That is right, you read correctly.
I accidentally stumbled onto a PS CC process that allows removing noise from JPG files or so it seems.
Before I post a tutorial on it I need to verify my workflow using examples that I do not control.
Out of focus capture ok. The process will not change the lack of focus but if it does correct the noise in poor conditions that will be further indication that my 'discovery works'
Please Check the box (store original) and do post unadulterated captures meaning unprocessed.
Thank you.
You can check what I am looking at in this
thread and more specifically this
post.
Capture use and credit courtesy of
rmalarz
A few kaka shots from an indoor gig I had to do. ISO 12,800 so lots of noise. if it helps f5.6 and roughly 1/15 sec.
Here's one that's not quite right.
pithydoug wrote:
A few kaka shots from an indoor gig I had to do. ISO 12,800 so lots of noise. if it helps f5.6 and roughly 1/15 sec.
Sorry, utter failure...
:shock: :hunf: :thumbdown:
If it's specifically jpeg artefacts that you're looking for, why not take any image and export it as jpeg a few times. The deterioration should be accumulative.
One of the tests for your process is whether it helps with banding or not.....
The trouble is that it does not really remove JPG artifacts (I wish it did)
Removing them would blur the picture, something I am fighting against.
Rongnongno wrote:
It works on yours but is this worth the effort??? ...
Probably not since a simple application of Neat Image produces the attached result. Go ahead and peek at these pixels.
That's for about $80 invested and two minutes of effort.
R.G. wrote:
One of the tests for your process is whether it helps with banding or not.....
It does not remove banding. That can be done with a simple gaussian blur, masking, opacity and blending if to limit the blur.
Fingerprint correction is an interesting concept. I've long had a suspicion that if you could somehow do a pixel-by-pixel correction for non-linearity and non-uniformity between pixels, you have the potential to produce perfectly rendered images where all of the sensor's imperfections and inconsistencies can be compensated for. It would probably take a fair bit of in-camera processing and it'd probably need some very large correction factor tables, but perfection would be the glittering prize.
R.G. wrote:
Fingerprint correction is an interesting concept. I've long had a suspicion that if you could somehow do a pixel-by-pixel correction for non-linearity and non-uniformity between pixels, you have the potential to produce perfectly rendered images where all of the sensor's imperfections and inconsistencies can be compensated for. It would probably take a fair bit of in-camera processing and it'd probably need some very large correction factor tables, but perfection would be the glittering prize.
Fingerprint correction is an interesting concept. ... (
show quote)
You may want to post this in
rmalarzs thread. I am sure he will appreciate your comment. I agree with you by the way.
R.G. wrote:
... One of the tests for your process is whether it helps with banding or not.....
The best way to deal with banding is to avoid it by doing your work on a 16-bit TIFF or the original raw file.
You are guaranteed to create banding by making graduated adjustments to contrast, etc. while working with an 8-bit image - JPEG, PNG, etc.
Having to remove banding means that you already screwed up.
Rongnongno wrote:
Sorry, utter failure...
:shock: :hunf: :thumbdown:
That is too much of a challenge.
Here are the Neat Image versions, not an utter failure, less than a minute for each one.
Not too hard to find noisy jpegs in my collection :lol: .
-
If you want to reply, then
register here. Registration is free and your account is created instantly, so you can post right away.