0

I have been reading review of the Canon EOS 1300D by The Digital Picture and I came across noise reduction.

The color blocks in this tool, being evenly-colored, provide a worst case scenario for noise levels. The samples provided for the Rebel T6 have no noise reduction applied, a key piece of information. Noise reduction can be applied in-camera (it is applied by default) or via software during post processing. The downside is that some detail is generally sacrificed with noise reduction application. The site's standard color block samples show what the camera is capable of without noise reduction.

Snip

As mentioned, in-camera noise reduction is always available in EOS DSLRs, and noise reduction is also available during post processing. Noise reduction is very effective at reducing visible noise, but it is also destructive to the details in your image and to image sharpness. Even at ISO 100, the standard out-of-the-box noise reduction will have an effect on your image quality. I usually apply light NR only to my very high ISO images.

How does Noise Reduction work in this camera? As far as I know its done by taking ten photos and blending them together.

  • 1
    1300D - That would be the Canon EOS Rebel T6 / 1300D – Alaska Man Jul 24 at 19:45
  • 1
    "As far as I know its done by taking ten photos and blending them together." That's not how it is done. – Michael C Jul 25 at 5:41
  • @MichaelC then what is that technique done for? – Delta Oscar Uniform Jul 25 at 8:42
  • 1
  • 1
    @JonathanIrons That technique is one of many ways to produce a final image with less noise than would otherwise be the case. It's not how Canon's (or any other DSLR maker's) noise reduction works for normal, single frame shooting. – Michael C Jul 25 at 14:13
1

The classical solution for edge-preserving noise reduction is to apply a bilateral filter: https://en.wikipedia.org/wiki/Bilateral_filter .

This calculates the average of a group of pixels in the immediate environment, but it ignores the pixels whose difference is too large, like on the other side of an edge. Thus it filters away the differences that are small enough to be attributable to noise, but not large enough to represent important details.

As noise is mostly visible on large flat areas, this should work well. The problem is with fine textures, i.e. weak details that fall below the noise floor, those will be lost. This gives artefacts like "plastic faces".

Noise reduction on such weak details can be done with temporal filtering, i.e. averaging between successive frames, but then you'll have problems with motion smear.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.