Camera resolution

[rongarrett]

The Nikon D700 is rated at resolving 2350 lines. The Canon 1DS Mark III is rated at 2800 lines. The new Sony A900 is rated at 3300 lines of resolution. The difference of approximately 1000 lines of resolution seems significient. At say 16 x 20 print size is this difference detectable by the average human eye?

[Franglais]

The Nikon D700 is rated at resolving 2350 lines. The Canon 1DS Mark III is rated at 2800 lines. The new Sony A900 is rated at 3300 lines of resolution. The difference of approximately 1000 lines of resolution seems significient. At say 16 x 20 print size is this difference detectable by the average human eye?

Where did you get these figures from? This sort of stuff used to interest me but not any more.

I can't print larger than A3+ (about 12x17 inches). Even at that size I'm unable to tell the difference between a 6Mpix camera and a 10Mpix camera in normal viewing. The cameras you mention are 12Mpix, 21Mpix and 24Mpix..

[Canuck935]

It may be a 1000 lines, but remember that these figures are lines per millimeter. It's unlikely you'd notice at anything but the largest of poster prints.

I would not let these figures be a factor in a decision on buying a camera. All of the current lines of DSLR's are fantastic camera's.

.

[photophorous]

I have not personally used any of these cameras, nor have I made a lot of digital prints, but based on the little experience I have, and a lot of reading, I'd say the answer is "it depends." The differences between the Nikon and the Sony should be noticeable in a print that large (assuming good lenses, good technique and close viewing). How noticeable would probably depend a lot on subject matter, the manner of display, and your expectations. I find certain subjects, like landscapes, require lots of resolution, because of the random patterns and intricate detail. With landscapes, I like to lean in and look at a print up close so I can see the details. Other things, like portraits or architecture, IMO, don't require as much resolution, so the differences would barely be noticeable, if at all. Plus, I'm much less likely to look at a 16x20 portrait from a few inches away.

Paul

[brmill26]

At say 16 x 20 print size is this difference detectable by the average human eye?

Absolutely not. 16x20 doesn't even stretch a modern DSLR. I've printed from my XTi at 20x30" and I can read text on decals on the race car and quite clearly see the individual mesh wires on the grill. I can even, only just, make out the direction of a screw head on the fender.

Those resolution figures are so high they're only going to matter in two cases. 1) Massive crops. You can crop further on the higher resolution image and retain lots of detail. 2) Printing at HUGE sizes. And by HUGE I mean you start measuring in feet.

So for those of us not shooting ads that are going to be put up in Times Square, those figures are pretty irrelevant to mere mortals' photography.

[rongarrett]

I can print up to 17 X 22. I rarely print any landscape at less than 11 X 14. I think a good detailed landscape image cannot be given credit at less than the latter figure. I like the 17 X 22.s even better

I'm not interested in pixel count, just resolution. Sometimes there is a higher pixel count but not higher resolution, {Sony A350 vs Nikon D700}. I know resolution is not the defining specification for image quality. My situation is that I currently own and use several cameras from 10mp to 14 mp. With the current crop of new cameras I just don't see a big advantage over what I already have. That doesn't stop me from trying to find a camera that actually does provide a sharper landscape image though. I truly believe that a kickin landscape image is usually one that has sharp defined detail. The sharper and more defined, the better the image?

I know I'm looking for a needle in a haystack, but I'm gonna keep on tryin.

Maybbe the Canon 5D Mark II be the needle I'm lookin for?

[Canuck935]

A good lens will do a lot more for your images than any camera body.

[photophorous]

I can print up to 17 X 22. I rarely print any landscape at less than 11 X 14. I think a good detailed landscape image cannot be given credit at less than the latter figure. I like the 17 X 22.s even better

I'm not interested in pixel count, just resolution. Sometimes there is a higher pixel count but not higher resolution, {Sony A350 vs Nikon D700}. I know resolution is not the defining specification for image quality. My situation is that I currently own and use several cameras from 10mp to 14 mp. With the current crop of new cameras I just don't see a big advantage over what I already have. That doesn't stop me from trying to find a camera that actually does provide a sharper landscape image though. I truly believe that a kickin landscape image is usually one that has sharp defined detail. The sharper and more defined, the better the image?

I know I'm looking for a needle in a haystack, but I'm gonna keep on tryin.

Maybbe the Canon 5D Mark II be the needle I'm lookin for?

I don't mean to offend anyone when I say this, but people who say you can make big (16x24) prints with an average consumer DSLR never seem to be talking about landscapes. I really think landscapes require more resolution than most subjects. Not just because the level of detail, but also because of the randomness. Digital does well with lines and uniform shapes, but film does better with random patterns because it has higher resolution. If making big landscape prints of excellent quality is your primary goal, and you don't want to shoot film, I'd get the camera with the most resolution that I could afford. Will the average viewer see the difference? I don't know...some probably won't. Is there a detectable difference that you will see and that keen viewers will see? I'm almost certain the answer is yes. Of course, if you have a cheapo lens and don't know how to use it, none of this matters.

Paul

[rongarrett]

I always buy the best lens available at the time I purchase, {Canon 400 mm / F2.8 for example for wildlife use}. Many of my lens cost way more than the camera. I own only two {out of some 15 total} DX lenses. Those two lenses happen to give stellar results I might add. I did so even though all my cameras are DX.

At 64, retired, and with some physical limitations, I fear I will be turning more and more to landscape photography. I just am not physically capable of crawling in the mud for an image anymore. Even with landscape I am learning I have to pick and choose. I recently visited a rarely viewed waterfall in the rugged North Carolina mountains. This waterfall is remote and at the end of a rugged very primitive trail that ascends very steeply at times. At my current skill level traversing large loose rocks on a steep trail is not my forte. Point is if I had been lugging an 8 X 8 film camera, I would have never made it. I must stick with digital. A 35K Hasselblad is out of reach $$$$$$ wise.

I'm still lookin for the needle I guess.

[rongarrett]

The Canon 5D Mark II will ship next week. Interesting to see what resolution figures it puts up.

[readingr]

I don't mean to offend anyone when I say this, but people who say you can make big (16x24) prints with an average consumer DSLR never seem to be talking about landscapes. I really think landscapes require more resolution than most subjects. Not just because the level of detail, but also because of the randomness. Digital does well with lines and uniform shapes, but film does better with random patterns because it has higher resolution. If making big landscape prints of excellent quality is your primary goal, and you don't want to shoot film, I'd get the camera with the most resolution that I could afford. Will the average viewer see the difference? I don't know...some probably won't. Is there a detectable difference that you will see and that keen viewers will see? I'm almost certain the answer is yes. Of course, if you have a cheapo lens and don't know how to use it, none of this matters.

Paul

Paul,

I have the Canon 5D and the Powershot Pro 1 so both are consumer products, the Pro 1 being compact but with L quality lens. The 5D is 12 and the Pro 8 Mpixels.

Using both cameras I have managed to print landscapes up to 24 x 16 and with the 5D it could have gone much bigger before loosing quality. All the landscapes were perfectly sharp. The thing is at this size people look at the photo's from a distance and therefore sharpness becomes subjective as the human brain compensates for the limitations of the printing .

If you've got a loupe on the picture and investigating the sharpness that way then I totaly agree and that is where the problems begin, because no matter how sharp the photo the printer cannot prevent the inks from merging so softens the picture.

I must admit; to print at that size I take care that the PP includes the sharpening and sizing of the image and I designed it for that size of picture. So I over sharpen which helps the printer compensate for the ink merging.

I have a couple of landscapes at that size on my wall in Basingstoke so if your ever here pop in for a look and a drink of course.

There was a test on one of the TV comparing film v Digital and they printed a picture (portrait of two people) at about3 or 4 stories high and Digital won for print quality. They were pro cameras though Nikon F? and the ?90 I think. Don't know my Nikon numbers that well.

Roger

[rongarrett]

Now with the Nikon D3x coming out it will really, really be interesting how it compares to the Sony A900 since they use the same sensor?

Just what controls resolving power of a sensor anyway? Can we expect the same resolving power from both these cameras?

Inquisitive minds want to know.

[rongarrett]

Interesting factoid: According to dpreview's resolution tests the new 50D, despite being 15 mp shows only 150 more lines resolution than the 10 mp 40D. Resolution isn't the whole story by any means, but it does mean the higher mp camera will not display more detail to any useful degree.

The same review says that the new 50D actually shows less high ISO performance.

I'll prolly be accused of negativity again, but the actual tests are far more accurate than manufacturer claims.

The hype continues.

[SmartWombat]

despite being 15 mp shows only 150 more lines resolution than the 10 mp

To double the resolution, you need 4x the number of pixels. So you'd need to go to 40MP !
You're also going from a cropped sensor to full frame, and so may be using a different part of the image for the measurement - I don't know how they determined it.
The crop factor of the smaller sensor only uses the centre (the good bit!) of the image.
So there's so many variables I can't be sure that comparison is meaningful.

DxO have a database of their sensor analysis.
It's interesting, but I don't understand how they normalise all the sensors to an 8MP print.
Even reading the small print in the technical bits of the site, I can't work it out.

[rongarrett]

Someone correct me if I'm wrong, but the 50D is an aps-c sensor isn't it? I was refering to comparing the resolution between the 40D and the 50D. Both are aps-c snesors and much of the controls are the same. The extra 5mp just don't add up to similar resolution gains. Supposedly the worse ISO performance by the 50D is because if the crowded sensor and smaller sensor size?

My point is that I am trying to say I sense a lot of questionable hype here lately put out by the camera manufacturers? Competition so hot they are fighting back any way they can? Dunno, just seems do to me.

I've just about talked myself into passing on all offerings this year in the hope things level out better next?

[AgingEyes]

A good lens will do a lot more for your images than any camera body.

Well, the camera body counts, too.

FWIW:

http://www.luminous-landscape.com/tu...solution.shtml

[rongarrett]

I'd like to qualify my statement above about 50D ISO performance. According to Pop Photo lab tests, the 50D actually shows more noise than the 40D up to ISO 800. Above ISO 800 the 50D slightly surpasses the 40D. I thought that was a little unusual.

[deckcadet]

That could be as simple as more noise reduction being performed by the camera processing at ISO 800 and above.

[Loupey]

I've just about talked myself into passing on all offerings this year in the hope things level out better next?

Like computers, if you're always waiting for the next big thing, you will always be waiting. Bite the bullet when you need to bite it. Cameras are so good these days (and yesterdays and tomorrows) that you shouldn't be disappointed.

Buy what makes sense to you and your needs. Don't worry about the rest.

[Loupey]

I'd like to qualify my statement above about 50D ISO performance. According to Pop Photo lab tests, the 50D actually shows more noise than the 40D up to ISO 800. Above ISO 800 the 50D slightly surpasses the 40D. I thought that was a little unusual.

I've used both the 40D and 50D extensively and I wouldn't say that the 50D shows more noise at the lower ISO's. If it does, one has to be splitting hairs and looking at every image at 200%.

I would say that the 50D keeps the noise progression somewhat linear as the ISO increases from 800 to 1600. Much like how the 40D had a more linear noise progression going from ISO 400 to 800 when compared to the 30D.

I would say that each successive model has improved the useable ISO up one stop. Now defining what a "useable ISO" is up to interpretation

[rongarrett]

My problem is I've bought a new camera every year for the past several years. What with the new cameras doing so much more now, it seems a lot harder to seperate the hype from actual performance in use.