LR vs PS: Workflow

[monkeydeus]

I know this discussion in some variant is available elsewhere, a la the yahoo color theory group, and I'm not trying to argue which tool is better at opening up a range, or expanding a range.  It's a little simpler.

If you *know* you are going to be editing a file in PS, say to add a blur layer, whatever, what is the minumum you would do in LR before export.  Should I just set the black/white points, white balance, and be done?  

I think I can do noise reduction better in PS with noise ninja, I have better curves control, more precise color control, I use Photokit sharpener, so I can probably get better sharpening; I'm just unclear what the "bare essentials" are for LR.

[digitaldog]

Do ALL the heavy lifting in a Raw converter. You simply don't have the control to render images in Photoshop, they are pre-baked into the color and tone, you can only shift around pixel values (slowly) and introduce lots of data loss. Dan on his list illustrates he doesn't have a clue about the differences between image correction and image rendering. Not at all the same. Mark Segal and Richard Wagner have recently posted very good explanations of the differences although I suspect many on the list don't 'get it' the host included.

Metadata, Raw renderings is quite fundamentally different from pixel based image correction. The host of that list only understands the later. This is why he suggests you zero out all the CR or LR rendering controls and then do the work in Photoshop. Dumb, slow, not anywhere as powerful and simply the wrong tool for the wrong job. Do EVERYTHING possible in the Raw converter, use Photoshop to tweak pixels; that's what it does well.

There's no reason why noise reduction ala Noiseware or output sharpening ala PhotoKit Sharpener couldn't be coded into LR or CR.

BTW, image correction, at least the big ugly fixes Dan teaches should be reduced by a huge factor if you simply render the Raw data as you wish from the get go. OK, someone hands you a rendered ugly image, you don't have the Raw, fix it in Photoshop. Rendering isn't fixing! Rendering is producing the color and tone you wish to express from the image as you work with the Raw converter. If you need a big, global color or tone tweak, go back to the Raw and rendering it correctly.

[johnwolf]

My rule of thumb is that if I can do it in LR, I do it there. That includes all global corrections, except those that are output dependent, like sharpening. Then I do the local edits (dodging/burning, spot removal, vignetting, blurring), and softproofing and printing in PS.  

I'm not sure how CS3 works with LR, but I use CS2, so once I open the image in PS and start adding adjustment layers for dodging/burning, sharpening, etc, I'm finished with LR.

John

[Mort54]

You simply don't have the control to render images in Photoshop, they are pre-baked into the color and tone, you can only shift around pixel values (slowly) and introduce lots of data loss.

You must not use layers and blends!!

[digitaldog]

You must not use layers and blends!!
[a href=\"index.php?act=findpost&pid=136003\"][{POST_SNAPBACK}][/a]

I do, that doesn't change what I wrote about baked pixels. There are rendering controls to create pixels that simply cannot be duplicated using Photoshop and existing rendered pixels, blends, layers or not. Rendering is pixel creation. Its not pixel value manipulation which is what Photoshop does.

Photoshop is about image correction (by altering pixel values). LR/CR and all Raw converters are about rendering pixels from data that has no color. Its not correction. Then there's the huge difference between metadata instruction based rendering and pixel based (immediate or upon flattening) corrections.