The truth about digital cameras...?

[Mr. Pickles]

David Pogue is a New York Times technology writer and future TV star. In his column "Pogues Posts", he writes about a comparison in digital camera print resolutions. Very interesting, but most won't like the answer. I loved it.

http://pogue.blogs.nytimes.com/2006/...ogues-posts-2/

[Jon]

One responder did make a valid point - what cameras were used in the test? Divulging that would go a long way further toward debunking the megapixel myth.

It sounds like I'm up in arms about it, but remember that I moved "up" from a 6mp main camera to a 4mp main camera...The 6 shoots at 72 points and the 4 shoots at 300. Big difference there.

[JohnC]

The resolution in pixels per inch doesn't come into play here. A digital camera doesn't shoot at a specific resolution (such as 72 or 180 or 300 or 360). That is an arbitrary number assigned to the image to determine the size of the print. A 6 megapixel camera image at 72 dpi is the same as one at 300 dpi concerning file content and quality. They will both look the same on screen, but the 72 dpi image will PRINT much larger.

Check out the discussion at http://www.photocamel.com/index.php/...html#msg121140

What does come into play here is the quality of the pixels. A 3 megapixel Canon D30 is still one of the better cameras out there. Good size sensor, clean pixels, great color. Compare it to a point-n-shoot 7 megapixel camera with a tiny sensor and high noise. My bet goes to the D30.

John

[Jim Sykes]

One responder did make a valid point - what cameras were used in the test? Divulging that would go a long way further toward debunking the megapixel myth.

They were all shot with the same camera, look at this quote and read it carefully:

We blew up a photograph to 16 x 24 inches at a professional photo lab. One print had 13-megapixel resolution; one had 8; the third had 5. Same exact photo, down-rezzed twice, all three printed at the same poster size.

They took a 13 meg image, then down sized it to 8 and to 5 megs. They then took each file and printed the same size.

What they are saying is that having only 5 megs of file printed the same as the 8 and the 13 meg file. Each one would have had to been upsized to make the print, the 8 and 5 needing more interpolation than the others.

Its a good test, but not entirely telling. Its good in the sense that it allows people who may use a quality lab to get the same results out of lower res photos as they could higher res photos.

However, there are variables that could have made the results turn out differently in another test. First off the photo may have been one that took to the resizing really well. It may not have had a lot of fine detail or any number of other reasons, but another photo may not have taken to the resizing as well as this one and in the end would have created a difference in the prints.

Also, this lab likely had a very good and expensive RIP program that would allow for high quality resizing and thereby making the large prints all look sthe same. If it was done at another lab or on a printer that didnt have as a good of a RIP, there may have also been a difference in the print quality.

So while in this case things worked out, in other cases the outcome may not have been the same.

What this does tell us though is a good lab and a good photo can indeed make prints in larger sizes that look really good and that does bust a myth that many film lovers try to perpetuate.

BTW, I'm one that does not believe that you need a 1Ds or digital back to have digital prints in large sizes, so I"m not trying to bust up this article, just stating some facts and variables that could cause a different outcome in another case.

[Sidebyte]

Impressive display for the buying public but not legitimate for photographers who understand pixel pitch and the benefits of large sensors vs. fingernail size P&S sensors. Before I sold it, my old D1h at 2.74 mp, it made awesome 8 X 10's and even 12 X 16's sometimes depending on the subject. Comparing that to my Kodak 14n, the only real benefit was the ability to crop. If you say you don't crop, you're either a liar or certainly a better man than me " Gunga Din".

[Jon]

Originally Posted by Jim Sykes

They were all shot with the same camera, look at this quote and read it carefully:

They took a 13 meg image, then down sized it to 8 and to 5 megs. They then took each file and printed the same size.

I never was very good at story problems! That makes sense.

[Blinky]

Seems fair. I don't doubt his results.

For my own purposes I've found that 150dpi at output size is enough pixels for maximum print quality (and 75dpi will get you by). For offset print I use either 1.41 or 2.0 times the linescreen at output size so a typical 133 linscreen magazine would go at 188dpi or 266dpi at the output size... a 5MP camera is plenty. What JohnC says about the quality of the pixels along with the pitch, the A/D converter, etc is crucial to good color. I'm not gonna even touch the sensor size debate.

But! The one thing the article seems to miss is that most digital images are cropped and manipulated. It's very nice to have 12bits of luminosity information and lots of extra pixels when editing image files; you downsize/sample later for the specific output device. Not everybody does that but I do on pretty much every shot that gets printed.

The old printing rule was scan at the output resolution at 1:1... saves on disk space and risks no resampling loss. I NEVER scanned at output resolution because I couldn't help myself, I screw around with every file by squishing and stretching the histogram, sharpening and, more often than not, color balancing. More was always better to me.

When picking my next camera I'll be looking at dynamic range first, sensor size second, then high ISO performance... as long as it has 5+ megapixels the pixel density is still down the list a bit. I don't print many posters though.

[Gbee]

Nothing really new here.

The point is who notices the differences. Printing from several different files from several different cameras, I am often amazed at the quality of the typical 10x8” print.

I’ve printed several 20x16” prints and I’d therefore be inclined to go along with the article, I’ve seen it myself, for myself, so I k now it can be done.

However, that’s all the article tries to say and many replies to this and several other threads on the same subject on every forum tends to read much more into the procedure than is actually expressed.

There are limits to everything and if one simply took a few files and ran them trough a poor quality printer then they will all look the same, this is pretty much what the article says, read it again, it’s clever but is says so little.