Originally Posted by
dblevitan
This is true. However, the whole point of RAW is to fidget with the settings on a photo-by-photo basis to theoretically produce the best rendering. Automatic rendering is easy...but fixing a photograph is not. And thus the question for the OP to answer would be whether he/she wants to fiddle with photos or whether he/she just wants to print them.
shooting jpeg is certainly convenient. for instance, images can be quickly emailed or printed at a kiosk. also, depending on the camera, one can generally shoot more jpegs until the buffer in the camera fills. on the other hand, there is no need to fidget with raw images. they show up on screen based on automatic settings, adjustments can be made if desired, and then the images can be printed, emailed, put in a web gallery, etc. both raw and jpeg have their place. my point is simply that raw is no more work than jpeg is.
True, and I haven't looked into the performance of the D40 since I wouldn't really consider buying it for myself. I did want to point out that the base sensitivity is lower which implies a good likelihood of higher noise at comparable ISOs between a D40 and D40x/D60. Simply the fact that the sensor has 4 MP more in the same area implies it will have more noise unless the quality is better (which I doubt it is).
noise is simply a function of pixel size. higher megapixel counts with the same size sensor and the same sensor technology means smaller pixels and thus more noise. newer sensors are less noisy than older sensors, so direct comparisons can't always be made by pixel count alone. only if one pixel peeps at high iso settings is it possible to tell the difference. shoot both at iso 100, they'll be virtually indistinguishable.
I think he makes many very good points on his website. I do not subscribe to all his theories but what I do like about him is that he judges cameras based more on their use than on their specifications. I have not really seen any other sites that do this to the same extent that he does. Now, if I really followed everything he did, I would be shooting wide on everything, only using basic JPEG, and using a D40 with 18-200, none of which I'm doing. I think that unfortunately his style puts many people off from reading his website which is a shame since there are good pieces of information scattered around that are harder to find elsewhere.
there is some useful stuff on his site, but it's difficult to weed it out from the garbage, especially for someone who isn't intimately familiar with the technology and visiting his site to learn something. as he says, he does it for fun, not for accuracy.
Regarding the megapixels...read the NY times article as well.
i have read it. david pogue took photos with an unnamed 13 megapixel camera (most likely a canon 5d, but it's strange that he never stated what it was) and then downsampled the images for the lower resolution images. he did that to avoid any issue with which cameras he chose, but the problem is that downsampling a higher resolution image will produce a better quality image than simply using a lower resolution camera. the result also is very dependent on the downsampling algorithm. his test was more about photoshop techniques than of three different cameras with different megapixel counts.
the benefit of more megapixels is that one can print larger and/or crop without losing quality. the rule of thumb is 300 pixels per inch for a very sharp print and 200 pixels per inch for an acceptably sharp print. thus, a 6mp camera (3008x2000 pixels for the d40) is suitable for an 8x10-11x14 inch print, and a 10mp camera (3872x2592 for the d40x) is suitable for an 11x14-16x20. even pushing it a little further to 150 ppi can sometimes produce acceptable results. most people print 4x6 inch prints, if they print at all, or they view the images on screen. for that, even 6mp is overkill. nevertheless, extra megapixels are nice for cropping when one doesn't want to lug a big telephoto lens.