Looking through my kit, I find one (Canon 50mm prime) with only 22 marks, an old Sigma 35mm prime with 8, 11 and 22. Otherwise, 7 others have nothing.

For your point 2, what you really want to consider is not that the sunset is at infinity, but what if anything in the foreground you want to have acceptably sharp.

Say you're shooting a sunset from the beach with a 50mm lens. The shoreline is in the bottom of your frame and is about 25 feet away. If you would like to capture it as well as the sunset, will they both be in the range of acceptable focus? You can pull out a smartphone with a DOF calculator app, or if you have a good feeling for the lens/camera you're using, you might know that a 50mm focal length at f/8 and a 1.5 or 1.6 crop sensor (i.e., Nikon or Canon but not full frame) the hyperfocal is about 50 feet so anything from 25 feet to infinity will be in acceptable focus. You're all set -- just focus on something about 50 feet away (easy to say). If there are some sea oats you want in focus that are 12 feet away, you may have to close down the aperture. If you set the aperture to f/16 (double the f-number) in this case you will halve the hyperfocal distance and anything that is 12 feet or more away will be acceptably sharp if you focus on something that is 25 feet away (the shoreline from before).

Hyperfocal distance is the point (determined by depth of field calculation) you focus on at which everything from that point to infinity is in sharp focus, and from that point halfway back to the camera is in sharp focus.

Find a better definition with the '1000 words' picture here:

http://www.dofmaster.com/hyperfocal.html

I think the back button AE lock will hold the exposure setting until the shot is taken or the camera 'gives up' and displays go blank. Until then you can use the half press shutter button to set the focus to wherever you like, including the hyperfocal point. You can confirm this by watching the exposure values while you do multiple half-presses of the shutter while aiming at different brightness areas. If AE lock is set to really 'lock', the aperture and shutter speed will not change.

Totally agree with Wanda - both the papers and supplier. Great looking prints.

I had one lens which took a bump and could no longer autofocus at infinity although it could be manually focused there. The autofocus would go crazy trying to focus on anything more than about 300 feet. The camera shop tech explained how the autofocus works and why the overshoot has to be there. He said every lens goes slightly beyond infinity focus. This allows autofocus to work and allows for thermal and manufacturing tolerances. The repair was quite simple -- something like changing a set screw position.

With a bit more study of the equations and physics, the resolution limit of the pixel size is the lens f-stop and the color. With a very good lens at f/4 and red light (700 nm) the smallest useful pixel dimension is about 3.4 µm which puts the full frame sensor at 75MP max. If the lens is stopped to a larger aperture, there could be even more pixels (smaller pixels). So as lenses get better the sensors can have more useful MPs. However, a not-so-good lens will make the higher number of pixels ineffective as the optical quality of the lens will limit resolution.

More resources...

http://theonlinephotographer.typepad.com/the_online_photographer/2008/11/diffractionquan.html
http://en.wikipedia.org/wiki/Airy_disk
http://snapsort.com/learn/sensor/true-resolution

Very good question -- time for some more study. The 36MP does go beyond the theory.

I've had a Nikon Coolscan V for 7 years and (1) I paid $495 in 2005; (2) manual focus is only required if the film strip is not flat -- which in a flatbed causes blurry scans and is not correctable; (3) it has native 4000 dpi, 12 bit color sensor which gives amazing results.

Buried deep in some theory there is a physical limit to the size of the photoreceptor in the sensor, related to the wavelength of light.

From Pixiq "Circles of Confusion: The answer to the question 'why?' in some basic matters of photography"

http://www.pixiq.com/article/circles-of-confusion

"...Which implies that if a sensor site is made any smaller than 5.4 μm it does not increase resolution.

"Eg. For a Nikon D3 sensor sites are 8.45 μm across and for the D3x they are 5.94 μm and for the D300 5.54 μm getting close to a theoretical maximum resolution.

"This is a good first approximation that gives a realistic idea of limits. In the world of sensors such things as pitch - the distance apart of sensor sites (pixel) centres is also important for and smaller sensors cram them in. This affects signal to noise ratio - also important when comparing sensor sizes and sensor site (pixel) sizes.

"**There are limits to how small sensor sites can be and offer higher resolution– not just the limits of manufacture but the limits of optics."

This implies that for a full frame sensor, the maximum useful size would be 30MP. To have more pixels you would need to go to a larger sensor (medium format digital back).

I'm not so sure about the physics here. A full frame sensor (36 x 24 mm) with 35 MP has about 42000 pixels per square millimeter. A 1.5 crop sensor (24 x 16 mm) with 18 MP has about 47,000 pixels per square millimeter. Each pixel in the latter 18MP camera covers a smaller viewing angle and thus has higher propensity to lose sharpness with movement.

Yes/no?

I got a package of cheap, double sided velcro from Home Depot. Has about 100 8" strips in a roll but they can be chained together to make a length long enough to go around.

I have enlarged a 10MP shot to a 48 x 72 print (in 6 frames). You have to assume that no one will be looking at the photo from closer than 3 feet and accept some visible artifacts if you get closer. I do the enlargement in Photoshop with a trick that was in a Scott Kelby book that he attributes to Vincent Versace. You continually increase the image size by just 10% using Bicubic. Then for the last 10% use Bicubic Sharper. I target for a print of 240ppi at whatever the final size will be.

The file will be large. You can save as JPG to save some disk space, but don't edit it again afterward. Also, normal sharpening practices are out for this kind of enlargement unless you want to spend a lot of time tweaking or reading about the physics of sharpening.

The last thing is then printing. A full 28 x 44 you will be expensive. If you can figure out a way to matte to something less than 23.5 inches wide it will be much cheaper to print, but the matte will be expensive.

I generally agree with cleaning only as required, but they do get more than just dust bunnies. They can get a layer of organic material (plasticizer, oil, lens cleaning solvent, etc) that needs to be wiped off with a cleaning agent. Mine get cleaned yearly minimum and before any big shoot. Inspection with a loupe shows the dust, but not the organic coating.

To the original post, the story from the shop sounds like BS as others have mentioned. Only if something heavy got in while changing a lens and then whacked the IR filter could you get a scratch. But then it wouldn't 'grow'.

Lightroom saves all the steps, but Photoshop CSx does not. However, Photoshop CSx does save a lot of data for each layer, mask, selection, etc. If you have an 18 megapixel photo (like Canon T3i) it may take only 3 MB of data as a JPG. When just loaded into PS as 8-bit mode, then saved as PSD (or TIFF), it will take about 18x3=54MB without doing anything. If you happen to load it as 16-bits per channel it will be about 100MB. If you add layers or masks it goes up even more.