PDA

View Full Version : Same NR Program but getting Different Results



mongo
06-11-2010, 10:06pm
While using Neat Image noise reduction program, Mongo was very surprised to get different results depending on how it was used.

WAIT – read on! There is more to this than you might expect.

Mongo found that using the same image through Neat Image program on its own (even at only 53% noise profile compatibility) gave noticeably better results THAN using Neat Image as a plug-in while in Photoshop CS3 at 64% noise profile compatibility !!!

This was no accident. Mongo just happen to notice one batch done one way on one day was noticeably different to another batch done the other way at another time. This made him carry out an experiment using the same original image (which was duplicated) for the tests. The results confirm what Mongo has said and found above.

Mongo is surprised and puzzled by these findings. Has anyone else noticed this either with Neat Image or any other form of program when used on its own as opposed to when it is used as a plug-in in conjunction with another program e.g. with Photoshop. Mongo needs to start understanding and taking note of what is going on???

Mongo has posed the original (no noise reduction), the Neat Image standalone noise reduction image and the Neat Image plug-in to photoshop image. Unfortunately, the difference between the last 2 images may not be so easily seen on these posted internet images. However, if you use Neat Image or similar programs as plug-ins, Mongo urges you to try the same experiment at home.

1. original (no noise reduction)
61511

2. Neat Image standalone noise reduction image
61512

3. Neat Image plug-in to photoshop image
61510

kiwi
06-11-2010, 11:29pm
In Photoshop there's no opacity on the layer is there ?

mongo
07-11-2010, 9:06am
In Photoshop there's no opacity on the layer is there ?

No - Kiwi.

Also just noticed that when you open un Neat Image as a stand alone it is automatically set to "RGB". For some reason, in Photoshop it has been set to "YCrCb" instead. That might explain the difference. Mongo is going to re-do the experiment with this in mind.

Kiwi, Mongo appreciates your interest in replying to this post and he will post his results later today just in case it was only a Mongo usage error