![]() ![]() I didn't apply any sharpening to the test images, to avoid obscuring the noise reduction effects. The curves adjustment had the effect of enhancing the noise, which makes it easier to see differences between the three test images in each comparison image. This adjustment was applied equally to all three test images, as they were combined into a single comparison image at this point. I then adjusted contrast and brightness using curves, in the same manner as I would typically adjust an image. I combined 300 x 300 pixel crops from each image into a single comparison image. Because I used the default noise profiles, this is to some extent a test of the supplied noise profiles, as well as a test of the software. Only for "special" images will I take the time to tweak the settings. For 98% of my images, I want to be able to apply the noise reduction filter automatically, without taking time for adjustments. I recognize that I could probably get better results by tweaking the noise reduction settings for each image, but that's not how I intend to use the software. Noise reduction was applied using the default image profiles with no tweaking. For this test, I used Noise Ninja 1.1.2 Beta, and Neat Image 3.0 Demo (the current versions as of January 1, 2004). All images are from a Canon G2 at ISO 400, and were converted from RAW format with the Photoshop CS RAW converter. The first test image in each comparison image is the original image, the second test image (the one with the watermark lines) is Noise Ninja, and the third test image is Neat Image. If that's the case, then it illustrates the importance of retaining embedded colour profiles.īelow are the seven images I used for comparison. The Neat Image images contained no profiles, which could be the reason for the duller images. Noise Ninja not only preserves colours, but also preserves embedded colour profiles. The colours in the Neat Image images are duller than in the original images and the Noise Ninja images. Neat Image's results look a liitle too smooth and "not real" for my taste. To my eye, Noise Ninja preserves more detail and texture. Noise Ninja filtered a 16-bit 4 megapixel image in 26 seconds, compared with 50 seconds for Neat Image to do the same image in 8-bit mode (unfortuantely, in the Neat Image demo support for 16-bit images has been disabled, so I couldn't test it on a 16-bit image). I also liked the ability to navigate in a 100% view of the image by holding down the Ctrl key to change the cursor to the hand for panning. ![]() Although the interfaces are similar, I found the filtering adjustments in Noise Ninja easier to understand and use, partly because Noise Ninja's help file is better-written. But I picked Noise Ninja for several reasons: To be honest, I'd be happy with either program. I tested the demo versions of each on several ISO 400 images from my Canon G2. I need to buy noise reduction software, so I tested Noise Ninja ( ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |