Let's continue with the red eyes' problem.
The physical contradiction that we obtained is (with a slight re-formulation which expresses better the reality) :
PC : The pupils are dilated because of the relative darkness, and the pupils should be non dilated in order to avoid the red eyes' phenomenon
There are canonical heuristics for the resolution of physical contradictions. As mentioned in "Engineering of creativity" (Savransky, 2000), Altshuller developed 11 such heuristics. Some TRIZniks extended this earlier set of heuristics. Among these one finds 4 separations methods which help eliminating physical contradictions, and two of them are very often used :
* separation in space : one part of the subsystem has property A and the other part has property non-A
* separation in time : the subsystem has property A during one period and afterwards it has property non-A
In our problem :
* subsystem = pupils
* property = dilation
* system (minimal) = eyes (including pupils) ; photo camera (including flash)
* harmful effect = red eyes' phenomenon
Is it possible to separate the dilated / non dilated states of the pupils ? Yes.
How ? In time.
Actually the relative darkness is underwent : it is part of the environment. But it is possible to induce when one wants the contraction of the pupils with a given resource of the system : the flash of the camera. Finally, in a chronological sequence :
* the pupils are originally dilated because of the relative darkness
* the photographer pushes the button, and a first flash occurs
* consequently, in a short time frame, the pupils contract
* automatically, after a pre-set short time period, a second flash occurs during the taking of the photograph
* consequently, the pupils contract (but this is not used ...)
* the pupils go back to their original state : dilated because of the relative darkness
The initial problem has been solved, and an additional positive effect may occur : the first flash desensitizes (partly in terms of probability) the propency of eyes to blink under a flash ; in other words, if the eyes are sensitive to the flash, they will blink during the first flash.
Note that we face now another problem created by this solution, that we may express as a pair of technical contradictions :
* TC1 : If one uses the double flash, one avoids the red eyes' phenomenon, but the photography is taken with some delay after the button pressing
* TC2 : If one does not use the double flash, the photography is taken when the button is pressed (**), but the red eyes' phenomenon is not avoided
(** : almost, indeed)
How to solve this newly induced problem ? This new problem is open.
Note also that in some conditions, professional photographers orient the flash towards the ceiling, so that :
* the eyes blink with a much decreased probability
* the red eyes' phenomenon does not occur
* additionally the photograph is of high quality, for the bodies' skins are not whitened by the flash
Is it possible to solve the newly appeared problem, in any darkness conditions, and for any amateur photographer ?
The physical contradiction that we obtained is (with a slight re-formulation which expresses better the reality) :
PC : The pupils are dilated because of the relative darkness, and the pupils should be non dilated in order to avoid the red eyes' phenomenon
There are canonical heuristics for the resolution of physical contradictions. As mentioned in "Engineering of creativity" (Savransky, 2000), Altshuller developed 11 such heuristics. Some TRIZniks extended this earlier set of heuristics. Among these one finds 4 separations methods which help eliminating physical contradictions, and two of them are very often used :
* separation in space : one part of the subsystem has property A and the other part has property non-A
* separation in time : the subsystem has property A during one period and afterwards it has property non-A
In our problem :
* subsystem = pupils
* property = dilation
* system (minimal) = eyes (including pupils) ; photo camera (including flash)
* harmful effect = red eyes' phenomenon
Is it possible to separate the dilated / non dilated states of the pupils ? Yes.
How ? In time.
Actually the relative darkness is underwent : it is part of the environment. But it is possible to induce when one wants the contraction of the pupils with a given resource of the system : the flash of the camera. Finally, in a chronological sequence :
* the pupils are originally dilated because of the relative darkness
* the photographer pushes the button, and a first flash occurs
* consequently, in a short time frame, the pupils contract
* automatically, after a pre-set short time period, a second flash occurs during the taking of the photograph
* consequently, the pupils contract (but this is not used ...)
* the pupils go back to their original state : dilated because of the relative darkness
The initial problem has been solved, and an additional positive effect may occur : the first flash desensitizes (partly in terms of probability) the propency of eyes to blink under a flash ; in other words, if the eyes are sensitive to the flash, they will blink during the first flash.
Note that we face now another problem created by this solution, that we may express as a pair of technical contradictions :
* TC1 : If one uses the double flash, one avoids the red eyes' phenomenon, but the photography is taken with some delay after the button pressing
* TC2 : If one does not use the double flash, the photography is taken when the button is pressed (**), but the red eyes' phenomenon is not avoided
(** : almost, indeed)
How to solve this newly induced problem ? This new problem is open.
Note also that in some conditions, professional photographers orient the flash towards the ceiling, so that :
* the eyes blink with a much decreased probability
* the red eyes' phenomenon does not occur
* additionally the photograph is of high quality, for the bodies' skins are not whitened by the flash
Is it possible to solve the newly appeared problem, in any darkness conditions, and for any amateur photographer ?
So in this new problem, you are removing the time dimension. So the solution must be based on the separation in space.
ReplyDeleteBecause of the flash the pupil is dilated. And it looks red on the picture.
Can the pupil be dilated, but the eyes not appearing red on the picture? Yes, by processing the picture after its taken manually or automatically in the camera.
Thank you Stephane for this extremely clear example of a physical contradiction resolution ! As quite often, the solution looks simple once it has been explained, but I guess it took some time to invent it at first. The new problem looks tricky though !
ReplyDeleteDear Bruce, sincerely I do not know the amount of time which was necessary to come to this idea which solves the initial problem.
ReplyDeleteIt is true that the solutions seem often (not always) simple once explained. This phenomenon illustrates what TRIZ calls "psychological inertia", concept which will be later introduced on this blog.
Till now, I do not have thought thoroughfully ( = systematically) about the new problem. Then I do not know if it is tricky. Maybe it is not. Try to apply the same approach, think in terms of functions, and propose your solution(s) !
Dear Serge, the time dimension is not necessarily removed. Please examine the different functions needed in the considered system.
ReplyDeleteHowever I accept your solution of correcting the red eyes afterwards with the help of a software like Photoshop. But this solution is not ideal :
* you have to guess the colour of the iris which will replace the red colour
* it is not adapted for people who do not know how to use such software
* it is not adapted for a part of the people who still use film cameras instead of numerical cameras
Which solution could be closer to ideality ?
Try to separate in space.. If your ligntning source is not aligned with the camera, you can make disappearing the red eye effect on your portrait. So we separate in space the two elements that are necessary to take the picture.
ReplyDeleteAnother way is to remove by itself the red eyes when shooting ... lest imagine, why not include the function of image manipulation software on the camera? near from ideality :-)
Dear Erwan, your first solution is interesting. It is close to the solution used by professionals (flash towards the ceiling).
ReplyDeleteAs Serge proposed before in the first commentary, your second solution is OK. If the correction is automatic, then it is close to ideality.
But I expect another, more ideal solution, where the photograph is not manipulated, or processed after it has been taken.