bushwacker wrote:
oh by the way... since we are talking about this... you posted that the 20d\30d has the same pixel size 1DsIII and 5D II --- from this we can safely say that image sharpness from 20D\30D is much greater [ compared to new - higher mega-pixel EOS bodies ] because it has more usable lens f-stop without hitting diffraction.?
Depends on the lens and the way you define sharpness. If the lens is sharper than what is required for smaller pixels, then there is no difference, if you dont stop too much.
However, when you print (to a fixed size), then more pixels mean more sharpness when the print size is large (and when the lens keeps up).
For inspection at 100% pixel size, the larger pixels are sharper if the lens cannot keep up with the smaller pixels. But the prints would still be equal.
Even when you want to stop for more depth of field, there is no difference in prints. Though, you can consider that an advantage to the camera with larger pixels - the file size is smaller but the result is still equal.
If you compare the 100 ISO samples posted at Imaging Resource, the 50D clearly outresolves the 12Mp bodies, including the Nikon D3/D700. High ISO samples (NR off) look pretty good up to 1600, and even 3200 is proabaly usable with some NR in post.
It's true these are in-camera JPEGs, but I'm sure that RAW files would retain more, not less, sharpness.
Yes, there may be diffraction limts and more apparent issues with some lenses, but I'm not sure what more people could want from this camera at this price point...
Next time you are selling a photo to someone taken with your 50D ask them if they can tell how many lines per mm they are getting and if it makes a difference to them. They will look at you and the picture and then think "This person is losing his mind" and then if the picture looks good most likely still buy it.
Honestly, I am surprised people are still debating this topic of diffraction even after so many threads about it. Sensors have ZERO effect on diffraction. A high pixel density sensor merely serves to tell the honest truth about what has happened AFTER lens diffraction has taken place. If there is no detail to be gleaned, then there's nothing to see. A low pixel density sensor merely gives the illusion that everything appears sharp 'cos it is incapable of showing the truth.
n0b0 wrote:
Here's a 100% crop image with effective f-stop of 20.
And here's one with effective f-stop of 48 though not a 100% crop.
In the first case you can clearly see the softening. For the second, hard to say as it is not 100% pixels. However, when you don't need 100% pixel sharpness, stopping down past the diffraction limit is not a problem. But when you need the utmost quality, it is good to keep the limit in mind.
astrolucida wrote:
In the first case you can clearly see the softening. For the second, hard to say as it is not 100% pixels. However, when you don't need 100% pixel sharpness, stopping down past the diffraction limit is not a problem. But when you need the utmost quality, it is good to keep the limit in mind.
The first image didn't receive any sharpening or any other post processing at all though and I think it's perfectly acceptable as it is, unless you're pixel peeping. A little PP and sampling down to 10MP and I'm confident I can get a sharper image than the one taken with a 40D at 100%.
There is no such thing as "the diffraction limit", it's not like the image stays sharp up to a certain f-stop and just get blurry afterwards. Diffraction increases gradually as we stop down. It's up to the shooter to decide at what point it is acceptable. Obviously pixel peepers will be limited by their own misconception much earlier than others.
n0b0 wrote:
There is no such thing as "the diffraction limit", it's not like the image stays sharp up to a certain f-stop and just get blurry afterwards. Diffraction increases gradually as we stop down. It's up to the shooter to decide at what point it is acceptable. Obviously pixel peepers will be limited by their own misconception much earlier than others.
I agree. However, apparent sharpness and resolution are two different things. You can increase apparent sharpness by sharpening in post-processing but you cannot increase actual resolution. I am not talking about the number of pixels but the amount of actual subject detail.
Whether you need the pixel-level detail, depends on what you do with the image. Low level of detail in the image limits how large you can print and still have enough detail.
Also, if your pixels do not contain any real information, why store the files with the "empty" (=devoid of real detail) pixels? In that case, using a smaller resolution in the camera would produce the same results when printing. In your first example, Medium size would capture all the same detail as Large does but in a file with a size 1/2th of the large one. If you need more pixels, you can uprez the file in Photoshop.
Edit: I performed an experiment with this file and it seems that 50% size dropped some actual detail while 75% size did not. My test was done by first downsizing the image to 75% and the uprezzing that file to 133% to regain the original pixel dimensions. I then flip-compared that file with the original. It was hard to get exactly the same sharpness - my first attempt actually looked better than the original!
With very sharp originals, even 200% size looks good, so there is some potential that the camera has but that gets lost through diffraction. Whether you really need that much resolution, is another question, of course. However, consider it as a digital 2x teleconverter with no loss of light, and you'll see the benefits, even to non- pixel peepers.
- Per-pixel detail not as good as on good 10 or 12 megapixel cameras
- High-end lenses required to get the most out of the camera
The first point is probably because DPR didn't use a good enough lens.
The second point suggests that the DPR reviewer knew how to do an appropriate review, but decided not to.
While not quite as sharp at the pixel level, the 50D is still pulling more detail out of the scene. The fabric example above show this. As you drag your mouse pointer up and down over the 50D and 40D labels, watch the tiny threads appear in the 50D image that are not present in the 40D image.
========================================
oh yeah the guy used a Canon EF 200mm f/2 L IS USM... well let's see the reults..
also about astrolucida's comments about diffractions--- the guy [ at digital-picture.com ] also created table for each EOS body's diffraction limits..
James B wrote:
I am having great success shooting high iso .jpgs with my 50d with saturation, sharpness zeroed and contrast at 1. The jpgs look a lot better than what I have been getting from lightroom.
After shooting with a new 50D this past weekend, I have the same experience. I'm a bit embarassed that my post-processing workflow isn't nearly as solid as I thought it was! The .jpegs are super sharp and rich with colour, I might actually start using auto mode for family snapshots I'm so impressed.
One of the guys in the store recommended I drop ACR 5.2 and use the DPP tool, as he felt it did a better job with the Canon output. I'll start working with it this week accordingly.
you just made me laugh out loud at work.