400D workflow
Been meaning to set aside some time to play around with different workflows.
The thing I’ve grown to hate about the 400D is digital noise and also the dynamic range in photos, two fairly important things seeing I’m quite nocturnal.
I did some experiments, not spending any real time on each process as I don’t usually like to spend much time going through photos. But think I’ve realised I’ve been using a bad workflow.
First experiment was noise reduction with a shot I took in very low light with 1600 ISO. I will point out I don’t mind film grain, it’s quite different to the ghastly coloured noise you get with the lower end digital dSLRs.

Here I’ve put DPP at ‘quality’ settings, noise ninja on auto with smoothing around 5 and camera raw at 50 for colour noise reduction and 25 for luminance. As you can see noise ninja looks quite artificial, raw looks chunky and DPP did a good job of reducing coloured noise and leaving just luminance. But to my surprise and without touching any colour settings in photoshop or DPP the colours and contrast are quite different.
Next I grabbed a shot I uploaded last week which I processed through lightroom (and maybe photoshop, can’t remember) and ran it through DPP, applied noise ninja with profile settings for 800ISO canon 400D and saved it out. I’m surprised to find that it came out with less contrast, nicer skin tones and also less noise on the face. The process took about 2 minutes.
Today:

Last week:

Wonder if any other 400D/350D users have processes they’ve found that works well?
3 Responses to '400D workflow'
Subscribe to comments with RSS or TrackBack to '400D workflow'.
Thanks for posting this Tim. Lately i’ve been thinking about “refining” my pretty terrible workflow. It’s a subject i dont really hear discussed that much. Maybe when people find a good one they dont want to give it away!
I also like taking photos at night, but have pretty much given up with my 400D, anything above 400iso is pretty bad.
I’ve only used Noise Ninja in the past, but will try out DPP next week. I’ll do a similar test to you and see what sort of results i get.
I’ve never really been able to get noise ninja to work for me that well. I seem to loose too much sharpness, and skin tones especially look a bit too “plastic” (which seems to be bit of a symptom of digital cameras anyway).
I wonder if theres a way of getting DPP to keep a similar contrast to the original shot? It seems to have bit of a mind of its own. Besides that it looks to work pretty well.
Thanks for the idea. I’ll see how i go.
cheers
arran 30 May 08 at 11:42 pm
Ah I actually preferred the contrast in DPP.. I don’t know why but it seems camera raw likes to crunch the levels a bit automatically. I just assumed it was how the 400D captures the tones. The skin tones from that night were also coming out greener for some reason in camera raw but DPP seemed to stay more faithful to them without me changing any white balance settings.
Anyway let me know how you go.. I’ll give the process a good work out next time.
And yeah noise ninja can be a bit hit and miss.. Make sure you’ve gone to the web site and got the Rebel XTi profile and loaded that in.. Auto profiling seems to be where it becomes overblown.
tim 31 May 08 at 12:02 am
Ok i gave it a shot taking your advice.
The results i got were a tiny bit different from yours, but i’d say it has alot to do with the actual image being used.
I quite like the result i got from DPP. It gets rid of alot of noise but keeps some grain so the image retains some sharpness and contrast (in my opinion). Noise ninja is pretty powerful, i think i’ll have to play around with it a bit more, but i like how it kept some edge sharpness. The bad bit was how it made skin look like plastic. But like i said, i think i need to play with the settings a bit more, you got a nicer result with it than i did (i got the profile like you suggested but probably went overboard by using 1600iso).
Camer Raw seemed pretty average to me. It looks like it just blurred everything a bit. I dont think its all that “sophisticated”.
Anyway, i put the results on my blog. The best indication is probably how NR works with full images not tiny sections, so i havent jumped to any solid conclusions yet.
arran 31 May 08 at 5:06 pm