ung. Wenn 
nit blossen 
m Betrach- 
Umständen 
vr ein. EW 
in ich den 
ld normal, 
. Wenn ich 
Brennweite 
dlich — die 
enfalls un- 
h 1,91 für 
>s hat Herr 
en 18/12, 
weite, 2/4, 
Ohungsfak- 
enn ich das 
2/5 = 2/0, 
g ähnliche 
genau das- 
it dem Un- 
meiner an- 
lgen, wäh- 
emacht hat 
n Zeit hal- 
| und kann 
. darstellen. 
only about 
wy, and un- 
ie rest of it 
yncerned is 
tion. I wish 
o eliminate 
| perception 
ocular fac- 
ı more im- 
hich exists. 
d with here 
n of recog- 
at part of 
h is trou- 
e to monu- 
:ated by the 
ve have two 
ular vision, 
stion: what 
it that we 
ind at once 
little about 
Ie is not at 
in fact, the 
comodation 
ion. This is 
  
  
  
something you can verify by experiment, and 
I shall not say very much about it. Some of it is 
in the paper I have presented. 
The next question which arises is: what is 
the nature of this perception? However, long 
before we ask this question, we can do some- 
thing extremely practical. If we know what is 
the nature of the information that is utilised in 
forming a perception of depth, if we wish to 
reconstitute the same perception of depth, all 
we have to do is to supply the eyes with the 
same information, and this is what we hope to 
do in a stereoscope. 
Now the problem of photogrammetry is more 
complicated because, as a matter of fact, it is 
absolutely impossible to make a complete over- 
all correction of the entire physical world in this 
kind of reconstitution which will give you the 
identical impression that the original object 
would have given from some point of view. In 
this paper I have chosen to take the following 
sort of ideas as a suitable correction. One 
should examine the transverse extent of the ob- 
ject which is seen through the stereoscope and 
ask what would be the apparent depth if you 
were in such a position unaided as to receive the 
same transverse impression. One can then find 
out two things: first, what the real situation 
looks like in such a condition; then one can 
also ask what correction best approximates this. 
Without going through the long analysis 
which would be required exactly to describe the 
distortion, permit me to show you one situation 
where I take a pair of six pole binoculars. In 
this case, the magnification of the base line — 
that is the factor by which you multiply the base 
line of the eye to get the separation of the ob- 
jective — is somewhat bigger than two. Clearly, 
it is out of line with the magnification of the 
lateral dimension since you interpret the base 
line as the only means of seeing depth. It would 
be wrong to say that all one has to do is to 
choose the same factor for the optical base line 
as you have for the ocular. The same magnifi- 
cation factor for the optics cannot be applied 
directly to base line because it is not true that 
this will correct the distortion completely. There 
will always be some residual distortion. In point 
of fact, if you have a very deep relief you have 
progressive foreshortening in approaching the 
observer. 
Let me show you two slides [not reproduced 
here] very quickly to show you what one can 
do. The first is a picture of the following type. 
You are looking through binoculars whose ob- 
jective separation is indicated by L star and R 
Star. These are six pole binoculars at an objec- 
QUELQUES QUESTIONS D'OPTIQUE PHYSIOLOGIQUE, DISCUSSION 
tive 50 metres away, which has a radial depth 
of 5 metres and a lateral separation of 2 metres 
at the front. 
One may multiply the base line by suitable 
factors to get the separation of the objectives, 
and this I call a corrected instrument. The im- 
pression one gets can be mapped through this 
corrected instrument. The real impression which 
best fits the magnification is shown from a 
given distance which is 123 metres, it is shown 
by the uninterrupted line. The dashed lines 
show the effect of correcting the instrument ac- 
cording to the criterion I have given you, and the 
dotted lines show the result of leaving the in- 
strument uncorrected. 
This, then, is in perceptual space. As you see, 
the correction is not perfect, nor can it ever be, 
but it is considerably better than the correction 
would be for an unmodified instrument. 
Professeur P. BAETSLE: Notre Président 
Monsieur Cruset va nous entretenir quelques 
minutes sur l'effet du chromatisme de l'oeil. 
Monsieur J. CRusET: Beaucoup d'études ont 
été faites depuis longtemps concernant le chro- 
matisme de l'oeil isolé et concernant les erreurs 
d'interprétation que ce chromatisme de l'oeil 
isolé peut provoquer sur, non pas la mesure, 
mais l'appréciation de distances. J'ai cherché à 
étudier l'influence de la composition de la lu- 
miére en vision binoculaire et pour cela je me 
suis attaché à mettre en évidence les aberrations 
chromatiques d'excentrement des centres des 
pupilles — vous savez que l'oeil est un instrument 
qui n'est pas parfaitement centré et on peut 
penser que la base de triangulation que notre 
paire d’yeux réalise en vision binoculaire est 
variable avec la radiation que nous examinons. 
Le phénomène est peu sensible, aussi ai-je 
cherché à réaliser artificiellement une dispersion 
des pupilles des yeux humains. La méthode qui 
vient d’abord à l’esprit consisterait à utiliser des 
feuilles de papier épais percées de trous et pla- 
cées devant les yeux. A mon sens cela aurait un 
grave inconvénient, celui de donner, puisque ces 
trous ne seraient pas dans le plan des pupilles, 
un effet directionnel aux yeux. Aussi ai-je pré- 
féré utiliser une paire de jumelles trés bien cor- 
rigées du chromatisme de facon à mettre en 
coincidence les pupilles de sortie de ces jumel- 
les avec le plan des pupilles de l'oeil, et à obte- 
nir une base variable en diaphragmant la pupille 
des yeux au moyen des cercles oculaires des 
jumelles mis à différents écarts. Le test utilisé 
était constitué par une boite à lumière dont la 
face antérieure était noire opaque à l'exception 
FR 
——————Á 
  
  
  
OEIL