Olympus-OM
[Top] [All Lists]

Re: [OM] Another NR datapoint

Subject: Re: [OM] Another NR datapoint
From: Tina Manley <tmanley@xxxxxxxxx>
Date: Tue, 12 Nov 2019 08:03:52 -0500
I prefer the grainy version, too.  I do have Neat Image and Topaz DeNoise
and Nik's noise remover but I don't use them much.  If I do, I fade the
effect by about 75%.  I just like grain with film images.

The blue tint to the photo is air pollution which was terrible in Berlin in
1989.  That's really what it looked like.

Thanks!

Tina

On Tue, Nov 12, 2019 at 7:16 AM Wayne Shumaker <om3ti@xxxxxxxx> wrote:

> The wall is a great shot. I do prefer Tina's original grainy version. It
> has more wabi-sabi. I hope she continues with her project, embracing any
> imperfection rather than rejecting it.
>
> At 11/11/2019 08:29 PM, Moose wrote:
> >When I saw Tina's photo of the Berlin Wall, I was interested in the
> noise. This is in no way a comment on her decision to keep the noise - but
> it's such an interesting case that I wanted to look into it.
> >
> >First of all, the noise in the sky increases the higher it goes. I don't
> know if that is down to the film or to scanning software.
> >
> >I've been quite interested in Topaz Denoise AI. I've seen it perform
> miracles - and fail.
> >
> >Here, it does a great job on the wall, the black coat, the monument/gate
> and the lower part of the sky, but does almost nothing to the top. <
> http://www.moosemystic.net/Gallery/tech/Process/NR/Wallnoise.htm>
> >
> >I tried heavier settings, to little avail. So I tried creaky old
> Neatimage 4, at settings that seriously blur detail, but clean up the sky
> nicely, and applied it as a mask to sky only.
>
> Non-comment noted...
>
> I found that Denoise AI starts to fail when the noise becomes too much. As
> if it starts to think the noise is important detail and may even enhance
> it. With most digital noise, the noise is at the pixel level. Film has
> grain, noise can increase in size with grain. With my A7iii at ISO 3200,
> Denoise AI works great, but fails totally at ISO 65535. If you look close
> at the denoise AI version, the small grain noise is reduced in the sky.
>
> I have one theory, although Denoise AI has AI in the name, the effect is
> almost as if it is doing an Inverse LCE (local contrast enhancement) at the
> pixel level. Hence its radius does not work with large noise, which
> requires larger radius which blurs the image. How is denoise AI to know
> what is detail and what is noise? It is only "artificially" smart.
>
> WayneS - Information Theorist
> --
> _________________________________________________________________
> Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
> Archives: http://lists.thomasclausen.net/mailman/private/olympus/
> Themed Olympus Photo Exhibition: http://www.tope.nl/
>
>

-- 
Tina Manley
www.tinamanley.com
http://www.pbase.com/tinamanley
<http://www.alamy.com/stock-photography/3B49552F-90A0-4D0A-A11D-2175C937AA91/Tina+Manley.html>
*https://www.alamy.com/portfolio/649596.html
<https://www.alamy.com/portfolio/649596.html>*
-- 
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz