Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: Philmar on August 18, 2007, 11:02:29 am
-
I have CS3 and use ACR 4.1 for processing my RAWs. I love it's intuitive easy workflow.
I have many many jpeg files from my pre-RAW days that I'd like to improve by doing slight adjustments i.e. crop, adjust tone, WB, set white point, contrast, saturation, colour tweaking,resize, sharpen. I do this so they can be displayed online on a Zenfolio or pBase account, NOT for printing.
I like ACR's workflow and myriad of adjustment options yet I wonder if it is a better platform for jpeg processing than Photoshop. For ease of use it clearly is, but what about end product?
For best results, Photoshop requires me to create a new adjustment layer for each adjustment. I DO speed up the workflow by creating Actions that convert the files to 16 bits as well as creating new adjustment layers upfront. I still have to apply batch processes to merge layers, resize, sharpen, convert back to 8 bit and save as jpg. But I still have to go in to each layer and make adjustments. ACR 4.1 is much easier because all the sliders are more easily accessible. But I wonder if it sacrifices results because I don't know if it is working in 8 or 16 bits. With ACR workflow is easier to recover shadow/highlight detail (though I am not sure if it is more effective).
For the best results am I still better off using PS for jpeg tweaking, despite its more cumbersome workflow, because I can work in 16 bits?
I really like ACR's workflow. But from what little theory that I do have it sounds like PS ability to work in 16 bits still gives it the edge.Or am I missing something here?
-
So does everyone use PS for jpeg enhancement? Nobody uses ACR?
What I am trying to understand is what is going on when one works on a Jpeg in ACR. Is it being manipulated in 16 bits? Or is it being worked on in 8 bits and THEN created into a 16 bit file AFTER the 8 bit manipulations?
With PS I can increase the file to 16 bits BEFORE I do and layered adjustments. Is this what is happening behind the scenes in ACR?
-
By all means convert JPEG to 16-bit mode as the first step after opening, but if you're using actions to batch process JPEGs, using adjustment layers is stupid. Adjustment layers are great if you want to try several different combinations of adjustment before deciding on a final result, but they are far more memory and CPU intensive than simply making a series of edits to the image. If you are batch processing, just do each adjustment directly to the image in 16-bit mode. You'll get the same result as a stack of adjustment layers, but the batch will run much faster and PS will use much less memory.
-
So does everyone use PS for jpeg enhancement? Nobody uses ACR?
What I am trying to understand is what is going on when one works on a Jpeg in ACR. Is it being manipulated in 16 bits? [a href=\"index.php?act=findpost&pid=134121\"][{POST_SNAPBACK}][/a]
Its being converted to 16-bit, linear encoded ProPhoto for any editing. I'm not sure this is a good idea myself. I think rendered images should have their own processing pipeline.
-
By all means convert JPEG to 16-bit mode as the first step after opening, but if you're using actions to batch process JPEGs, using adjustment layers is stupid.
[a href=\"index.php?act=findpost&pid=134156\"][{POST_SNAPBACK}][/a]
I use the actions to apply setting to batches i.e. create layer adjustments for saturation, curves, levels ect.
THEN I go in to each file and adjust each layer. The recording action only creates the many different adjustment layers for each jpeg (one saturation adjustment layer, one curves adjustment layer, one levels adjustment layer...ect). I still have to adjust each layer of each jpeg individually. The action just saves me the bother of having to create several separate adjustment layer.
Before this step I also have converted every 8 bit jpeg to a 16 bit jpeg.
But what I REALLY want to know is what bit depth does ACR do its manipulations in...8 or 16?
-
There's no such thing as a 16-bit JPEG. 16-bit requires JPEG2000, PSD, or TIFF format. And your 16-bit question has already been answered by digitaldog.
-
But what I REALLY want to know is what bit depth does ACR do its manipulations in...8 or 16?
[a href=\"index.php?act=findpost&pid=134241\"][{POST_SNAPBACK}][/a]
16-bit (probably really 15 bit like Photoshop).
-
There's no such thing as a 16-bit JPEG. 16-bit requires JPEG2000, PSD, or TIFF format. And your 16-bit question has already been answered by digitaldog.
[a href=\"index.php?act=findpost&pid=134261\"][{POST_SNAPBACK}][/a]
Sorry General Sir! I meant to write 16 bit 'file' not 'jpeg', General Sir! Thanks for the correction.
Thanks for the clarification, digitaldog
I will now do all jpeg edits in ACR.
-
16-bit (probably really 15 bit like Photoshop).
[a href=\"index.php?act=findpost&pid=134308\"][{POST_SNAPBACK}][/a]
I can confirm it's 15 bits, just like PS. Or at least the "16-bit" TIFF files generated straight from ACR 3 with its own Save option are 15-bit (see the holes in the 1:1 histogram):
(http://img214.imageshack.us/img214/7109/testacr16o15bitshislf4.gif)
-
Many thanks to all of your learned responses!!!
Methinks I'll do all my JPEG edits in ACR and reserve PS for any localised edits.
As an aside I did get this from Jeff Schewe in the Adobe ACR forum:
"...if you open a JPEG (or 8 bit tiff) it is converted to 16 bit, linear Pro Photo RGB for editing. It's been shown (by Lightroom) that while limited in recovery (cause the original ain't linear) the fact that the totality of the edits in CR are in linear ProPhoto you could see a real benefit to processing camera JPEGs (as apposed to already edited JPEGS saved out of like Photoshop) over doing similar edits in Photoshop. "
-
Its being converted to 16-bit, linear encoded ProPhoto for any editing. I'm not sure this is a good idea myself. I think rendered images should have their own processing pipeline.
[a href=\"index.php?act=findpost&pid=134165\"][{POST_SNAPBACK}][/a]
So would you still recommend working on JPEGs in Photoshop (after conversion to 16 bit depth)?