PDA

View Full Version : Destructive or non-destructive


rozmbow
August 2nd, 2009, 07:40 AM
I shoot mainly RAW and use PSE 6 and Adobe Camera RAW to PP them. I prefer PSE over the Lightroom.
My work flow is as follows -
First open the image in ACR and make the initial changes.
Second step is open the image directly in PSE and make further adjustments.
Last stage is save as a JPEG file.
Is this considered non-destructive or destructive method?
As I understand as long as the file is not saved it should be a non-destructive method.
Correct me if I am wrong.
Is there a better non-destructive work flow?

Codebreaker
August 2nd, 2009, 08:09 AM
It's non-destructive in the sense that nothing effects your RAW file. However, the save to a JPEG is somewhat desctructive. Or to use the more popular term - Lossy.

JPEG is a lossy format in that it compresses the data to make smaller image files. Colour information gets compressed more than the luminosity. However, modern compression algorithms are very good and provided you use the highest quality setting you'll get good results.

There is a possibility that when you do another save on your Jpeg file some additional data gets thrown away.

Safest option would be to choose a Loss-less format like TIF which can be compressed anyway. However, don't give up yet on Jpeg if it suits your needs - just be careful in its use.

BTW: I'm suprised by your comment about PSE v Lightroom. LR is far more powerful and provides a more effective workflow. It is different to use and takes some learning but is worth the effort to get to know it, in my opinion.

Colin

Jeff Perry
August 2nd, 2009, 12:05 PM
To add to the good advice Colin has offered, they way you edit in PSE can have as much to do with destructive/non-destructive editing as the file format you save in.

Example, if you edit your image thru the Enhance Menu options, all of the is being destructive to the pixel layer (your image information) making otherwise permanent alterations (once saved). Of course during the same editing session you can UNDO all of the changes (at least to the limits of the undo cache), but once saved, even as a PSD with layers, all those "Enhance" menu edits have permanently changed your pixels.

As an alternative, make all your adjustments in PSE using Adjustment Layers whenever possible. Almost all of the most used Enhance Menu adjustments can be replicated using non-destructive Adjustment Layers, e.g., Levels, Hue/Saturation, etc., and with a lot more flexibility (with fine tuning, masks and blend modes).

Part of the attraction to the Enhance menu tools is the set of "Auto..." enhancements, and easy-to-use slider adjustments, which is OK for quick and dirty edits, but once again, these are all destructive.

And as Colin has said, LR2 is an extremely powerful "editor" and has tools to perform just about every manner of adjustment without hammering a single pixel. Even non-destructive cropping (which can also be done in ACR).

Jeff

Codebreaker
August 2nd, 2009, 01:01 PM
Use of adjustment layers is good practise although I wouldn't classify them all as being non-destructive.

Of course that depends on your meaning of non-destructive.;)

Levels and Brightness adjustments I would classify as Destructive. Call up the Histogram display and then make a Levels adjustment. You will see gaps appear in the histogram which indicates that those particular Levels no longer contain Pixels.

So, whereas you original image had 256 Levels the adjusted version will have less. This is one good reason why 16 bit images are better. They start out with 32768 levels and a smaller percentage get lost.

Colin

Jeff Perry
August 2nd, 2009, 01:13 PM
Colin, my interpretation of non-destructive is that which does not physically alter the pixel layer, and can be undone at any time. Non-destructive adjustment layers are applying their "treatment tot he underying layer. Save it as a PSD, come back a year from now, open the file, delete the Adjustment layer and you are back to square one. Flattening or Merging will turn non-destructive adjustments into destructive actions without a doubt.

Same goes for things like non-destructive dodging a burning. The actual pixel layer is never altered, unlike using the D&B tools or using the menu Enhance functions.

Jeff

Codebreaker
August 2nd, 2009, 01:29 PM
Jeff....

Indeed this is the frequent interpretation. I point out the destruction of Levels as a main difference between using similar adjustments in RAW and Non-RAW. Doing similar in RAW does not reduce the number of Levels.

RAW is truly non-destructive.

Hair - sharp blade - divide by 2 ;)

Colin

Jeff Perry
August 2nd, 2009, 05:50 PM
Colin, point well made. I assume you are talking about an original histogram that has gaps to the left of the black point and/or to the right of the white point, and by moving inward either of the two points, you spread the tones and get the gaps in the middle. In that case, we never really had all those tones to begin with, and those "gaps" on the ends are now interspersed throughout?

jeff

Codebreaker
August 3rd, 2009, 02:40 AM
Jeff....

Yes, if your histogram shows the pixels don't extended to the Shadow and Highlight ends then you have less levels to start with.

However, moving the end points in widens the gaps between Levels - so its not just the number but the difference between Levels that may introduce banding.

Also moving the mid point will alter the Level count.

Colin

ImTryin
August 3rd, 2009, 04:01 PM
Use of adjustment layers is good practise although I wouldn't classify them all as being non-destructive.

Of course that depends on your meaning of non-destructive.;)

Levels and Brightness adjustments I would classify as Destructive. Call up the Histogram display and then make a Levels adjustment. You will see gaps appear in the histogram which indicates that those particular Levels no longer contain Pixels.

So, whereas you original image had 256 Levels the adjusted version will have less. This is one good reason why 16 bit images are better. They start out with 32768 levels and a smaller percentage get lost.

Colin

Please correct me if I'm wrong - - but my understanding is that the same percentage will be lost. But it's still much better with 16 bit because there is a much larger gross amount to begin with so the same percentage reduction will not be so visible.

Codebreaker
August 4th, 2009, 02:52 AM
Ron....


Yes, bad description :o. I'll go with your's.

Colin

rozmbow
August 4th, 2009, 01:29 PM
Just to make my question more focused - My workflow - after the ACR I just hit the "open image" button. It opens im PSE, I make some more adjustments (usually in the Guided" mode) and after all is OK - save (ONLY ONCE) as Jpeg in the highest quality.
Is this process "lossy" or "destructive" or is it OK?

Jeff Perry
August 4th, 2009, 08:33 PM
You are talking about two different things. Saving as a JPG, once, twice, thrice (is that even a word?), is saving work in a lossy file format because the JPG file "process" by design introduces a lossy form of compression. If you shoot JPGs in your camera, the image processing engine in your camera introduces a lossy compression when it saves the image to your memory card. If you shoot RAW, generally it is a lossless compression used by manufacturers (there is still some debate about that in at least in the world of Nikon). TIFF is also a lossless format, but eats catrd space like nothing else. That's why compression is used.

The term "destructive" is most often reserved for the process of irrevocably "hammering" your pixels thru the editing process. Change a colored pixel to B&W, dodge pr burn, etc., and unless you use adjustment layers you are permanently altering the pixel data. It can generally be undone during the same editing session.

Non-destructive editing/adjusting, via Lightroom, ACR or PSE/CS4 using adjustment layers stays non-destructive in the PSD format, but as soon as you save in JPG (which automatically flattens your layers) your edits can not be reversed. So in essence you are using destructive editing by flattening all your layers into a single layer for saving as JPG, and then you are using a lossy format when you actually save the file to disk as well.

Clear as mud, right?

Don't worry about it, have fun, capture some great images, do some cool editing and save your work as native PSD files, and then output to JPG when needed.

Jeff

ImTryin
August 4th, 2009, 08:58 PM
Ron....


Yes, bad description :o. I'll go with your's.

Colin

No problem, Colin - - I'm just trying to understand this stuff and some of it is quite subtle.

Cheers,

Ron