The Megapixel Myth (info for the new people)

[Hickeroar]

I know MOST of you have read articles similar to this, but I figured it's nice to stick this sort of thing out there for any new people who aren't aware of this industry-wide "scam." I definitely fell for it lock-stock-&-barrel when i was back in my "point and shoot camera" stage.

http://www.kenrockwell.com/tech/mpmyth.htm
Make sure you follow the first two links in the first paragraph regarding the NYT writers test of different MP levels. The last section on 35mm (and other formats) vs Digital was pretty eye-opening.

Ken also did a comparison of the Nikon D40 (6MP) and the Nikon D40x (10MP version of the same camera) and ended up concluding that the D40 was the better of the two...especially at the price. He said you'd be better off getting the D40 and putting the saved $200 (now $250) toward a nicer lens.

Just some info for anyone who's interested!

[Ronnoco]

Actually not a myth. If you read carefully around half way through he admits that there is a difference in sharpness. His initial comparison of the bench at 4 megapixels and a blow up of the section which he said was fine....really was terrible. It looked artificially sharpened and there was muddy colour. Obviously, he does NOT have an experienced photographic eye.

Sure other factors are important as well in photo quality but when Popular Photography in their lab compared a number of 8megapixel DSLRs with each other and a few 6 megapixel DSLRs, the lower quality images were from the 6 megapixel DSLRs.

I know I can certainly identify from the photo which megapixel camera I used, just by looking at the sharpness, and that is on screen.

Ronnoco

[Hickeroar]

LOL... This guy has had photos on display all over the US and won all kinds of awards. To assert that he has no experienced photographers eye is absurd...

The point isn't that there's NO difference, it's that it's barely noticable, if at all, in any circumstance but a certain select few. The whole megapixel race by p&s camera companies is meaningless. If you have a sensor half the size of my pinky fingernail...those 10mp shots are going to look like crap no matter what.

Megapixels, passed a certain point, aren't NEARLY as important as is sensor quality and size as well as the quality of the optics in question.

If you take my 6MP D40 and take some shots and then go out with samsungs new 8mp SLR and take the same shots, my camera will produce FAR better shots than the samsung will.

[Rocket_Scientist]

Well, I have to admit that I could not stay interested in the article, but I am a little confused. In the one linked article, it seemed the author was saying they compared blowups of three pictures at 13, 8 and 5 MP, but that they had all been down "rezzed" (resolution? resized?) from the same photo, presumably the 13 MP. Whatever the merits of the argument, that does not sound like the same test as shooting the picture with different cameras having 13, 8, and 5 MP. For one thing, it includes how good the software was that did the interpolation. Oh well, I like my 6 MP, and it is certainly better than the 0.8 MP I used to have.

[Hickeroar]

You're talking a HUGE difference between 800KP and 6MP. That would be obvious. The bottom line with the downsize is that there were far less PIXELS left over when the shots were downsized. If purely the number of pixels mattered, the two smaller shots should have been obviously different...but they weren't.

There was also a second test done using a different method.

You can't do an accurate test with different cameras. There are differences from camera to camera that any eye could see. The same shot from the same camera is needed. There are probably hundreds of articles out there regarding the "myth" that you can read as well...

[Medley]

I disagree with this "myth"- to some extent.

I got into photography from a post-processing background, which is kind of backwards when you think about it. I've spent a great deal of time learning post-processing, and have specialized in image sharpening and optimizing for print. That is, after all, the goal of most post-processors: better prints- right?

I'm going to try to keep this post as non-technical as possible, so forgive me if I assume too much.

First of all, an image sharpened to look good on screen always prints too soft. No exceptions. There are a variety of reasons for this, not the least of which is that the printer is laying down round dots of ink according to a map of square pixels. Even if you've done everthing else right, the image will lose sharpeness when printed. So if you're asking if I can take an image that most people print, upsample it, and retain the same image quality, the answer is "yes, probably, because the 'average' viewer ( such as you would typically find in a 'street' experiment ) leaves a large margin of error initially."


Being able to sharpen an image onscreen does little good if you can't control the amount of sharpening in the print. That involves being able to print the sharpening halos at a specific size without having them either upsampled (which produces an oversharpened and pixellated effect) or downsampled (which produces image softening). Doing so means understanding that there is a HUGE correlation between image resolution (pixels per inch) and average print resolution (dots per inch). I say average because most printer manufacturers advertise the maximum print resolution of their printers, not the average. In most cases, this is accomplished by either upsampling the image or downsampling it.

Any person with experience in post-processing will tell you that downsampling is infinately easier and less complicated than upsampling. In addition, downsampling an image correctly has the effect of sharpening it, while upsampling has the effect of softening it (though the degree of softening can vary widely, depending on the method of upsampling.) Both will likely still need additonal sharpening, the only difference being in how agressive that sharpening must be.

Bottom line, more mp equals less agressive sharpening. And yes, the differences are going to be subtle. But it is exactly those subtle differences that can separate a good image from an outstanding one. In fact, subtle differences are really what sharpening is all about.

If you have questions, feel free to ask.

- Joe U.

[Ronnoco]

You're talking a HUGE difference between 800KP and 6MP.

There are probably hundreds of articles out there regarding the "myth" that you can read as well...

Yes, but NOT believable to anyone who has worked with television screens, digital projectors and cameras with different pixel ranges. Differences show up in postprocessing as well in Photoshop etc.

Ronnoco