Mark Metternich
Offline
• • • • • •
Upload & Sell: On
|
SergeyT wrote:
That's misleading at best.
* PSD and TIF are containers, so the RAW formats are. Neither PSD nor TIFs "rasterize" anything. They can be used to store raster images without loss of data caused by compression algorithms applied to image data in other format, such as JPEG.
* RAW files form digital cameras contain "rasterized" image data to begin with, with no more than 16-bit per channel integer precision, in many cases as low as 14 or 12 bit per channel.
* "Adjustment Layers" in Photoshop indeed can be viewed as formulas. They do not modify the underlying layers (image pixel values) until either the layers merged or image flattened. By creating and modifying "Adjustment Layers" no image data is lost or even touched. So the edits performed on or with "Adjustment Layers" are as loss-less and non-destructive as it gets.
* Photoshop does all the calculations with al least 32-bit floating point precision for ages, no special "intervention" is required. This is plethora for minimizing or even avoiding rounding errors during calculations.
* Majority of displays have precision of 8 or 10-bit per channel
* Majority of printers have 8-bit per channel precision, a few have 16.
* Before an image can be "sent" to either display or printer it must to be flattened and pixel values rounded to "fit" into the output device precision, or bit depth. There is no way around it. It will happen to images regardless of how they stored or processed.
...Show more →
gdanmitchell wrote:
I'm surprised to read something like this here.
I'll give you credit for your "instruction layers" term. That does make some sense, as the adjustment in smart layers are essentially instructions (or operations) that are applied to the image data without altering the basic underlying data. These adjustments do not alter the underlying image data, they simply affect how it is impermanently (if you choose) interpreted when the image is displayed.
"Traditional" destructive edits, whether in layers or not, alter the fundamental underlying image data in Photoshop and cannot be undone.
My workflow is almost entirely (with only a few rare and unusual exceptions) entirely non-destructive from ACR through Photoshop, and I can alter virtually anything I do to the image at any point, including going all the way back to the original unaltered raw file if I choose. All changes that I make in ACR or in Photoshop (again, with a few extremely rare changes in Photoshop) can be turned on and off, scaled (for example, with opacity settings), changed at any point in the post-processing workflow, edited later on.
What "non-destructive" means in the normal context of image editing is that the fundamental, underlying image data (the original raw file) is never lost or altered. All editing is done by applying operations to that original data, and those operations can be turned on or off, or edited at any point.
Analogy from the audio world:
Back in the day, we spliced magnetic tape. (Yes, I did.) You could, for example, splice out a word in a recording of speech. We did this by literally cutting the tape, removing the section containing the unwanted word, and spacing together the remaining words on either side of the missing word. That editing was destructive, in that there was no way to get back the perfect original recording. (Even after replacing the excised portion evidence of the operation remained. We did have fancy tricks, like editing a copy rather than the original, but the copy was degraded from the original...)
Later, with digital audio techniques, we have been able to do such things non-destructively. Essentially we give instructions to the playback process to skip over sections of the data that we don't want included. The effect is more or less the same (though way more powerful and precise!) than the older destructive splicing... and the underlying data are not modified at all.
That's what Photoshop smart layers let us do. We can apply all sorts of operations to the base layer as (to use your correct description) "instructions" on how to interpret or operate on the original data while leaving it intact.
In the end, differentiating between the familiar and well understood term "smart layer" and your new term "instruction layer" is a distinction without a difference.
That second sentence (the one beginning, "Simply put...") is misleading to anyone who is trying to understand the difference between destructive and nondestructive editing is. If what you are saying is that all changes to an image file's data are destructive because of what happens at the final output stage, I'm not quite sure where to even begin... (Try applying it to my two audio examples and see where you end up. ;-) )
I know this is a lot of words, but in my experience the non-destructive workflow from using smart layers is an extraordinarily powerful and useful thing, one that is not well understood by enough photographers.
While I'm at it, the term "lossless" is not usually used the way seen here. It usually refers to compression systems that make files smaller without actually losing image data. I suppose that we could logically call it a non-destructive compression method, but this ends up confusing things given how the terms are usually used.
Dan...Show more →
I'm not going to debate here, if you would like to debate, you can PM me.
But I will say that there is nothing misleading here, IMO. In fact, the opposite! It might be a big eye opener for some folks wanting to push the boundaries of quality processing, especially because this is the direction the industry is heading.
BTW SergeyT, I agree with a lot of what you have said above, but NOT all of it.
I stand by my initial statement, and will repeat it here:
FYI, heads up for quality minded folks:
Photoshop "Smart Objects" and "Adjustment Layers" are not truly lossless or "non destructive" edits. Simply put, they are Instruction Layers, in a Photoshop Layer Stack, to apply destructive adjustments (just as destructive as the real adjustments) to the image once it is Rasterized (turned into pixels - like Tiff, Psd...).
These adjustments are not applied at the lossless 32-bit Raw ("pure data" - as some have called it) level.
The unprecedented 32-bit Raw Layer Workflow that Adobe has given us (and is developing/expanding into the future...) exploits the phenomenal tools Adobe has given us, to be able to use 32-bit RAW Lossless Layers to get everything or nearly
everything done to the native file with NO data loss and NO lossy/destructive edits (even for those who make a lot of adjustments or use lots of creative techniques). One can even use superior tools like 32-bit masks! 😮
For those who want the highest quality ("future proofed" as Robert Park of Nevada Art Printers calls it) workflow, or photo files, or those who want top quality enlargements - there is simply nothing like it. The significant improvements show up in, and are, real world.
Lastly, those who do Black and Whites... there is no workflow like it!...Show more →
So, to keep it as simple and clear as I can for those who may read along:
Yes, of course it is true (and I said it in my statement above) when you adjust a “smart object” or an “adjustment layer” in a Photoshop Layer Stack, NO it is not immediately degrading/destructive.
But, YES, all the accumulative degrading/destructive edits happen/get applied when the image is Rasterized/Flattened.
Who in the world does not Flatten their images for Print or Web output?
Also, when a person brings their image from Lightroom or Adobe Camera Raw, into Photoshop (even if imported as a “smart object”) NO, these aforementioned Adjustment Layer or Smart Object Layer adjustments do NOT get applied in Camera Raw's lossless, higher bit depth, 32-bit Raw Data computation, which is very different from the pixel pushing done in Photoshop!
The aforementioned adjustments get applied to either the 16-bit or 8-bit (depending on what workspace people are using) rasterized pixel data during the Flattening. Such adjustments are NOT lossless, higher-bit depth, or raw parametric editing.
Whether a person does a ”real” Levels adjustment (Image/Adjustments/Levels) or an “Adjustment Layer'' Levels adjustment…
Or whether a person does a “real” Gaussian Blur adjustment, or a “Smart Object” Gaussian Blur adjustment, you still get the same lossy/degrading adjustment applied during Flattening.
When teaching post-processing, I often find photographers confused about the issue, I think partly due to what can be quite confusing terms thrown around on social media, such as “non destructive” used of both “smart objects” and “adjustment layers” as if they were the same as Camera Raw computation. I have found that many photographers wonder if they could have a nearly almost endless amount of “adjustment layers” and “smart object” layers of any type, with masks applied, to their precious image and never have any data loss, artifacts, color fidelity loss, banding, noise build up, or any other type of real world degrading... This is NOT true. I often find the term "non destructive" confusing or even misleading for many photographers.
Doing post-processing full-time for top class photography gallery enlargement (where the rubber meets the road in my life) we see the results of this data loss/degrading all the time! This is especially noticable making enlargements. I have also had the incredible privilege of helping many clients fine tune their workflow for very much improved results!
No greater jump in quality has come from anything, more than exploiting the awesome higher bit depth computation we have been given. Results change literally over night.
As a quite strong Adobe Advocate (although, I will advocate for other software too) I advocate for a “smart object” and “layer adjustment” workflow for most people because of the massive flexibility built in. Why? Because people can come back to their imperfect, or unfinished layer stack and make further changes to an extremely wide array of adjustments and/or fixes without extra/additional lossy/degradation.
BUT, with the Ultimate Quality Adobe 32-bit Raw Layer Workflow, a person can have that same type of flexibility built in as well, yet without the massive data loss or truncation/quantization issues/errors. For the highest quality fine art processing (this workflow is not for speed) it makes a lot of sense (and a big difference) to make the most of our bits!
No, the math is NOT the same (not even close) whether a person is making image adjustments on a rasterized 16-bit file or 8-bit file (with adjustment layers or smart objects) compared to the superior computation going on in Adobe Camera Raw to the initial Raw Data (and a person can make a linear raw file/profile if they choose - but usually not necessary). Camera Raw’s 32-bit float point raw data parameter editing has a higher bit depth and is by far superior (especially for those who make a lot of adjustments - which has become very common - or the norm for many - in post-processing today).
For those who make just one curves and saturation adjustment, and call it a day, and do not intend to really push image optimization, then yeah, who cares? Not to make a straw man, I'm just saying... But I find that A LOT of people do care about maximizing image quality, and many want more, or all the quality they can get out of their precious photo files! I estimate that at least 80+% of the photographers in my workshops are eager to learn and then begin implementing the Adobe 32-Bit Raw Layer Workflow with great results. Most want to keep moving forward with the technologies given to us.
YES, we see a real world difference in image files (especially for enlargement, big time) for many who start moving away from a workflow where they do a lot (or most) of their adjusting/editing in Photoshop, vs an Adobe 32-bit Raw Layer Workflow. I see the difference every day!
As an example, just today I worked with several black and white files (for gallery print) that had adjustments leading to severe banding/posterization in the typical aforementioned 16-bit workflow. This is very common with a lot of tonal work done in black and white photography. But when processed (by me) pushing the boundaries of the Adobe 32-bit Raw Layer Workflow... no banding/no posterization! In fact, smoother tone, a total lack of artifacts, no noise build up, no detail degrading, no common shadow yuckiness and so forth... It is awesomeness!
Rajan Parrikar wrote:
Sublime! Superbly processed.
Huge thank you very much Rajan! ✌🏼 I hope you and your family are doing well during these times…
junglialoh wrote:
Wonderfully presented in skillful technique.
Thank you very much for the encouragement. I appreciate it.
Brian Woolf wrote:
Lovely image, I find that I enjoy all the greenery, makes a beautiful compliment to the falls. And thanks for sharing all that info and data, it is appreciated.
Brian
Hi Brain. Thank you for your encouragement and response. It is my pleasure! 👍🏼 Keep up your own great work.
rrklepper wrote:
Very nice, I really like the feel of this photo.
Thank you very much. I really appreciate it. Some have asked me about these very long exposures, and that is it… they sometimes have a feel to them that I really love. Thank you and I hope you and your family are doing well during this time. 👍🏼
Grahamc wrote:
Excellent image.
Sorry to hear about the family - it certainly reinforces the need to get the jabs when they are offered!!
Thanks you very much!
I agree. Everyone who did not get the jab, now is holding on for dear life with Covid. Even my wife (her and I both have had the vaccine) now has it (she has been taking care of 5 family members in close quarters, and has had a HUGE viral load). I’m hoping that having had the vaccine, gives her body an advantage.
saaketham wrote:
What a beautiful place and image
Thank you. It is certainly one of the most beautiful places in the world. Thank you for the compliment. 😊
dbehrens wrote:
Mark - gorgeous pic with such a simple, clean, stunning comp! Really a masterpiece! f/8, ISO 100 @ 340 secs? Were you using CP + ND? In any case you got my vote! 
Hi Dave! Well, coming for YOU, that means a ton!!! 😁 I wish metadata would record the filters used… 🙂 I know I had a 9 or 10 stop in Hoya ND on there. I very much may have also had a Circular Polarizer. I love to, and quite often experiment like that. But the CP I am not certain about. I took a whole bunch of exposures of this, even in the 30 second or 1 minute range… but the nearly 6minute exposure here was definitely different. I’m sure you know how this waterfall moves around a lot. Blessings man!
|
