sort of have more pop to them and sharper.

07-06-2010, 06:02 PM	#1
bugsy603 Inactive Account Join Date: Jun 2010 Posts: 12	excessive noise or pixel peeping? I got a new K7 about 4 months ago, which I've been painstakingly trying to learn & master. I thought it was great until I went on a photo trip last week with a friend of mine who was shooting with his Nikon d70. He uploaded his photos to my laptop, and later I added mine. When I went into Photoshop Elements to look at the various pictures, I noticed two things: If I change the View to 100% his photos barely get larger & they are so crystal clear with no visible noise at all. My K7 photos get much bigger (a real zoom!) and the noise is visible even at ISO 200 with excellent histograms. I then pulled up some shots I'd taken when I had my K20d, and I didn't notice nearly any noise with similar settings. In smaller format, or views my photos look great...I wouldn't even know about the noise if I didn't 100% view (or view actual print size)... I read an article that said that rest assured, print would be fine, and that the added noise was due to the higher MP count. I don't understand why this would be. Can someone explain this to me, or discount this theory?

07-09-2010, 05:52 AM	#4
bugsy603 Inactive Account Join Date: Jun 2010 Posts: 12 Original Poster	I think I'm beginning to see the light... smacks head of course the sizes of the photos are different. The Nikon produced a 3008x2000 image, and my K7 produced 4672x3104. I guess what I still don't quite get though (forgive me if I'm being obtuse) is that when I open my Photoshop Elements 8, and change view to actual print size...why aren't the images the same? Isn't a print size a specific size regardless of the original size of the photo? I know this is probably really a question for a PSE forum... If I enlarge to 100%, I can see that my two photos are still different in size, my K7 being much, much larger so I get that I'd see more noise there. Its just if I am going to compare my two systems and their image quality...so I can tweak my settings if need be...I want to make sure I'm comparing apples to apples. I've read that adjusting the High/Low Adjustments settings & the Sharpness settings, as well as bringing the Noise Reduction at ISO level down to the 200 would help...I want to be able to accurately examine the photo quality in comparison to other makes/models... Thanks for your help,

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Say no to pixel peeping - Bring back the latent image!	ChrisPlatt	General Talk	6	09-06-2009 06:22 AM
Excessive Noise?	jezza323	Pentax DSLR Discussion	7	04-17-2009 06:37 AM
Pixel Peeping justified?	beaumont	General Talk	6	09-18-2008 04:57 PM
The Pixel Peeping Syndrom	benjikan	Photographic Technique	23	06-08-2008 01:56 PM
extreme pixel peeping..	trog100	Pentax DSLR Discussion	16	12-27-2006 12:33 PM

07-06-2010, 06:44 PM	#2
SpecialK Pentaxian Join Date: Dec 2006 Location: So California Photos: Gallery Posts: 16,482	Viewing an image 100% and looking for flaws is pixel-peeping. Whether that's good or bad is up to you. The D70 is a 6MP camera, so images from it will not be as large as your 14.6MP K7 at 100%. More pixels jammed onto the sensor generally increases noise. We'd have to see a sample of the noise you refer too to make a knowledgeable judgment about it.

07-07-2010, 10:07 AM	#3
Marc Sabatella Veteran Member Join Date: Sep 2006 Location: Denver, CO Photos: Gallery Posts: 10,685	To put the above in other way - if you blew up his images as large as yours, they'd be just as noisy if not noisier. It's somewhat of myth that more pixels means more noise - viewed at the same size, the difference should be insignifant. The main reason the myth got started was because of people viewing images at wildly different sizes, as you did. There is a slight extent to which more pixels could mean more noise even at the same viewing size, as more pixels means the spaces between pixels is probably more significant as a percentage of total area. But sensor makers have done a pretty good job of making that scale well through advancements in technology. Since you also see a difference between the K-7 and K20D, that's presumably not on account of any difference in viewfing size. The K-7 sensor is slightly noisier (one of the penalties of adding video support, I assume). But I'm surprised you can see a difference at ISO 200. We'd have to see unresized crops from the images to say if you're just being unusually picky about tiny differences, or if there is some sort of problem. note the images should be exactly the same on all counts in order to be a valid comparison - there are a ton of variables that affect the appearance of noise. So you can't really say anything when comparing a picture of one type of scene taken with one type of lighting and processed one particular way with a picture of another type of scene taken with another type of lighting and processed in a different way.

07-09-2010, 10:15 AM	#5
Marc Sabatella Veteran Member Join Date: Sep 2006 Location: Denver, CO Photos: Gallery Posts: 10,685	"Actual print size" means nothing to me when viewing on a monitor. I'd have to know more literally wha it was actually doing. But the best way to get an apples-to-apples comparison is to size both pictures to the same dimensions - say, 1200x1800 (the dimensions of a 4x6 print at 300dpi), or even 2000x3000 - then compare both at 100%.

07-11-2010, 04:10 AM	#6
MPrince Veteran Member Join Date: Sep 2006 Location: WV Posts: 1,495	Another good way to compare the images is to actually print them.

07-11-2010, 10:41 AM	#7
Votesh Inactive Account Join Date: Apr 2008 Photos: Gallery Posts: 817	Yeah, the best thing to do would be to print the photos at the same size. Viewing on a monitor only goes so far.

07-11-2010, 11:05 AM	#8
dragonfly Veteran Member Join Date: Jun 2009 Location: on the wall Photos: Gallery Posts: 715	This thread isn't going to help without pics.

07-11-2010, 10:33 PM	#9
bugsy603 Inactive Account Join Date: Jun 2010 Posts: 12 Original Poster	yes, thank you. I know I can print them to compare (and have)...my question was more geared towards viewing in Photoshop. & thanks to the earlier posters, my question was thankfully answered.

07-11-2010, 11:43 PM	#10
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by bugsy603 I don't understand why this would be. Can someone explain this to me, or discount this theory? To understand this in theory is easy. Think of a sensor as of a photon counter. Now compare this to a vote counter in an election: Wouldn't the vote counter with a few large electoral constituencies have much less variation (aka noise) from constituency to constituency than the vote counter with many small constituencies? Still, the counter with many small constituencies does the better job (collecting more information for later analysis).

07-12-2010, 12:50 AM	#11
Deni Veteran Member Join Date: May 2007 Location: Tirana, Albania, South Europe, Planet Earth Photos: Gallery \| Albums Posts: 621	I've seen the same thing when comparing photos taken with my K100D and the K-X. The images from the K100D are more pleasing to look at 100%. They sort of have more pop to them and sharper.