MCAsan wrote in post #18485345
It is not just a matter of recording history. When you leave Develop module and go to Photo module you are rendering the raw image plus any Develop adjustments into a pixel image. The only time I want to render a pixel image from my non-destructive edits is when I have to print the image or create a jpg to post.
BigAl007 wrote in post #18485710
Staying in RAW for as long as possible is handy in that it allows you to put off things like choosing a colour space etc. I would always prefer to be creating a first generation copy when exporting my finished images. Every additional conversion step between the original data, and the finished image is a potential point for conversion losses.
Care to demonstrate the "conversion losses" that you dread, outside of a handful of stereotypical examples that can be handled with a little bit of finesse? Is a 16bit RGB file converted into the ProPhoto color space not big enough to start editing in a pixel editor?
I think most peoples' desire for a parametric "non-destructive" workflow has gotten to the point where parametric "editors" like Lightroom have become bogged down with attempts to make them perform the duties of a pixel editor. I do not need to point out the recent blow back from users in this regard.
Folks "want their cake" but probably cannot articulate why, beyond the notion that they want their cake for convenience.
I would always prefer the highest quality raw conversion into an appropriate working color space for retouching and grading in a capable pixel editor - period. Raw conversion and grading/retouching are two distinct tasks that benefit from being done in two distinct environments. I cannot understand peoples' need to have these things done in a unified application where both tasks potentially suffer (quality, workflow speed, precision) as a result of what is probably the notion of convenience over quality - jack of all trades, master of none. At some point you are going to have to commit to your edits, even if you have multiple versions of your envisioned output.
Can you make a good image without ever leaving a Lightroom-like environment? Sure - and many people swear by this workflow. Can you make a better image by converting the data in a high-quality raw converter and finishing it in a pixel editor? 99.9% of the time, yes, most of the time with edits that are burned into the original pixels. If you do not need to make a "better" image, then staying with Lightroom-like edits is a good solution. But the idea that there are losses and artifacts that can only be avoided by staying in Lightroom, or similar "raw" environment until the very last possible moment is probably a belief rather than a fact. And most of the time, altering the original pixels in a way that simply is irrecoverable is a risk that most folks are going to take to make the best image possible, regardless of the technology.
Of course, I am open to data that would prove otherwise.
EDIT - using a mask is "destructive" in the sense that you are altering some pixels more than others in the downstream rendering pipeline. Just because you can create variants or versions in a raw converter and then apply different local corrections does not make the workflow more legitimate compared to working on pixels that have already been rendered. So, wouldn't you rather have more control over the "mask" by having access to channels that, most of the time, provide more precision in the local corrections you want to make? Applying a Lightroom adjustment with a brush is also "destructive" in this sense. Of course you can undo it by deleting that local correction "pin" but you can also delete the analogous adjustment layer in Photoshop. In other words, layers in Photoshop provide the same, and more flexible, non-destructive editing capabilities that Lightroom may provide, but, by necessity, work on pixels that already have committed to a particular color space.
Some differences in this workflow are the precision of the correction and the resulting file size of the working document - but these may be offset by the speed at which the user can render and work with the "masks" and their local corrections, as demonstrated by the growing frustration of users who reported that their machines came to a standstill as they tried to apply "masks" for local corrections and local spot healing. And, while the new CC Classic "masks" include luminance and color refinements, the user has relatively little control over this refinement compared to the power of being able to use RGB, CMYK and Lab channels to target specific areas of an image for mask construction. These more advanced mask-making strategies are probably beyond the ken of the average user of Lightroom, which indicates that, although Lightroom may be used by "professional" photographers, it is likely less a tool for professional image processing and image making users who depend upon complete control over the elements of the image processing and retouching pipeline. The recent emphasis by Adobe on the mobile Lightroom CC platform also would tend to confirm this distinction.
Again, you can make a perfectly good image in Lightroom, but you can make that image better by editing your good Lightroom image in a pixel editor like Photoshop, or some of the more recent hybrid tools like Luminar, Topaz Studio, etc.