No way I can use a 100mm or longer lens and still get a bit of the sky and foreground in the picture. Backing up is not an option, as I am shooting from a lookout point on a hill. Anyhoo, the point is that a 21mp camera will give me more resolution and detail than a 12mp camera, everything else being equal, and that makes the 5D Mk II a better camera than the 5D, for landscape photography.
floris wrote:
You could use a 100mm, or a 135mm, or a 150mm... what's stopping you? It just means you'll be stitching more, which is a pain, but it will get you a lot more detail, and that's what you're after...
RDKirk wrote:
I wouldn't call it "pixel peeping," but for sure unless physically restrained, viewing audiences will get as close to a photograph as possible--up to reading distance--regardless of the size of the image.
The limiting factor is the resolving power of the eye. It can resolve fine detail at reading distance, but as viewing distance increases the fine detail will blur together and seem flat compared with an image with less detail but more contrast. So notwithstanding the inclination of people to get up close and pixel peep they can only see the entire image from a greater distance and as the viewing distance increases it is necessary to change the contrast and resolution of the image to maintain the same overall perceptual effect.
Painters realized that fact long ago. One of my favorite painters is John Singer Sargent who did larger than life portraits which have stunning clarity and detail when viewed from across the room. But up close at the gallery you find he actually applied most of the paint with broad strokes of palette knife and the sharpness was created with the very clever use of contrasting highlights much in the way unsharp masking creates the illusion of increased sharpness in a digital image.
The optimal situation is to have enough capture resolution to make up-sampling unnecessary. My first 2.1 MP digital, did good job with a 4 x6 print at 300 ppi. My 5MP Minolta D7Hi produced a decent 5x7. My 20D a comparable 8x10. But when I first got my DC290 I did some experimenting making a 24 x 30 print two different ways. First I kept the file size the same pixel dimensions and lowered output resolution. For comparison I rescaled the image by resampling to the same 300ppi resolution. Its a big like a bargain with the devil in that neither produced an optimal result, but I found that the preserving the original file detail, plus optimal sharpening, produced a better looking image perceptually than the resampled version. LIke the Singer paintings the one with the "bigger dots" with more contrast fooled my eye into thinking it was sharper. Sharpening the resampled file didn't have the same effect because the resampling had diluted the original image detail. Its an interesting experiment which shows the role human perception plays in the process. Try it some time
Now when I edit a photo I save an unsharpened, un-cropped version as a TIFF as the "master" then open, crop and resize before sharpening. The bigger the print, the more USM I apply to maintain the same perceptual image quality. Up close a large print will look over-sharpened but if you was to hold a 4x6 copy at reading distance and then look at the 11x14 with higher USM from several feet away they look about the same perceptually.
bobbytan wrote:
No way I can use a 100mm or longer lens and still get a bit of the sky and foreground in the picture. Backing up is not an option, as I am shooting from a lookout point on a hill. Anyhoo, the point is that a 21mp camera will give me more resolution and detail than a 12mp camera, everything else being equal, and that makes the 5D Mk II a better camera than the 5D, for landscape photography.
But you're shooting a panorama so you can use whatever focal length you want. The longer the focal length the more images you'll need, and yes, you might need to do a multi-row pano, but that hasn't stopped people in the past.
I think the point that was being made is that while 21 MP will certainly be better than the 12 on a 5D (I plan to upgrade), it won't compare to a big pano shot with a 5D, or 20D for that matter. If you shot a multi row pano you could easily break 100MP, which is a heck of a lot more than 21... all you need is a multi row pano head, some decent software, and a relatively powerful computer (if you don't want to wait forever). Of course the problem with panos is that if things are moving then it's pretty tough if not impossible to get it to look right. That (and ease of use) is where the high res sensors are required.
I am totally unwilling to go through all that trouble, nor do I want to create such massive files. I shoot strictly for fun and don't sell my images. So the 5D Mk II would be great for my purpose - shooting and stitching simple multi-frame panos just for kicks.
floris wrote:
But you're shooting a panorama so you can use whatever focal length you want. The longer the focal length the more images you'll need, and yes, you might need to do a multi-row pano, but that hasn't stopped people in the past.
I think the point that was being made is that while 21 MP will certainly be better than the 12 on a 5D (I plan to upgrade), it won't compare to a big pano shot with a 5D, or 20D for that matter. If you shot a multi row pano you could easily break 100MP, which is a heck of a lot more than 21... all you need is a multi row pano head, some decent software, and a relatively powerful computer (if you don't want to wait forever). Of course the problem with panos is that if things are moving then it's pretty tough if not impossible to get it to look right. That (and ease of use) is where the high res sensors are required.
cgardner wrote:
The limiting factor is the resolving power of the eye. It can resolve fine detail at reading distance, but as viewing distance increases the fine detail will blur together and seem flat compared with an image with less detail but more contrast. So notwithstanding the inclination of people to get up close and pixel peep they can only see the entire image from a greater distance and as the viewing distance increases it is necessary to change the contrast and resolution of the image to maintain the same overall perceptual effect.
Painters realized that fact long ago. One of my favorite painters is John Singer Sargent who did larger than life portraits which have stunning clarity and detail when viewed from across the room. But up close at the gallery you find he actually applied most of the paint with broad strokes of palette knife and the sharpness was created with the very clever use of contrasting highlights much in the way unsharp masking creates the illusion of increased sharpness in a digital image.
The optimal situation is to have enough capture resolution to make up-sampling unnecessary. My first 2.1 MP digital, did good job with a 4 x6 print at 300 ppi. My 5MP Minolta D7Hi produced a decent 5x7. My 20D a comparable 8x10. But when I first got my DC290 I did some experimenting making a 24 x 30 print two different ways. First I kept the file size the same pixel dimensions and lowered output resolution. For comparison I rescaled the image by resampling to the same 300ppi resolution. Its a big like a bargain with the devil in that neither produced an optimal result, but I found that the preserving the original file detail, plus optimal sharpening, produced a better looking image perceptually than the resampled version. LIke the Singer paintings the one with the "bigger dots" with more contrast fooled my eye into thinking it was sharper. Sharpening the resampled file didn't have the same effect because the resampling had diluted the original image detail. Its an interesting experiment which shows the role human perception plays in the process. Try it some time
Now when I edit a photo I save an unsharpened, un-cropped version as a TIFF as the "master" then open, crop and resize before sharpening. The bigger the print, the more USM I apply to maintain the same perceptual image quality. Up close a large print will look over-sharpened but if you was to hold a 4x6 copy at reading distance and then look at the 11x14 with higher USM from several feet away they look about the same perceptually.
But remember that when talking about painting, we are talking about a totally different medium. When the Pointillists attempted to copy photography with their technique, you couldn't "see" the image close up either. But film those days had large grain, even in 4 x 5 film, which was about as small as one got.
But photography evolved rapidly. As photographs became better, artists also began to copy the "perfection" with ever increasing realism in their own work. At some point, they realized it was a hopeless quest, and Impressionism and modernism were two results.
Basically, people don't expect paintings to be perfect. They expect photos to be perfect.
This is the reason why the 5D mkII has almost two times as many pixels than the older 5D.
Some photographers will use pixillation as I and others used grain many years ago. But, most want to eliminate all artifacts of the photographic process, and end up with a file they can manipulate the way they want, not limited by the defects they have little or no control over.
The job of the hardware and software companies is to give us products that, at the time of introduction, come as close to that ideal as possible, and let US decide how we wish to use it.
photogenix wrote:
Anyhow I am starting to wonder, now this is supposed to be 3 weeks away in theory, still without a pricetag, whether Australia will receive any in the one-and-only 2008 shipment at all.
I had a call about an hour ago from Michaels saying that although Canon still haven't announced the RRP yet, they're still expecting the delivery to be late November.
think theyre waiting for the dollar(s) to settle. In NZ the price most retailers were told was NZ4300 but that was before the dollar dropped recently. If it stays at 4300 that'll be $2590 USD which Id be very happy to pay. In Aus thats $3760 at the moment.
Weird they havent locked it in yet so close to shipping date though....
Mel Gross wrote:
Basically, people don't expect paintings to be perfect. They expect photos to be perfect.
Yes, but my point was that resolution which makes an image perfect both in the technical and perceptual senses at 18 inches will not look "perfect" or the same perceptually at a distance of 18 feet. In other words there is a disconnect between perfection in the technical sense of pixel peeping at resolution charts or measurement and what its judged to look best perceptually by eye.
I think digital photographers, in general, get so obsessed with technical perfection that they lose sight of the perceptual aspects and the fact the only thing that matters is is the illusion created by the gray matter behind the eyeballs. Making an effective facsimile image of reality with technology requires an understanding of human perception. Much of the underlying technical magic based on human perception --- like the fact equal amounts of red, green, and blue light cause us to see white light, or CYMK produces a similar image on paper --- is intuitively obvious. Other aspects such as how perceived image sharpness changes with image resolution, contrast, and distance are less so and need to be verified with testing to be believed.
Basing reproduction parameters on the assumption the photo has to look perfect at 18 inches is a good one if printing a 4in x 6in print for a scrap book but rather myopic in the non-optical sense if printing a 4 ft x 6ft display for a store window. If the most effective way to reproduce an image on a billboard sign in Times Square was to print it a 600 ppi output resolution common sense tells us that the billboard industry would have figured out a way to do that by now. The fact a printed billboard may have dots the size of golf balls isn't due to an inability to print with greater resolution, but because the image is perceived more normally from street level when reproduced that way.
So for any intended viewing distance for a photo there will be an optimal balance between up-sampling to print at higher resolution or keeping the native capture resolution and making the "pixels" bigger and spaced further apart. Because the variable in the equation is human perception the only way to really know what that optimal balance is would be to make test prints and look at them from the intended viewing distance.
What is the optimal viewing distance to plan for? The distance at which the entire image fits into field of view. Ergo the bigger the print gets, the further away a person will stand to look at it. If they move closer and it appears to become pixelated brain will tell feet "Hey, it doesn't look as good up here!" and feet will move eyes back to the optimal perceptual viewing distance. Even that distance may vary from person-to-person depending on their eye sight. Next time you go to an art gallery observe how far people stand initially to view different size images and you will that field of view / viewing distance dynamic in action. Its why big painting are hung in big rooms and small ones in smaller ones
AUD.
It will almost def be higher than that though, due to exchange rate changes.
I have no idea if Canon AU has told its dealers anything but Canon NZ estimated NZD4300 when the camera was announced. (you can do math from there, NZ and AU will probably lie fairly similar in their values)
I very much doubt it will be above AUD4000-4100 though.
Printer
Wich is the best Printhead dpi settings for your printer.
Some of them will print better in 600 then in 300 dpi modes, with others you really cant see a big difference.
Color:
More important is that printer will not posterize, specialy from light dark to the darker areas. Needs more ink color above 6 to 12 (better) or 16 bit printing from also 16 bit pictures!
File size:
In photoshop you can see View: actual pixel size. There you can check with softproofing if the sharping is ok or not, also check for some jaggies. Then step back from your Monitor to see. The dpi from the file is not the same as the Printheadpixel density!
Media:
If I print on Canvas( Sihl Lisa J45) the dots from my 600x 600 dpi printhead resolution is completely absorbed. If I print on 3927 matte paper I can see the grain.
With my Kigamo 6000 pixel trilinear array ( no bayer) resolution 6,000 x 8,488 pixels File size in 24-bit RGB 145.7MB i can print 54 x 118 inch and nobody is complaining with pixel peeping.
If you print really big like 54 inch you need a lot of RAM and a fast RIP or highspeed Harddiscs for spooling. Pictures with 54 inch dont need 600 dpi files, my Designjet 3800 needs only 150. Bigger settings than 150 will be trown overboard.
In a printed magazine nobody sticks his nose so close to see the srceening, but with your own prints they do so, stupid!
I was just at my local camera store. They said they just yesterday got definite pricing from Canon and they are still not sure when the first batch will arrive.
I have a piece of good news concerning the LP-E6 battery for the new 5D Mk II. As you know, currently you can only order the original Canon battery, and it cost $100 each .... yes, Ouch! Well, Sterling Tek has just informed me that their LP-E6 battery will be available in 3-4 weeks. Yay! I am sure their battery will retain all the functionality of the original Canon battery - but I cannot guarantee you that.
