I'm assuming the parameter the OP is referring to is the in-camera Processing Parameters for sharpness, saturation and contrast. Manjade asked about minimising noise a few days ago and I advised then that for JPEG shooting, reducing the settings in-camera for sharpness and contrast especially have a big impact on the resulting noise - not only locally on a per pixel level, but reducing contrast particularly also reduces the depth and severity of shadows leading to an overall more pleasing looking result with high ISO images by softening the depth of shadows.
Excellent article. What you show is that the medium and technology is different from film, but many of the goals are the same. Ask anyone who's spent time in a darkroom about dodging and burning, or controlling contrast through water bath development - different approaches to dealing with the same issues.
Thank you for sharing the test results. I am wondering why f/36 was chosen. What would the noise be like at f/8 or even wide open as this is the more likely scenario one would use during the low light situations.
I've done a couple of series for this purpose and as my interest is available light, mine were taken wider.
If anything, the converse is true, but if that's the conclusion you've drawn, I've explained or illustrated something very badly indeed.
Indeed, I'll help if I can. Post them at DPR rather than e-mailing them as this gives others the option to contribute too.
Boo,I've done a couple of series for this purpose and as my interest is
available light, mine were taken wider.
Shots below both taken at 1/40, f2.8, 800 ISO with Sigma 28-70mm
lens at 45mm. I took two sets of images in manual with the
settings given, one set with -2 Processing Parameters, the other
set with +2. At the time the reason for these frames was to post
in a thread discussing whether the PPs had any impact on the RAW
files, so they were shot in RAW and the embeded JPEG used for
relative comparison of the appearance of the images with and
without Parameters coming into consideration. Noise was only one
of the factors being considered.
Full frame of the medium embeded JPEG resampled, taken at +2
in-camera Processing Parameters:
Full frame of the medium embeded JPEG resampled, taken at -2
in-camera Processing Parameters:
1:1 pixel crop from medium embeded JPEG:
1:1 pixel unsharpended crops from the two RAW develops:
I did another series at 1600ISO with the 50mm lens at f6.3 - I
actually thought they were wider until just checked - but now I
remember setting it at that to get a dark area in the background
just going out of focus. That set were done specifically to
investigate the impact Processing Parameters have on noise. Those
are at the bottom of my 'Working with high ISO' tutorial at:
http://www.zenadsl5251.zen.co.uk/photos/highiso.html
All of the comparative examples shown were given exactly the same
workflow in terms of resampling routine.
--
So many photos, so little time . . .
http://www.peekaboo.me.uk - general portfolio & tutorials
http://www.boo-photos.co.uk - live music portfolio
http://imageevent.com/boophotos/ - most recent images
Please do not amend and re-post my images unless specifically
requested or given permission to do so.
I can definitely see the difference. The one shot with parameter 1 has the most noise and the one the custom parameter which I assume it -1,-1,-1,0 has the least color noise. The parameter 2 shot is in the middle.
I don't use Photoshop at all, my essay is general in terms as I don't know the specifics of how my techniques would translate - they will, but I couldn't tell you more details.
On the web page, the series comparing Processing Parameter settings were 1600ISO. The ones posted higher in this thread were 800ISO.
I'm using a CRT and can clearly see the difference between not only the noise levels, but the general differences in the data captured, the high parameter images are darker in shadow areas with less detail and more towards being blown in highlights - an obvious result really when choosing to make an image more contrasty. It's simply less forgiving with high ISO images, which perhaps need nursing along a little more than lower ISOs, so any helping hand you can give the data is worth considering. People will need to do their own comparisons and see what suits them, my page is in response to all the people who e-mail me daily asking about high ISO and noise and how to either prevent it, or treat it.
Me neither as I use RAW, but plenty of people don't and they're usually the ones asking such questions. I still use the embeded JPEGs and I want them as good as possible for when I do.
This is the 300D/350D forum, so how these cameras perform is what is being discussed here. I can't comment on the 20D images as I don't have one. With the right handling - which was the point I was making - the 1600ISO imges from the 300D are clean too.
