Noise Ninja vs. Neat Image?

[gesper]

Advertisement (gone after registration)

What is the current view of forum members on noise reduction software?

[carstenw]

Don't forget DFine from Nik Software, which seems to be getting good reviews. In the past, Noise Ninja was ahead, then Neat Image was ahead, then... They are both meant to be very good, and from release to release, they swap order, so I wouldn't worry too much, and buy the one you like.

[gesper]

Thanks Carsten. I forgot about Dfine. I like Nik Sharpener, so I will add this to my list.

[Guest M8-Freak]

I use Noisware Professional.

I am very happy with it.

[jrgeoffrion]

What is the current view of forum members on noise reduction software?

 

NoiseNinja has a slight edge over NeatImage. It can be seen mostly in the very dark and uniform areas where NeatImage will leave very low frequency bands.

[leicaman94044]

Andrew Rodney reviews Noiseware here: Review: Imagenomic Noiseware (Bonus)

 

Scroll down to the images of the white dog... where he writes "Photo: A before and after of an image shot at ISO 1600. Notice the detail in the dogs hair in both. (Click image for larger view.)"

 

I clicked on the image, looked at the detail in the dog's hair and there is none in the image treated with Noiseware. Grain is entirely absent from the treated image... but after showing this example to several friends, all agree that the dog looks better in the image not treated with Noiseware.

 

Rodney is infinitely more knowledgeable in this area than I am, and he cites Greg Gorman as the person who turned him on to Noiseware. Both are very credible in my book. I just don't understand why he would chose an example that illustrates the opposite of what he alleges in his review.

 

Until I saw this example, I was ready to order Noiseware. Now I strongly feel that more research is in order.

[gesper]

Advertisement (gone after registration)

Several reviewers cited DFine 2.0's retention of detail versus the competitors. I am thinking of doing the 30 day trial.

[gesper]

Andrew Rodney reviews Noiseware here: Review: Imagenomic Noiseware (Bonus)

 

Scroll down to the images of the white dog... where he writes "Photo: A before and after of an image shot at ISO 1600. Notice the detail in the dogs hair in both. (Click image for larger view.)"

 

I clicked on the image, looked at the detail in the dog's hair and there is none in the image treated with Noiseware. Grain is entirely absent from the treated image... but after showing this example to several friends, all agree that the dog looks better in the image not treated with Noiseware.

 

Rodney is infinitely more knowledgeable in this area than I am, and he cites Greg Gorman as the person who turned him on to Noiseware. Both are very credible in my book. I just don't understand why he would chose an example that illustrates the opposite of what he alleges in his review.

 

Until I saw this example, I was ready to order Noiseware. Now I strongly feel that more research is in order.

 

 

I thought both of his examples were bad and showed unacceptable loss of detail (which wasnt great to begin with). Also, this article is sixteen months old, so not sure how relevant it still is.

[leicaman94044]

Gesper,

"Also, this article is sixteen months old, so not sure how relevant it still is."

 

The article, as you correctly state, is sixteen months old. On the other hand, the only changes to the version reviewed in AR's article are the following:

 

- Added Windows Vista support

- Added Photoshop CS3 support

- Added Universal Binary for Mac OS X

- Memory related fixes

 

None of these would suggest that the software (with respect to noise reduction effectiveness) has been improved relative to the reviewed version. In addition, the fact that Imagenomic currently refers to AR's review on their website would suggest that they are in agreement with his assessment. That would indicate to me, at least, that Imagenomic views the article as relevant.

 

I agree wholeheartedly with your observation that both examples were bad and showed unacceptable loss of detail. If as is said, the devil is in the details... I'm not impressed.

[jrgeoffrion]

Andrew Rodney reviews Noiseware here: Review: Imagenomic Noiseware (Bonus)

 

Scroll down to the images of the white dog... where he writes "Photo: A before and after of an image shot at ISO 1600. Notice the detail in the dogs hair in both. (Click image for larger view.)"

 

I clicked on the image, looked at the detail in the dog's hair and there is none in the image treated with Noiseware. Grain is entirely absent from the treated image... but after showing this example to several friends, all agree that the dog looks better in the image not treated with Noiseware.

 

Rodney is infinitely more knowledgeable in this area than I am, and he cites Greg Gorman as the person who turned him on to Noiseware. Both are very credible in my book. I just don't understand why he would chose an example that illustrates the opposite of what he alleges in his review.

 

Until I saw this example, I was ready to order Noiseware. Now I strongly feel that more research is in order.

 

The image shown in the above article has been overly smoothed. In other words, there is too much luma noised removed.

 

Denoising is part art part science. The trick is to remove MOSTLY chroma noise and leave the luminance noise. Yes, the image will still be "grainy", but that's exactly it, "grainy", not "noisy". With both NoiseNinja and NeatImage, the user can control the amount of chroma and luma noised removed independently of each other.