“Undoctored” photographs
For the first time in history there are billions of images that look like “undoctored” photographs but are not. What exactly is an “undoctored” photo, and why do they matter?
-
601. Why does people care whether realistic-looking images put before the public are “undoctored” photographs?
That is addressed in this brief.
-
602. Why aren’t TTG photographs called “unmanipulated” instead of “undoctored”?
Because no digital photographs are completely “unmanipulated.”
The term “unmanipulated” is basically meaningless, because every digital photograph — including newsphotos — undergoes numerous “manipulations” instantly, inside the device, before the photographer even sees the image on the review screen (see #4 and 5 in this brief).
-
603. If TTG allows numerous changes (or “manipulations”) to be made to TTG photos, then what defines an “undoctored” photo?
It’s basically a photo that shows “nothing other than what the camera lens saw at the moment the picture was taken.”
The official definition of “undoctored” is here.
-
604. Aren’t there are a lot of arbitrary definitions floating around out there for the term “undoctored”?
It often seems that way.
But nothing about TTG is arbitrary, including the derivation of the definition of the term “undoctored.”
-
605. What’s the difference between an “undoctored” photograph vs. an “unaltered” photograph?
An “unaltered” photograph is unchanged from what comes out of the camera after minimal in-camera processing.
An “undoctored” photograph can undergo any and all of TTG's Allowable Changes and still be considered “undoctored.”
(As explained in this brief, like newsphotos TTG photos must be “undoctored” as per P2, but they need not be “unaltered.”)
-
606. If “unaltered” is a stiffer non-manipulation standard than “undoctored,” does that mean that all “unaltered” photographs qualify as TTG?
No, because completely unaltered photographs can still be deceptive; see #613 below.
See also #10xx.{the VUO question on why VUO tech needs TTG}
-
607. Why not just label photographs as “undoctored” instead of as “TTG”?
For two reasons:
1. Because without TTG behind the definition, different photographers' interpretations of the term “undoctored” could allow any manipulations imaginable.
2. Because there are 8 other characteristics (besides “undoctored”) that affect viewers’ level of trust. It is easy to deceive viewers with an undoctored photo (see #613 below), so merely saying that a photograph is “undoctored” does not mean it is “trustworthy.”
-
608. Why does TTG say it draws the line on photo manipulation based on the behavior of “light”?
Because whether a photograph remains “undoctored” depends in large part on whether changes made to the photograph are “light”-related or non-“light”-related.
In ensuring that a photograph remains “undoctored,”
— P2 allows whatever “light”-related changes are necessary* to ensure that the photo meets P7 for not misrepresenting the appearance of the scene pictured...
— but P2 allows for only a few categories of non-“light”-related changes (see the list here).
-
609. Why are “light”-related changes treated differently than non-“light”-related changes?
That is explained in the background brief on Light.
-
610.Why do photographers so often want viewers to believe that doctored or aigmented photographs are actually undoctored photos?
-
611. Why are single-exposure, undoctored photographs the reference point when discussing trustworthy photographs?
Because all around the world those have historically been the most-widely trusted kind of photographs.
-
612. The TTG label is obviously only for images that are not doctored. But why not label images that ARE doctored instead of labeling images that are NOT doctored?
This brief explains why such a label would be unrealistic and unworkable.
-
613. What about photographs that are not “doctored” but are “deceptive”?
Photographs need not be doctored — or misrepresentative of a scene’s appearance — to be “deceptive” about inapparent circumstances behind the photo (that’s been the case since photography was invented; it’s not just a “digital” thing).
A list of common examples of photos that are potentially-deceptive-even-when-undoctored is under {“B”} in the “IC” alerts guide.
FAQ #7 looks at one cause of many potentially deceptive photographs, “subject manipulation.”
-
614. What about photographs that are not “deceptive” but are “doctored”?
A photograph need not be “deceptive” to be disqualified by TTG.
Merely being “doctored” — that is, failing to meet P2 — disqualifies from TTG millions of photographs that might not be deemed “deceptive.”
TTG is based on what causes viewers to trust some photographs more and other photographs less, and viewers are less likely to trust a photograph if they know or suspect that it is doctored. (More on this)
-
615. What if circumstances beyond a photographer’s control cause changes that fall under the definition of “doctoring”?
The reasons don’t matter. If a photo doesn’t fully meet the Trust Test, it is disqualified from TTG. Period.
To use an American idiom, viewers don’t care to hear about “how the dog ate your homework”; they only want to know whether a photo that looks undoctored is undoctored.
Common rationales that don’t cut it with TTG
-
616. Is it possible — using any and all of TTG’s Allowable Changes — to convert a doctored photo into an undoctored photo that will be eligible to qualify as TTG?
TTG’s Allowable Changes
TTG’s definition of “doctored”
It usually is not possible to “undoctor” a doctored photograph because the manipulations required to “undoctor” it would disqualify the image from the Trust Test. But...
• If the unaltered original was saved, the photographer can work with that.*
• If the photo was taken on a smartphone, the photo can often be Reverted on that phone (though not reliably on another device).
• If the disqualifying factors are “light”-related, they can usually be sufficiently corrected or the image can be converted to monochrome.
___________
*The safest thing for any TTG photographer to do is to save a completely unaltered copy of the original photo.
-
617. Why is TTG’s definition of “undoctored” so strict?
(The definition is here.)
TTG’s definition of “undoctored” may seem strict to some, but the TTG label can reassure viewers that a photograph is “undoctored” only when viewers are confident that there is no stricter definition of “undoctored” in wide use.
Viewers will not regard as “undoctored” any photographs made using a definition of “undoctored” that is less strict than their own definition.
That is why, if a definition of “undoctored” does not incorporate the strictest aspects of all commonly accepted definitions of the term “undoctored,” then various audiences will dismiss the definition as not strong enough for them to trust.
Could I create a standard that is like TTG but less strict?
-
618. Why isn’t TTG’s definition of “undoctored” more strict?
(The definition is here. Some photographers have expressed the opinion that only single-exposure photographs can be trusted, for example, while TTG allows combining exposures if numerous conditions are met.)
The answer to #618 is “because the Trust Test was designed to be a helpful tool for real-world use, not mere theory.”
More
-
619. Why does the public increasingly assume that impressive photographs have been “Doctored or aigmented unless labeled otherwise”?
Because people expect that a “non-news” photographer will improve a photograph’s appearance in every possible way that cannot be detected by viewers.
Outside of “news” settings, the more impressive a photograph looks, the more likely viewers are to assume that it is doctored or aigmented.
TTG can play an important role in identifying exceptions to the public's assumptions about impressive photographs; see how TTG “levels the playing field” in #4 here.
-
620. How do photographers feel about assumptions that their best photos are “Doctored or aigmented unless labeled otherwise”?
Interestingly, those who know best how to doctor photos without detection — advanced photographers — often encourage the public to assume that non-news photos are “Doctored or aigmented unless labeled otherwise.”
More
Questions #1201-1203 discuss the terms “undoctored” and “doctored” in relation to P2.
The numbering of the FAQ questions will not change — any new questions are added at the bottom and given new numbers — so users can safely make a link to any specific question.