Apollo 11 Photogrammetry
Photogrammetric Analysis Of Apollo 11 Imagery:
New camera-station map with improved locations
Copyright (C) 2010 by Vladislav Pustynski and Eric M. Jones.
All rights reserved.
Last revised 16 December 2013.
During 2009,& one of us (VP)& got interested in precise
photogrammetry of A and, through several Russian
forums, got an opportunity to collaborate with a person nicknamed
"N.A.", who was using photogrammetric software ImageModeler
to identify and locate several boulders visible in the direction
of West Crater in the Apollo 11 surface photography.& Although, ImageModeler is
intended for architectural purposes, but it is useful for photogrammetry of
the Apollo images.& It has a few drawbacks.& For example, it sometimes has
trouble with photos containing only benchmark points distributed
on a relatively planar surface - such as parts of the lunar
surface - a problem that can be remedied by the introduction of
virtual benchmarks a few meters above the surface. ImageModeler also can
become stubborn or even unstable when the number of features in
the scene under construction gets large.&
Nonetheless, ImageModeler is excellent software for obtaining good
quantitative results in a simple way.
in the Apollo
11 Preliminary Science Report is a camera-station map
compiled by R.M. Batson and K.B. Larson at the USGS soon after the
mission.& (A camera station is the location of the camera
when a particular photo was taken.)& The map is dated 11
August 1969.& See, also, a re-drafted, color version by ALSJ Contributor
Thomas Schwagmeier.& The methods used by Batson and Larson
were necessarily simplistic but were chosen to allow rapid
compilation of a map adequate for the Preliminary Science Report
and for other early studies of the landing site.& They wrote
in the Preliminary Science Report:
The surface traverse and sampling activities of the
astronauts have been reconstructed from clues provided by the
voice transcript, from review and analysis of the lunar
television pictures, from analysis of the 16-mm pictures taken
with a time-sequence camera mounted in the LM cockpit, and from
detailed study of photographs taken with Hasselblad cameras
before, during, and after the EVA. The camera stations for
Hasselblad survey panoramas taken on the EVA were located by
photographic triangulation from mosaics. Horizontal angles
between the LM footpads were measured on the photographs as a
function of& the known field of view, and the angles were
drawn on tracing paper. The paper was then& manipulated
over a scale drawing of the LM until the lines intersected the
appropriate pads& at the proper place. Once the panorama
locations had been determined, azimuths were measured from two
or more panoramas to conspicuous features on the surface, and
the positions of the features were plotted by triangulation to
produce the map of figure . Individual photographs were located,
and their orientations were measured by similar methods, using
for control both the LM footpads and other features on the lunar
surface that had been located by triangulation from the
panoramas.
Because the graphical method by which these data have been
obtained is fairly crude, azimuths shown for individual frames
may have errors of 3° or more. Positions of most of the camera
stations are probably within a 1.5-m circle centered at the
point shown. The determinations are sufficiently accurate,
however, to provide a useful control net for an overall view of
the astronauts' traverse and a starting point& for more
rigorous analytical photogrammetric measurements.
A 1978 USGS/Defence Mapping Agency site map, discussed in section 2.3, is probably derived from the "more
rigorous analytical photogrammetric measurements" for which time
became available in the years after Apollo ended.
Visual study of the 1969 map convinced VP that it is quite schematic and not very precise.
An initial photogrammetric
analysis with ImageModler produced improved locations and
orientations for each image, along with locations for
boulders and the US Flag, TV, LRRR, and PSEP.& The initial
analysis made use of 116 of the 123 photos Neil and Buzz took
during the EVA.&&As
with the work of Batson and Larson, all locations are calculated
relative to structures on the LM.& Of the seven images that were not used, 5904 is an accidental image of one of
the spacesuits, and is the only one of the 123 images that can
not be placed by any means.& Five of the others are 5876
to 5880, photos documenting the bootprint penetration experiment
Buzz did for the soil mechanics experimenters.& As discussed
in Section 5.2, the two bootprints Buzz made at this spot have
been identified in photos taken later in the EVA and out Buzz's
window after the EVA.& This allows placement of the five
documentation photos, which are shown with empty arrowheads.&
Finally, 5966A is the second of seven images exposed when Neil
advanced the film at the MESA, before removing the magazine.&
It shows only an underexposed, blurred image of the MESA interior
but can be placed with confidence close to the MESA. 5966A is also
shown with an empty arrowhead.
Of the 116 images used in the
initial photogrammetric analysis, the eight photos Neil took
near the southwest rim of Little West, 5954-61, were not
accurately placed because the locale is up-Sun of the LM
and because only distant benchmarks can be identified in the Pan 5
images.& However, as is
discussed more fully in section 2.1.5, these eight stations can be
placed with good accuracy using local photogrammetry and specific
features seen in both the Hasselblad images and LROC images of the
site.& The Pan 5 images are shown with empty arrowheads.
In summary, accurate camera-station locations were determined in
the overall photogrammetric for a set of 108 out of the 123 photos Neil and Buzz
took during the EVA.&
Seven of the 108 are LM
inspection photos that were not shown on the 1969 map.& Of
the remaining 15 images, 14 could be placed using local
photogrammetry and/or other information.& Of the 123 EVA
photos, only 5904 could not be placed at all.
Note that the map also
includes the rims of Double Crater and eight small small
craters.& These rims were defined using photgrammetric
determinations of the locations of small rocks on the
rims.& These craters are discussed in Section 4.4.
This 24 January 2012 revision
of the Apollo 11 camera-station map is based on
photogrammetric analysis done for 116 of the 123 Hasselblad images
taken during the EVA.& Six others were placed by
alternate means.& Only one photo, 5904, could not be
placed.& The arrows show
azimuthal pointing of the images, with arrow length
indicating vertical tilt.& Each small dot represents
a location for a camera station (blue) or a rock (red).& Boulders labeled with
Roman letters can be identified in LROC images. For Pans 1, 2, 3, and 4,&
all of the camera locations and orientations were
calculated in the overall photogrammetric analysis. The
Pan 5 stations were determined in a separate analysis
discussed in Section
2.1.5.& In each panorama, the
central location of the stations is plotted as a violet
dot.& A violet circle
shows the characteristic distance of the
stations from the pan's central point; and the dashed circles
indicate mean-square deviations.& Because of the
large number of camera stations near the LM ladder, Pan 1
information has been moved to an empty area northeast of
the LM.& See,
also, an& between the original 1969 map
and the revision.& The major differences are: (1) the locations of the
LRRR and PSEP and associated camera stations, which are
farther south of the LM in the new map than in the
(2) camera station locations near the MESA (Quad
IV) on the northwest side of the LM; and (3), the detailed
distribution of camera stations within each panorama are
now available. (Click on the image for the
full map in PDF format.)
Batson and Larson estimated that
camera station locations in the original 1969 map were accurate to
about 1.5 meters and that azimuths were accurate to 3
degrees.& Comparing the original map with the revision,
although we see that there are location differences great than 1.5
m, in many cases it is a good estimation, especially for camera
stations not far from the LM. As for the 3-degree azimuth-error
estimation, in most cases photogrammetric azimuths differ by less
than 3 deg from the 1969 determinations. Azimuthal differences are
larger mostly for camera stations& with large location
differences between the 1969 and 2010 results.& The
accuracy of camera-station locations may be lower for shots
where only relatively distant boulders are visible, particularly
in the case of
photos taken toward the east.&
As mentioned above,& the special case of Pan 5 required a
separate photogrammetric analysis and is discussed in detail in
Camera stations for the
five panoramas.& For each pan, the central
location of the stations is shown with a violet dot
and the characteristic distance from that central
point is shown with a violet circle.& Mean-square
deviations are indicated by dashed circles.&
Azimuth arrows for photos taken by N
those taken by Buzz in green. (Click on the image for
a larger version in PDF format.& See, also, an overlay on
the Dec 2009 LROC image, MR)
The following table lists the panoramas and the camera stations in
Pan No. and Location
No. of frames
Frame Sequence (AS11-40-)
Characteristic distance
from central point (meters)
Mean-square deviation
Pan 1, Foot of LM ladder
Pan 2, 10m west of ladder
on the rim of Double Crater
Pan 3, 10m north of
plus-Y(north) footpad
Pan 4, 20m southeast of SEQ
Pan 5, rim of Little West
Crater (partial)
While taking a panorama, the astronaut took one frame and then
turned about 30 degrees before taking the next frame.& Except
for the special case of Pan 5, all the pans were taken in
clockwise order, with the astronaut turning to his right between
On a relatively level surface, the astronaut is able to turn
without moving much off one spot on the ground. If we could see
the pattern of footprints made while the pan was taken, it would
be a tight pattern.& However, because both Neil and Buzz took
their pans with the Hasselblad out in front of the RCU - Neil with
the camera on the RCU bracket and Buzz holding his camera out in
front - camera stations (identical to the center of the film
plane) will be on a circle 0.5 to 1.0 meters in diameter around
the footprint pattern.& Among other things, the outward
displacement of the film plane from the turn axis will generally
produce some stereo separation between sequential frames.&
Remapped images generated with panorama-stitching software Hugin
give an impression of the direction and amount of the stereo
separation and, as well, can indicate cases in which the astronaut
moved forward or back between frames.
Pan 5 was taken in a very different manner.& Neil wanted to
get a portrait of Little West Crater and was in a hurry.& He
took one photo, moved about a meter east and took three more,
turning to his left between them, and then moved another meter
east and took two more, turning to his left between them.& He
then moved west and took two final frames toward the west, turning
left between them.& Two pairs of images give excellent
2.1.1 Pan 1
Pan 1 camera-station map,
with the location of the MESA-mounted TV camera
marked (x).&
Times when
Neil turned to his right between frames are
discussed in the text.& (Click on the image for a
larger version.)
Neil took this pan just west of the ladder.& Pan 1 is the
only one of the five pans for which we have a TV record.&
The Apollo 11 EVA video was digitally restored in
2009, which gives us good detail of Neil's
movements.& Colin Mackellar has produced a 2 min 31 second
video clip ( ) which starts at about 109:30:49, four seconds before Neil tells
Houston, "I'll step out and take some of my first pictures
here."& Neil begins by adjusting the f-stop and/or focus on
the Hasselblad, then takes up-Sun photo AS11-40-5850.& He makes the first turn to his
right 49 seconds after the start of the video clip and then,
five seconds later, sidesteps a little farther from the
MESA-mounted TV camera.& This sidestep may explain the
relatively large WSW displacement between the 5850 and 5851
stations. In a , the 5850
station is an outlier.& Neil's subsequent turns are at
1:08, 1:22, 1:32, 1:36, 1:50, 1:55, 1:59.& Because this was
Neil's first lunar panorama, he may have taken his time while he
was taking the first few frames.& The interval between the
first turn/step and the next turn is 14 seconds.& The next
two intervals are 14 and 10 seconds.& With one exception,
subsequent intervals between turns are 4-5 seconds, perhaps
indicating that Neil had gained confidence that he could turn
between frames without trouble.& The exception is the
14-second interval during which Neil took 5855. As indicated on
a , photographed at the National Air and
Space Museum in 2006 by Ulrich Lotzmann, Neil planned to change
the f-stop setting from 5.6 to 8 or 11 when taking down-Sun
photos. In the video, it appears that, in preparation for taking
5855,& he increases the setting at about 1:39,& takes
5855 at about 1:44, and goes back to 5.6 after 5855 at
1:45.& The unprocessed scans are available on the LPI website.
On the first lunar pan, Neil got& significant overlap only
in the sequence 5855-58.& These four images have been
combined as a .&
A side-by-side comparison of the remapped images generated by
panorama-stitching program Hugin has been made by horizontally
aligning a horizon feature near the center of the overlap region
for each pair of images. In each pair, a vertical white line is
drawn downward near the center of the overlap region in the
lefthand image.& A horizontal white line is then drawn low
in the left hand image.& In the righthand image, yellow
traces of the white lines are drawn from rock-to-rock,
shadow-to-shadow. & An anaglyph made from a pair is
displayed between the two images.& For these three pairs of
images, the yellow traces of the vertical lines slope down to
the left, indicating that, as expected, the camera station for
the righthand image is to the right of the camera station
corresponding to the lefthand image. & Terrain,
particularly relatively large craters, can produce local
distortions of the yellow trace of the vertical white
line.& In lower part of each righthand image, upward
displacement of the yellow traces of the horizontal white line
in the lefthand image would indicate that Neil moved back
between a pair of frames.& Downward displacement would
indicate that he moved forward.
Side-by-side
comparisons of images 5855-58, indicating that, as
Neil turned to his right between frames, the
center of the film plane also moved to the
right.& In these three pairs, Neil got enough
overlap to provided unintended stereo. An anaglyph
is displayed between each pair.& Note the
effect on the vertical yellow line of the crater
on the left in 5858.& Comparison of the
horizontal white and yellow lines indicate that
Neil did not moved noticeably forward or back
during this sequence.& (Click on the
image for a larger version.)
2.1.2& Pan 2
Pan 2 camera station
map.& The distribution of stations is elongated
roughly along the trend of the raised crater rim,
indicating that Buzz was careful where he stood.&
The rim of the
western component of Double Crater is drawn at the
outer edge of the relatively steep inner wall.&&
(Click on the image
for a larger version.)
Buzz took this pan near the
southern edge of the LM shadow about 10 meters from the
pluz-Z footpad.& During the post-flight Technical Crew Debriefing, he said that
he took the pan with the camera handheld. In the next
figure, the camera stations are plotted on a detail from the
deconvolved
version of the& 22 December 2009 LROC image,
MR.& This part of Double Crater has a
raised rim, as can be seen in a mini-pan
assembled from three frames from Neil's ladder pan, AS11-40-5852-54.& The footprints he made
while taking the pan can be seen in oblique views provided
by various frames he took out his window after the
EVA.& An example is a detail from
A11-37-5488, which is shown along with the camera
station map with scaled bootprints added.& The sole of
the boot is 33 cm long by 15 cm at its widest, so a 3 x 3
array of bootprints would cover a rectangular area about 1.0
by 0.5 meters on a side.& Other than the elongation of
the overall pattern, there are no outliers in the
distribution of stations.& Details are listed in
Superposition of the Pan
2 map on a deconvolved detail of LROC image
provided by GoneToPlaid and a detail from MLR.& MR was taken with the Sun
low in the east and emphasizes the break in slope at
the top of the inner, eastern wall of the
crater.& The elongation of the pattern of
camera stations along the crater rim suggests
that Buzz was careful about where he stood.& (Click on the
image for a larger version.)
The 22 December 2009 LROC image
was taken with the Sun 8.2 degrees above the eastern horizon and
shows the eastern half of the crater filled with shadow.& Pre-EVA
(AS11-39-5756) and post-EVA (AS11-37-5501) photos taken
out Neil's window show the crater with the Sun 10.9 and 15.1
degrees above the horizon, respectively.& In the pre-EVA
photo, some of the inner, eastern wall of the crater is in
shadow, whereas the post-EVA photos, taken 8 hours later, show
very little shadowing,& These images suggest that the inner
slope at the eastern end of the crater is 10 to 15 degrees.
There are two sequence of frames showing adequate stereo
separation:&
() and 5888-90 (mini pan).
Side-by-side comparison
of frames 5882a to 5884, showing the view from
northwest to north. Although the camera station map suggests
that Buzz took 5883 forward of both the 5882a and 5884
locations, comparison of the horizontal white and
yellow lines within each pair do not support that
conclusion.& The indicated relative error of
photogrammetrically-determined 5883 location is
about 1 meter.& As expected, both vertical yellow lines
slope downward to the left.& (Click on the
image for a larger version.)
The 5883 station position may contain error of ~0.5 m due to a
slightly-wrong focusing distance set at the time of its addition
to the IM scene model. IM refuses to accept a correction.
Side-by-side comparison
of 5888-90 showing the eastern component of Double
Crater and
part of the western component.& The yellow
horizontal line in 5889 is lower than its white counterpart in
5888, indicating that Buzz moved a bit forward
between the two frames. Note the significant
effect of the craters on the shapes of the vertical
yellow lines.&
(Click on the image for a larger version.)
2.1.3& Pan 3
Pan 3 Camera Station
Map.& Buzz took this pan about 8 meters north
of the plus-Y footpad, on the rim of an 8-m crater.
from the& 22 December 2009 LROC
image shows the camera station location relative to
the crater.&
(Click on the image for a larger version.)
As can be seen
Neil's Pan 4 photo AS11-40-5932, the Pan 3 crater does not have
a raised rim, so Buzz did not have to be quite as careful with
his footing as he was when taking Pan 2.& As was the case
with Pan 1, the stations are well ordered around the
mean-distance circle,& The exceptions are AS11-40-5915 and
5916.& Both show parts of the LM, so the station locations
are well determined. In 5916 we see the back of Neil's
PLSS.& Neil is just to Buzz's left, a bit closer to the LM,
and is using the Apollo Close-up Stereo Camera. What may have
happened is that, after Buzz took 5915, he realized he would
have to go around Neil on the north side and get far enough west
to finish the pan.& Neil may have backed in the
field-of-view just as Buzz was taking 5916.& Another factor
may have been a fist-sized rock on the south edge of the
mean-distance circle.& Details from the best photos we have
of this rock - ,
taken as part of Pan 1; and
and& , taken as part of Pan 4 - suggest that
it sits on the southwestern rim of a small, fresh crater.&
The rock and fresh crater may have contributed to Buzz's
decision to move around them on the north side.& In 5932,
note the patch of undisturbed soil east of the rock, which is
another indication that Buzz moved from the 5915 station to the
5916 station on the north side of the rock.
Side-by-side comparions for
5907-11 (), covering the Pan 3 crater, labeled "α" in the camera station
map.& The crater isn't very deep, so it doesn't
have as large an effect on the vertical yellow traces
as Double Crater did in the Pan 2 comparisons. The photogrammetric map
indicates that Buzz moved back between 5908 and 5809,
forward again before taking 5910.& The
side-by-side comparisons do not support that
result.& The implied position error is about 0.5
m.& (Click on the image for a larger version.)
Pan 4 Camera Station
Map.& Neil took this pan about 17 meters from the SEQ Bay on an
azimuth about 30 degrees south of east.& (Click on the
image for a larger version.)
Neil took this pan after taking
a few photos of Buzz offloading the EASEP packages.&
Details of this pan show the importance of benchmark
distance in determining camera station locations. As can be
seen in a ,
the five camera stations (, 40, and 41) for
showing parts of the LM and/or boulders near the LM are well
ordered along the circumference of the mean-distance
circle.& The other 7 camera stations in the pan rely on
distant benchmarks and are not well ordered.& As
detailed in the , the
mean distance of the 12 stations from the central point is
0.428 meters, with a standard deviation of 0.272
meters.& The five stations on the west side of the pan
have a mean distance of 0.310 meters and a standard
deviation of 0.054 meters.& The remaining seven
stations have a mean distance of 0.461 meters and a standard
deviation of 0.305 meters.& Neil demonstrated while
taking Pan 1 that he was able to turn in a systematic way
without trouble, suggesting that the distribution
differences between the two groups of Pan 4 camera stations
is due unavoidable photogrammetric uncertainties when only
distant boulders are available for use as benchmarks.
Note that the largest displacement of a camera station from
its expected position on the violet circle is about 1.0 m,
in the case of 5934.
With the exception of the up-Sun images, 5936 and 37, Neil
got enough overlap between adjacent frames for assembly of a
nearly complete panorama ().& Side-by-side comparisons
of the ten overlapping frames demonstrate that Neil took the
pan much as he did Pan 1, turning on about the same spot
with little motion forward or back between frames.
Side-by-side
comparisons of frames 5930 to 34.& In the
photogrammetric analysis, the 5933 and 5934
camera stations are not part of a well-ordered
sequence comparable to those of Pan 1 and
Pan 3.& The side-by-side comparisons
demonstrate that Pan 4 was, indeed, taken in a
very orderly fashion and, that, as expected, the
scatter in the photogrammetrically-determined locations
is due to the availability of only distant boulders for doing
triangulation.& (Click on the image for a
larger version.)
Side-by-side comparisons
for frames 5937 to 41, plus 5930.& This
tells the same story: Neil took a well-ordered
pan.& Note that Neil actually took 5941 from a spot to the
right of 5930.& To make a comparison with 5930
attached to
the end of the sequence,& the white lines were
drawn in 5930 and the yellow traces on 5941.&
Because of Neil's shadow, two vertical lines were
used. Note
that the vertical displacements of the horizontal
lines in 5930 and 5941 indicate that Neil ended the
pan a very
short distance back from where he started.&
(Click on the image for a larger version.)
Neil took Pan 5 about one
crater diameter from the rim of a 4.5-m crater just
inside the rim of Little West Crater, about 38&
This map shows the station locations determined in
the separate analysis done for Pan 5.& See,
also, a version superimposed on details from
LROC images MR and MLR.
& (Click on the image for a larger version.)
The Pan 5 station locations calculated in a full, 116-station
photogrammetric analysis had some obvious discrepancies,
particularly with regard to their locations relative to the
prominent 4.5-meter crater just a few meters north of the spot
where Neil took the pan.& These discrepancies were due to
(1) technical issues in using IM when the number of benchmarks
and (2) the fact that the only benchmarks visible in
photos 5954-59 were distant boulders.&
To remove these discrepancies, we performed a three-step process
to determine the correct station locations:
(1) We did a dedicated
photogrammetric analysis for
using about fifty
local rocks as benchmarks.& A labeled detail
show a set of fifteen of these
rocks. The analysis demonstrated that the relative positions
and azimuths of frames 5954 and 5956-59 differed only
slightly from those found with the Pan& photos included
in the full photogrammetric analysis.& The differences
are well within the range of expected errors and are not
shown.& The local analysis also showed that, although
the azimuth calculated for 5955 was virtually the same as in
the full analysis, the station location was virtually
identical to that of 5957, instead of being about 2 meters
NNE of 5957.& As discussed below, detailed
intercomparisons of the
images confirm that the
the three were taken about the same spot.
(2) The local photogrammetric analysis is not directly tied
back to images of the LM and, therefore, only gives relative
locations of the six stations.& However, Neil took Pan
5 near a sharp-rimmed, fresh crater just inside the
southwest rim of Little West Crater.& Consequently, we
included in the local analysis ten small rocks on
or near the rim of the 4.5-m crater.& The
photogrammetric locations of these ten small rocks was then
fit with a circle.& The pattern - circle and stations -
was then scaled so the radius of the circle matched the
4.5-m radius of the small crater and was then shifted to
place the circle over the crater rim.
(3) As a final step, stations 5960 and 61 were moved in
tandem along a line defined by the 5961 central azimuth
until the azimuth from 5960 to the true position of boulder
j2 agreed with the value of 15 degrees to the right of the
5960 central azimuth derived from the 5960 image.& The
two stations were moved 0.9 meters toward the other Pan 5
Animation illustrating
the three-step process used to place the Pan 5
stations at their proper positions near the 4.5-m
crater just inside the southwest rim of Little West
Crater.& Note that the red dot near the top
left in the figure indicates the actual position of
boulder j2, with the sunlit face on
the right (east) and the shadow on the
left.&& (Click on the image for a larger
The result can be confirmed by
noting that, in LROC images taken with the Sun above 30
degrees elevation, Neil's track out to Little West shows up
quite well.& Examples include MR,
23:50:40 UTC on 1 October 2009 with the elevation 87.8;&
03:50:15 on 1 March 2010 with the solar elevation& 62.3;
taken at 09:34 UTC on 5 November 2011; and& MR,
14:44:11 UTC on 25 November 2009 with the solar elevation
35.8.& Note that, with the Sun above 30 degrees or so,
there is relatively little surface-brightness variation.&
Increasing contrast to bring out subtle features like
astronaut tracks produces images that are "grainy" near the
resolution limit which, for three of the images is about 0.5
meters.& M was taken from 25-km altitude and has
a resolution of about 0.25 m. An animation compares the new Pan 5 station
locations in M, M, and M. &It
is pleasing that the new analysis puts Pan 5 at the end of the
Neil's track.
2.1.5.1& Camera Stations
As indicated in the map above, Neil began&Pan 5 with
5954, taken about 1.5 meters WSW of the ALSCC. &He then
moved about a meter southeast and took a set of three frames,
starting with 5955 which shows a view aimed a bit more
easterly than 5954, and then turned to his left to take 5956,
and left again for 5957. &Neil then moved about a meter
ENE and took 5958 and 59, turning left between frames.
&He finished the sequence by moving to a spot beyond the
ALSCC to take 5960, which shows the TV and other benchmarks
north of the LM, and 5961, which is aimed at the LM.
In the following, a series of side-by-side comparisons confirm
the relative locations of frames 5954 to 5959.& We begin
with two pairs - 5956-57 and 5958-59 - with excellent stereo
separation.
Side-by-side comparison
of AS11-40-5957 (left) and 5956 (right). As shown
in the photogrammetric map, 5957 was taken
slightly to the left of 5956. This is confirmed by
the fact that the yellow trace in 5956 of the
vertical white line slopes down to the left.&
In addition, there are only very small vertical
differences in the& intersections of the
horizontal yellow lines with the vertical yellow
line& compared with their white
counterparts.& This indicates that the two
camera stations have no appreciable forward or
backward position difference.&& (Click
on the image
here for a full version of the anaglyph.)
While the side-by-side
comparison of 5957 and 56 confirms the relative locations of
the two camera stations, the anaglyph made from the two images
shows two major breaks in slope which conceal portions of the
crater floor beyond each of them.& These breaks in slope
are on the right in 5957 and across most of the image in
Comparison of details
showing breaks in slope in 5957 (top) and 5956
(bottom).& The dashed red-lines approximate breaks in
slope.& (Click on the image for a larger
The breaks
in slope can be identified by three techniques.& First,
at some places there is a discernible change in image
clarity across the breaks& The images were taken at
74-foot focus, so the surface on the near side of a break in
slope will be more out-of-focus than the surface beyond.
Second, patterns of rocks and shadow on one side or the
other of a break in one of the images can be identified at
the same relative location in the other image.&
However, a pattern across a break in one frame will have a
lateral discontinuity in the other frame.& For example,
in 5956 a pattern beyond a break in slope will be displaced
to the the right compared to its location in 5957. And
third, rocks and shadows that are visible in one frame of a
pair may not be visible in the other if the latter was taken
behind the first.
One other pair of frames, 5958 and 5959, form an excellent
stereopair.& A
confirms& the
relative placement of the camera stations in the
photogrammetric map, with 5959 having been taken just to the
right of 5958 and with no significant forward/backward
displacement.&
for a full resolution version of
the anaglyph.
The photogrammetric map indicates that 5958/59 were taken to
the right and forward of 5957/56.& Image 5956 has no
overlap with either 5958 or 59, except right at Neil's
feet.& Image 5957 has a useful amount of overlap with
both 5958 and 59.
Side-by-side comparisons
of 5957 (center) with both 5958 (right) and 5959
(left). The
white vertical and horizontal lines in 5957 are
traced in yellow on each of the other
images.&& (Click on the image for a larger
The side-by-side comparisons
between 5957 and both 5958 and 5959 show, not surprisingly,
nearly identical yellow traces of the vertical line,
indicating nearly identical displacement of 5958/59 to the
right of 5957.& The fact that the yellow traces of the
foreground horizontal line both slope down to the right is
another indication that the 5958/59 stations are to the right
of 5957.& Both stations are also forward from 5957, as
can be seen by the fact that the intersections of the
foreground yellow traces with the vertical yellow traces are
lower (farther from the horizon) than the corresponding
intersection in 5957.& The forward displacement is large
enough to preclude good stereo of the foreground.
Comparison of details of
5957 and 5958 showing the 4.5-m crater. The dashed portions of
the yellow traces indicate parts of the surface that are visible in 5958
but are hidden in 5957 by the crater rim. This is a clear indication that
5958/59 were taken forward of 5957.& (Click on the image for a larger
The special photogrammetric
analysis done for 5954-59 shows that the 5955-57 stations are
tightly bunched. &We have seen that 5957 was taken a
short distance to the right of 5956, with enough separation to
give good stereo. &A comparison between 5955 and 56 (next
below) shows that the intersections between the yellow traces
in 5955 are slightly above the corresponding intersections in
5956, indicating that 5955 was taken slightly behind 5956.
Side-by side comparison
between 5956 (left) and 5955 (right).
(Click on the image for a
larger version.)
In addition, portions of rocks
visible in 5956 are hidden in 5955.& A detail
showing the break in slope that hides part of the central
mound in both images clearly shows that more of the central
mound is visible above the break in slope in 5956 than in
5957.& A second , covering the break in slope nearest
Neil, shows similar differences.& The fact that the
yellow trace in 5955 of the vertical white line is very close
to vertical indicates that the 5955 station is virtually on
the backward extension of the 5956 aiming azimuth.
The photogrammetric map shows
5954 and 5956 where aimed in the same direction, with 5956/7
having been taken about a meter to the right.& It appears
that 5956 may have been taken a short distance farther into
the scene than 5954.
Side-by-side comparison of
5954 (left) and 5956 (right).&
(Click on the image for a
larger version.)
The fact that the horizontal
yellow traces slope down to the right and the vertical yellow
traces slopes down to the left confirm that 5956 was taken to
the right.& In addition, the yellow intersections are
below the white intersections, indicating that 5956 was taken
forward (in the northeast direction) of 5954.& See, also,
of the anaglyph.
2.1.5.2& Camera Stations
Side-by-side comparison
of .& Stereo information
available from the image pairs is strongly influenced
by relative displacement of the camera stations
along the azimuths defined by the centers of the overlap
The center-line azimuth of the 5961/60 overlap is roughly 15 degree
and the center-line azimuth of the
5960 overlap is roughly 40 degrees north of
west. The anaglyphs have been rotated left so that
the red and blue images of prominent features line up
horizontally.& In each case, a rotation of 45
degrees gives reasonable stereo.& (Click on the
image for a larger version.)
Of the final two images,
5961 shows the LM and various benchmarks near it,
while& 5960 shows the TV camera near the left edge.
As indicated by the slight
leftward slope of the vertical yellow trace in 5960 (above),
5961 was taken slightly to the left of 5960.& The
upward displacement of the horizontal yellow traces in 5960
indicate that 5961 was taken a short distance farther into
the scene than 5960.& The leftward slope of the
vertical yellow trace in 5959 indicates that 5960 was taken
to the left of 5959.& The leftward displacement of 5960
from 5959 is significantly great than the leftward
displacement of 5961 from 5960.& Similarly, the upward
displacements of the horizontal yellow traces in 5959
indicates that both 5960 and 5961 were taken much farther
into the scene than 5959.
To summarize, comparisons between
pairs of Pan 5 images that have some overlap demonstrate that
the apparently haphazardous distribution of camera stations is
actually a reasonable representation of how Neil took this set
of photographs.& Specifically, Neil took 5954 near the
ALSCC, then moved a meter east, perhaps to get farther away
from the ALSCC.& He took 5955, turned slightly to his
left - moving forward a bit in the process, took 5956, and
then turned left again to take 5957.& Next, he moved
about a meter farther east and took 5958 and 5959, turning
slightly to his left between the frames.& Finally, he
moved toward the ALSCC,& took 5960, aiming a bit north of
the LM, and finished off with a shot of the LM, 5961.
General comments: Figure 3-15 in the Apollo
Preliminary Science Report (hereafter, the Preliminary map)
is in good general agreement with the new photogrammetric results.
Most differences between the two maps are within the ranges of
uncertainty estimated by Batson and Larson& Notable
exceptions include& (1) tilt angles of many stations shown in
the Preliminary map are e (2) some camera
stations are misplaced, such as , and 5967-70; (3)
azimuths are sometimes wrong, as with 5942-44; (4) distances from
LM to the PSEP and associated camera station
and (5) individual stations within each panoramas were not
determined by Batson and Larson.
: This is one of a few photos in Pan 1 and
Pan 2 that include in the field-of-view a very old, eroded
~22-meter crater just next to the Double Crater on the southeast.
This crater is visible in LROC images - such as a labeled
detail from 22 December 2009 - but its presence is difficult
to pick out in EVA photos. This crater is between the foreground
and the ~6-m& ι crater (seen just in front). Boulders
which are identifiable on LROC images, are near the horizon to the
left from the center, boulder a is to the right from the
center, the top of boulder s seen just below a to
its left.&
5855-56: A sharp, fresh crater is seen on the righthand edge of 5855 and at the center of 5856. & Photogrammetric analysis confirms
that this is the crater identified in a details from
LROC images ML and MR.& The crater is about
70 meters from the LM.& See, also, a labeled version
of a pan Buzz took out his window after the EVA and a
similarly-labeled
made from nine LROC images of the site.
:& The new locations show that Neil
moved about 3.8 meters north from the Pan 1 location before he
took 5859.
: In a map detail, the three camera
stations marked with red dots.& These indicate that Neil
moved slightly to his left after taking the first of the three
frames, then farther left and forward after the second.& A
side-by-side comparison of 5860 and 61 shows that there is no
significant forward or backward separation between the two.&
5859 has signifi 5860 but 5861 is
the only one of the three what is reasonably sharp.
:& These two images of Buzz
crawling out of the LM hatch were not shown in the Preliminary
:& A tight cluster of
stations,& less than 0.5 m from each other.
:& The bright boulder visible above
the MESA and beyond the inside edge of the north strut can also be
seen in 5903, the well-known, full-length portrait of
:& All three photos made from
nearly the same place, with only slight azimuth and tilt changes.
Made from the rim of the older, eastern component of Double
: During early stages of the
photogrammetric analysis, the 5871 station location depended
primarily on distant boulders.& During September 2010, a
number of very small pieces of rock were added to the benchmark
pool so that station locations could be determined for 5865 and
other close-up photos.& One of us (VP) realized that some of
these small rocks were in the 5871 field-of-view.& Forcing IM
to accept these new benchmarks in re-positioning 5871 required
painstaking work and about 40 recalibrations.& The net result
is that the 5871 camera station shifted about 0.5 meters to its
present location.& This exercise provides an
order-of-magnitude estimate of the uncertainties in station
location dependent solely on distant boulders.
: The azimuthal difference between these
two photos is less than 2 degrees, with Neil moving forward along
that line after taking 5872.& West Crater is near the horizon
on the lefthand side of 5872 () and can also be seen in 5873. The
landscape is so flat that it is difficult to estimate
distances.& As can be seen in a
from the 22 July LROC image, the
southern portion of Little West Crater lies in the same direction
as the portion of West Crater visible in these two images.& A
show Little West, as labeled.& The southern portion of Little
West blocks our view of the lower parts of various boulders
located between the two craters.& See, also, a discussion in
section 3.4 of a shadowed wall inside the east rim of an old,
heavily-eroded crater immediately southwest of West Crater.
: Several boulders identifiable on LROC
images are present.& The largest is boulder a.
:& These Boot Penetration Soil
Experiment photos are not included individually in the camera
station map because determining the locations would be a very
tricky photogrammetric task.&& However, their positions
and azimuths are found approximately from photos where BPSE area
is clearly visible: 5884/85, 5905, various post-EVA window photos, and the
16-mm film shot out Buzz's window.
: Panorama 2 made from the rim of the
younger and sharper, western component of Double Crater.
: The station position may contain an error
of ~0.5 m due to a slightly-wrong focusing distance set during
modeling. Correction of this error would be tricky. Actual
location is to the south of the location on the map, so the actual
location is closer to the panorama central point.
: In a , the old eroded crater mentioned
in the comment for 5851 (above)& is in the center j