Renders out of focus when cropped?
intrinsicanomaly_262dc77de2
Posts: 132
in New Users
I have just discovered that if I try to crop a saved render, it becomes out of focus. I'm really surprised by this, is it normal?
Comments
Not generally, no.
However, there is a "crop" tool in Photoshop, (I have CS6, it may have changed since then,) and this tool allows me to set the size I want, then position it. When I apply the "crop", it cuts the unwanted pixels off the edges and gives me the size I specified. However, if the area I am targeting is actually smaller than the size I want, it will process the image and make it larger. That can result in an image becoming blurry, depending on how much the image needs to be sized up.
I'm not familiar with other photo editing software, but many of them try to emulate Photoshop. Could this be happening in your program?
Yes, it is standard for cropping tools to cause the image to become larger absolutely. The amount you can crop and maintain image quality is one way of measuring the quality of your camera. I just can't understand how it woudl be that an artificially produced image would not be able to do this I would have thought you could crop as much as you wanted and have it stay sharp but it seems quite the opposite, it can't be cropped even a tiny bit without destroying the image.
In my experience with Photoshop, (started when 4.0 was the flagship. That's 4.0, not CS4.) sizing an image larger is going to lessen the image quality. Sizing algorithms are much better now, but you still have code extrapolating what should go where when you're asking the program to create pixels that weren't there in the first place.
For example, if you have an image that's 500 pixels square, crop it to 400 pixels and then ask the program to size it up to 1200 pixels square, the program is going to put 9 pixels where the original had only 1 pixel. The image can't help but be "out of focus".
On the other hand, if you're starting with an image that's 3600 pixels square, crop to 3500 pixels square and size it back to 3600 pixels square, you're going to get a much better result. There is more data for the algoritm to extrapolate from, and the increase is actually fairly small.
I now use Photoshop CS6, (not a fan of the subscription only model,) and I still don't like the results of increasing image size most of the time.
Reducing the size works far better. In fact, when Iray was first introduced, one of the tricks to get better results faster was to render the image twice the size for half the time, and then use the image editor to size the image by half.