Adobe Camera Raw workflow question #2901

Question

Now that I have my new digital camera-dedicated rig up and running I have been using ACR and CS2 to process my backlog of Canon 30D CR2 files. I am trying to get them ready for posting on the internet. Currently I am using ACR only to tweak Exposure and White Balance, then I move the file in 16 bits/AdobeRGB to CS2 where I do noise reduction, levels, curves adjustment, saturation then resize, sharpening then convert to an 8 bit, sRGB Jpg.
I have adopted this workflow because it was similar to the one I used when my old creakin' PC required me to use DPP to do my RAW conversion (and I subsequently did PP in CS with a 16 bit TIFF)
But I notice ACR has many choices and options such as contrast, shadows, saturation, sharpness, noise reduction controls that one can perform prior to RAW conversion. I have read in the past theoretical comments that adjustments are best made to a file during RAW conversion rather than while as a jpg. Does the same apply to a 16 BiT TIFF? In the workflow above, am I working on a TIFF?
I am wondering if I am better off to make all of the adjustments in ACR window instead of the more lengthy process in CS2. Are there any advantages to doing these adjustments prior to RAW conversion? Or is it better to do PP on 16 bit files using layers?
What are the relative merits of processing the file in either ACR window or as an 16 bit file in CS2?
If it matters, I work on a PC and prefer to PP each file separately rather than in batches. But I DO use batch runs in CS2 to save time using Actions with around 10 file batches to do noise reduction and resize/convert to sRGB/save as.
As always, thanks for the time you take to help me out!!

Answer 1

>"What does Pixelgenius PhotoKit Sharpener do that I can't do, perhaps with more trouble, by applying what I've read in his book? And would I have a problem trying to use it with CS3? "
I have a few observations and perhaps Jeff can comment. With images taken with a digital camera at high ISO, noise reduction may be necessary in the capture phase of sharpening. PhotoKit offers the sharpen and smooth option for this purpose. One can exert some control over the smoothing process by changing the opacity of the smoothing layer and the blend if sliders.
In his sharpening book Bruce uses Photoshop's reduce noise and despeckle filters to reduce noise (sometimes using the despeckle filter multiple times), but I don't know what filter is used in the PhotoKit smoothing operation or what parameters are used for the reduce noise filter if it is used.
If you use Bruce's book and the "roll your own" approach, you have more control over the noise reduction process and also the possibility of using a third party NR product such as the Noise Ninja plugin. Furthermore, just as an edge mask is used to restrict the sharpening to the edges, a surface mask may be used with NR to keep the smoothing away from the edges. The surface mask may be derived by inverting the edge mask used for sharpening, but Bruce says that some tweaking may give improved results.
An alternative would be to use the third party NR prior to using PK sharpener. However, you would still need to make a surface mask for optimum results, but I doubt that many users take the trouble of doing this.
In my own work with the Nikon D200 (which has rather high noise at high ISO), I find that I often get unacceptable noise and artifacts with PK Sharpener when used with high ISO images.
Also, many landscape photographers mask off areas of clear blue sky and foliage that do not need sharpening, thereby avoiding accentuation of noise secondary to sharpening in these areas.
>Bruce gave you the recipies...he didn't give you the exact numbers to use. So, you'll have to do what he and I did...trial and error, to arrive at optimum numbers...
For output sharpening, the PK defaults seem to work quite well and the trial and effort of rolling your own is usually not worthwhile, IMHO.