Update: In January 2003 the 8-bit vs. 16-bit topic came up yet again on Dan's Color Theory list http://www.ledet.com/margulis/ACT_postings/ColorCorrection/ACT-more-16bit.htm. My site in general, and this page in particular, came into the discussion. The thread closed with Dan making this statement:
"As they can't produce any such images, their only alternative is to go ad hominem, as Bruce and others unfortunately have. The site speaks for itself."
The first sentence makes me think that perhaps Dan does not understand the meaning of the phrase ad hominem. For everyone's benefit, here are Webster's definitions:
Keep these definitions in mind as you read this page, which contains my original, unchanged message.
With regard to the second sentence, I agree with Dan: let this site speak for itself.
Finally, Dan has never contacted me about this page. My goal in preparing it was to be truthful and objective. If it contains errors, please let me know.
Update (December 2005): Here is another comment by Dan Margulis on 1 November 2005:
"One of the more prominent advocates, Bruce Lindbloom, was so frustrated by his inability to produce a persuasive image that he posted a web page that accused me of sabotaging my 16-bit images before testing them. Also, he asserted that I kept my results private and that nobody else could verify them. Both statements are categorically false, and Lindbloom knew that they were false when he posted them."
Here's the link, if you want to read the whole thing yourself: http://www.retouchpro.com/forums/showthread.php?s=1c37c843d7742f7bc3114f4b4cf96e40&p=104205 Search for "Lindbloom".
First of all, I am not a 16-bit advocate. If you read my discussion below (still unedited from its original publication), you will see that I am not a "prominent advocate." I try to hold to an objective view, recognizing the benefits and pitfalls of both positions.
Second, I have no frustration caused by any inability to produce a "persuasive image." As I state in my discussion, my position is that it is foolish to get drawn into this challenge in the first place. My advice to others is to avoid it, and you know what? I take my own advice. There is no frustration in steering clear of chasing red herrings.
Nor do I experience frustration from Dan's 8-bit advocacy position. Everyone is entitled to a view on this controversial subject, including Dan. I will admit though, to frustration caused by the method by which he comes to this conclusion, and the method by which he subsequently defends it.
Here's another excerpt from the same link that is worth looking at, related to an image submission made from an unnamed person (not me):
"Then, he acquired the image in ProPhoto RGB, an ultra-wide gamut RGB definition that is rarely used in professional work. The image contained a large area of sky. Applying the drastic curves that were needed to lighten the image to the 16-bit file resulted in a perceptibly smoother and more attractive sky than the one done by converting the file to 8-bit in Photoshop and applying the same curves there. When the same image was captured with the same sabotage in Camera Raw into either of the RGB definitions that most of us use--the narrow-gamut sRGB or the wide-gamut Adobe RGB--there was no significant difference between correcting in 8-bit or 16-bit."
This is a classic example of one of Dan's favorite techniques (see section V2 below). If an example is presented that shows an 8-bit/16-bit difference, a rule is immediately created, on-the-spot, that disqualifies the image. None of Dan's original six conditions would disqualify a ProPhoto image (you can read these conditions below in section I), but it appears as though ProPhoto images are no longer acceptable. If one takes this technique to its logical conclusion, Dan's 16-bit challenge would become "When considering all images showing no 16-bit advantage, 16-bit images show no advantage."
I have nothing against Dan Margulis. As I have repeatedly stated in various venues, he is a knowledgeable person who has helped many people by way of his many contributions online, in publications and through his seminars.
But making statements such as those quoted above, involving accusations of sabotage and deliberate postings of false information, must be challenged. Why should anyone believe these statements? Where is the supporting evidence? My single most important message on this page is found in the last paragraph at the very bottom of the page, which highlights the difference between an assertion and an argument. Learning to recognize assertions when you encounter them, and then forcing the other person to turn it into an argument is a very valuable technique in many aspects of life. Learn it.
Finally, Dan Margulis has never contacted me personally about this page. Nor has anyone, through personal contact, ever challenged the truthfulness of anything said on this page.
Dan Margulis has, on numerous occasions, and over a long period of time, asserted that 16-bit color correction offers no advantages over 8-bit correction. There is a certain group of people that claims the opposite (the 16-bit Advocates). These two factions have engaged in many heated arguments on various online forums. As an engineer with 25 years experience in color science and digital imaging, I thought that there must be some objective way to determine which side was right, so I investigated further. Here is a summary of my findings and opinions.
For the record, Dan's challenge consists of the following six conditions (this is a direct, unedited quote from Dan, so the "I" and "me" in the following is Dan speaking):
I had always thought the test was this:
This understanding was wrong. The fifth of Dan's conditions ("Condition 5") modifies step (a) as shown here:
I specifically asked Dan about this on 17 September 2001 and Dan confirmed it on the following day (read the archive messages on his "Color Theory" list on those dates to get the full context). Dan further elaborated on his rationale behind Condition 5 (again, a direct quote from Dan):
No, you have it right, possibly with an intervening step. There seem to be four possibilities.
- Working in 16-bit is better because the extra data is valid and valuable;
- Working in 16-bit isn't better because the extra data is valid but irrelevant;
- Working in 16-bit isn't better because the extra data is bogus;
- Working in 16-bit is better even if the extra data is bogus, because there's something inherent in the calculation method that favors it.
Anyone advocating working in 16-bit has to be, IMHO, saying that #1 is true. The test method above is to find out whether #4 might not be true instead.
Dan is testing #4 above and the so called "16-bit Advocates" are testing #1. Dan's test strips off any "extra data", right out of the starting gate, whether it is valid or bogus.
I think they may well be correct that in some circumstances, performing color corrections in 16-bit will yield higher quality results than performing the same corrections in 8-bit. However, this conclusion is true only in cases where all of the original 16-bit data is retained (i.e. the "extra" data has not been discarded). I am quite certain that none of those holding to the 16-bit advocates position is discarding the extra data like Dan is doing in his tests.
I think, given the six conditions of his test (especially Condition 5), Dan is probably also correct. Both sides of his test use only 256 levels in each channel, and because of this, I would be surprised if any 16-bit image would look substantially different from its 8-bit counterpart. In fact, I'm surprised that Condition 1 is even necessary. But realize that this is a fundamentally different test than what is performed by the 16-bit Advocates.
Don't do it. I'll admit I was tempted, but after investigating the nature of the test, there is absolutely no way I way I would ever get involved. Why not? Here are some of my reasons:
A final word to the 16-bit Advocates that have taken Dan's bait in the past: Shame on you! You should know better than to get enticed into such a fiasco without doing your homework first. The challenge is set up as a self-fulfilling prophecy — you are guaranteed to lose.
There is some question about the "extra data" contained in 16-bit images. Does it contain "valid" or "bogus" information? I think we should have a method for making this determination, or even further, to be able to measure how much of the extra data is valid and how much is bogus.
Another very important aspect of this test has to do with what the 8-bit or 16-bit numbers represent, that is, are they linear with respect to intensity or do they represent some companded form, such as gamma 2.2 or L*? This has a strong bearing on how the error differences appear to the eye (which is what the test is all about).
I also think the role of noise should be investigated. Images from scanners and digital cameras have noise, while most computer generated images do not. I suspect this is the reason for Dan's first Condition, although he never actually says that. Furthermore, converting a 16-bit image to 8-bits in Photoshop introduces noise into the image, as do transformations through profiles and mode changes. In the context of color correction, noise helps "break up" banding (dithering) that may otherwise occur, but this comes at the expense of a noisier image.
An assertion is a statement made without supporting evidence. An argument is a statement made with supporting evidence. In my opinion, Dan tends toward the former. Granted, he is a very experienced and knowledgeable person who has a large and faithful following, but at least in his infamous "16-bit Challenge," he is a fish out of water, showing no understanding of the Scientific Method.