SF-0005Photographs That Need Explaining
NASA published the Face on Mars itself, and called it shadows in the same sentence
The suppression story attached itself to a photograph its alleged suppressor put out — six days after it was taken, in a release that pointed at the face before anybody else did.
The frame that started it: Viking 1 Orbiter, 25 July 1976, photographed from 1,873 km with the sun about 20 degrees above the horizon. The black speckles are bit errors — data lost in transmission from Mars — and two of them fall inside the "eye" and "nostril". This is the picture as NASA released it, at the resolution that made the face.
NASA's Viking 1 Orbiter spacecraft photographed this region in the northern latitudes of Mars on July 25, 1976 while searching for a landing site for the Viking 2 Lander. The speckled appearance of the image is due to missing data, called bit errors, caused by problems in transmission of the photographic data from Mars to Earth. Bit errors comprise part of one of the 'eyes' and 'nostrils' on the eroded rock that resembles a human face near the center of the image. Shadows in the rock formation give the illusion of a nose and mouth.
- Image
- NASA / Jet Propulsion Laboratory — Viking 1 Orbiter, Visual Imaging Subsystem Camera A
- Taken
- 1976-07-25
- Location
- Cydonia, northern latitudes of Mars
- Archive ref.
- NASA Photojournal PIA01141 (JPL release P-17384, frame 35A72)
- The image itself
- Authentic, unaltered
NASA/JPL, Viking 1 Orbiter. Photojournal PIA01141. Public domain.
Verdict
"In 1976 NASA photographed a giant face on Mars and then quietly buried it."
The claim is demonstrably untrue, and the evidence against it is clear.
Both halves of the claim fail, and the second one fails on a document that has been public since 1976.
TRUE, AND NOT IN DISPUTE: the photograph is genuine and the resemblance is real. NASA's catalogue entry for the frame describes "the eroded rock that resembles a human face near the center of the image", 1.5 kilometres across, photographed from 1,873 kilometres with the sun 20 degrees above the horizon. Anyone who found it striking in 1976 was looking at something genuinely striking.
FALSE, ON THE RELEASE: NASA did not bury it. JPL issued the image as photo P-17384 on 31 July 1976, six days after Viking 1 took it, and the caption is the reason the face is famous at all — "The huge rock formation in the center, which resembles a human head, is formed by shadows giving the illusion of eyes, nose and mouth." One sentence, containing both the observation and the explanation. The suppression story is attached to a photograph that reached the world because its alleged suppressor pointed at it.
FALSE AGAIN, ON THE FOLLOW-UP: NASA announced the re-imaging before it happened. A JPL release of 26 March 1998 said the agency would attempt to photograph "features of public interest" including Cydonia, and quoted Carl Pilcher: "Most scientists believe that everything we've seen on Mars is of natural origin. However, we also believe it is appropriate to seek to resolve speculation about features in the Cydonia region by obtaining images when it is possible to do so." Ten days later the Mars Orbiter Camera imaged it at 4.3 metres per pixel, which NASA described as ten times the resolution of the best previous Viking image.
WHAT THE BETTER PICTURES SHOW: an eroded mesa. The camera returned to it on 8 April 2001 at about 2 metres per pixel, and in 2007 the HiRISE camera imaged it at 29.9 centimetres per pixel, where the University of Arizona describes it plainly as "an eroded mesa". The face does not survive the sharper pictures or the changed lighting — which is what the 1976 caption explains rather than predicts, since a shape made of shadows depends on where the sun is.
THE PART THAT IS FAIR: twenty-two years separate the photograph from the first re-imaging, and a reader who says NASA was in no hurry is not wrong about the interval. NASA's own 1998 wording treats Cydonia as speculation to be resolved rather than as a scientific question of its own. That is worth stating, and it is a long way from suppression.
False requires 1 primary or authoritative source. This file cites 7.
In short
On 25 July 1976 the Viking 1 Orbiter photographed a mesa in Cydonia that, at that resolution and a 20-degree sun angle, looks like a human face. The photograph is authentic. What the "NASA buried it" story cannot survive is the document that carried it: JPL released the image six days later, and the caption points at the resemblance itself — "the huge rock formation in the center, which resembles a human head" — then explains it in the same sentence, "formed by shadows giving the illusion of eyes, nose and mouth".
The picture is real
There is no forgery here and no retouching, and this file is not the kind that ends with “it was faked all along”. The Viking 1 Orbiter photographed the Cydonia region on 25 July 1976 while searching for somewhere to land Viking 2, and one of the mesas in the frame looks like a face. It looks like a face now. It looked like a face to the people at JPL who wrote the caption.
What makes it look like a face is stated in NASA’s own record of the frame: the sun was about 20 degrees above the horizon. A low sun fills shallow depressions with long shadows, and at 1,873 kilometres a mesa a mile across is a smudge with dark patches in it. Two of the darkest patches are not even terrain — they are bit errors, data lost on the way from Mars, and NASA’s caption says exactly where they fell: “part of one of the ‘eyes’ and ‘nostrils’”.
Part of the face is missing data.
The part that does not survive
The story that travels with the photograph is that NASA buried it.
The photograph was released on 31 July 1976, six days after it was taken, as JPL photo P-17384. The caption is short, and this is the operative sentence:
The huge rock formation in the center, which resembles a human head, is formed by shadows giving the illusion of eyes, nose and mouth.
Read it twice. The agency accused of hiding the face is the one that pointed at it. The observation and the explanation arrive in the same sentence, in the document responsible for the picture being famous at all. Nobody discovered this image in an archive — it was handed out.
What happened when they went back
The stronger version of the claim is not that NASA hid the photograph but that it refused to look again. That version deserves testing, and it has a real fact behind it: twenty-two years passed before anyone re-imaged Cydonia.
When it happened, NASA said so in advance. A JPL release of 26 March 1998 announced that Mars Global Surveyor would attempt to photograph “features of public interest” including the Cydonia region, and quoted Carl Pilcher:
Most scientists believe that everything we’ve seen on Mars is of natural origin. However, we also believe it is appropriate to seek to resolve speculation about features in the Cydonia region by obtaining images when it is possible to do so.
Ten days later the Mars Orbiter Camera returned an image at 4.3 metres per pixel — ten times sharper than the best Viking view, by NASA’s own comparison. It went back on 8 April 2001 at about two metres per pixel. In 2007 the HiRISE camera photographed the same ground at 29.9 centimetres per pixel, where the description is four words long: an eroded mesa.
Ten times sharper than Viking, then twice as sharp again, then finer still: 4.3 metres per pixel, about two, then 29.9 centimetres. The face does not survive it, and the 1976 caption is the reason that is unsurprising — a shape made of shadows does not keep its shape when the light moves.
What we will not claim
That NASA behaved impeccably throughout is not established here, and the file does not say it. Twenty-two years is a long time. The 1998 wording treats Cydonia as speculation to be handled rather than a question to be asked, and a reader who finds that language dismissive is not misreading it.
There is also a claim, which would matter if it held, that NASA overstated its evidence in correspondence with Congress in 1993 — citing a second Viking frame it could not produce. We went looking for the primary document and did not find one; the trail ends at a single advocacy book. So it is recorded in the uncertainties, where unverified things belong, rather than used or dismissed.
None of that touches the sentence from 31 July 1976, which is the whole of this file: the face was published by NASA, and NASA called it shadows in the same breath.
The evidence
The 1976 JPL release points at the face itself. Photo P-17384, issued 31 July 1976: "The huge rock formation in the center, which resembles a human head, is formed by shadows giving the illusion of eyes, nose and mouth." The observation and the explanation are in one sentence, in the document that put the picture into the world.
Cited sources:S1
The interval between the photograph and its release was six days. Viking 1 acquired the frame on 25 July 1976; the release is dated 31 July 1976.
NASA's catalogue record gives the geometry that makes the face: a feature 1.5 kilometres across, photographed from a range of 1,873 kilometres, "with the sun angle at approximately 20 degrees". A low sun casts long shadows into shallow hollows.
Cited sources:S2
Two of the "features" are transmission errors rather than terrain. NASA: "Bit errors comprise part of one of the 'eyes' and 'nostrils' on the eroded rock that resembles a human face." Part of the face is missing data.
Cited sources:S2
NASA announced the re-imaging in advance and said why. JPL, 26 March 1998, quoting Carl Pilcher: "Most scientists believe that everything we've seen on Mars is of natural origin. However, we also believe it is appropriate to seek to resolve speculation about features in the Cydonia region by obtaining images when it is possible to do so."
Cited sources:S3
The 1998 image was ten times sharper than the best Viking view. NASA: the Mars Orbiter Camera picture of 5 April 1998 "has a resolution of 14.1 feet (4.3 meters) per pixel, making it ten times higher resolution than the best previous image of the feature, which was taken by the Viking Mission in the mid-1970's."
Cited sources:S4
The camera went back on 8 April 2001 and imaged the landform at about 2 metres per pixel, as MOC frame E03-00824, describing it as one of "thousands of buttes, mesas, ridges, and knobs in the transition zone between the cratered uplands of western Arabia Terra and the low, northern plains".
Cited sources:S5
The most recent authoritative imaging is finer again. HiRISE photographed the feature on 5 April 2007 at an original scale of 29.9 centimetres per pixel, and the University of Arizona describes the subject as "an eroded mesa".
Cited sources:S6
Evidence that cuts the other way
Findings that complicate or argue against the verdict above. They are listed because leaving them out would make the file look stronger than it is.
Twenty-two years passed between the photograph and the first attempt to re-image it. That interval is real, and a reader who concludes NASA was in no hurry has the arithmetic right even though "buried it" is wrong.
NASA's own framing in 1998 treats Cydonia as speculation to be resolved rather than as a scientific question worth asking. Pilcher's sentence begins by saying most scientists believe everything seen on Mars is natural, and offers the imaging as a response to public interest. That is not the language of investigation, and this file does not pretend it is.
Cited sources:S3
What we still don’t know
- A claim exists that NASA overstated its evidence in the early 1990s — that a document sent to Senator John Glenn and other members of Congress in May 1993 asserted the face disappeared in a second Viking frame taken under different lighting, that the document did not reference the frame, and that a revised draft in June 1993 dropped the reference. It is repeated in a peer-reviewed media-studies article, but that article cites exactly one source for it, an advocacy book arguing NASA failed to investigate. No primary NASA document and no congressional record was located. It is recorded here because it would genuinely complicate this file's conclusion if true, and it is not repeated as fact because nothing read here establishes it.
- How much attention the 1976 release received at the time. This file establishes that NASA published the image and the explanation; it does not establish how widely either travelled, and no contemporary newspaper coverage was consulted.
- The exact resolution ratio between the 1976 and 2001 images. NASA states the 1998 image at 4.3 m/pixel was ten times the best previous Viking view, and the 2001 image is about 2 m/pixel. Roughly twenty times is arithmetic on those two figures, not a statement either agency has made, so the file quotes the figures rather than the ratio.
- Which Viking frame PIA01141 is. The Photojournal caption gives no frame number. Its measurements — the date, the 1.5 km feature, the 20-degree sun angle, the 1,873 km range — match the 1976 release's description of frame 35A72 exactly, so the identification rests on agreeing parameters rather than on a stated frame number.
- Whether the suppression framing originated with DiPietro and Molenaar, with Hoagland, or with neither. The artificiality reading is traceable to them; the accusation about NASA's conduct is not traced here and is therefore attributed to nobody.
Sources
Every source below was read in full before it was cited. Archive snapshots are recorded at research time, because a file whose sources have evaporated is indistinguishable from one that never had any.
S1Tier 1Read in full
JPL caption for the "Face" release — photo P-17384 (frame 35A72)
NASA Jet Propulsion Laboratory — photo release caption, 31 July 1976 · 1976-07-31
The load-bearing document. A JPL photo release caption issued six days after the frame was taken, containing the sentence that both names the resemblance and explains it as shadows. Note the hosting: the transcription sits on the Malin Space Science Systems education site rather than on a nasa.gov address, because JPL's own 1976 release pages are not online. The text is a US government work; MSSS hosts it, and does not own it. Quoted as text only — no image is taken from this source.
S2Tier 1Read in full
Geologic "Face on Mars" Formation — PIA01141
NASA Photojournal — Jet Propulsion Laboratory · 1976-07-25
The catalogue record for the image published here, and the second NASA description of the same frame. Establishes the acquisition date, the instrument, the 1.5 km feature size, the 20-degree sun angle, the 1,873 km range, the bit-error speckle including its intrusion into the "eye" and "nostril", and the "NASA/JPL" credit line the rights position rests on.
S3Tier 1Read in full
Mars Global Surveyor to Attempt Imaging of Features of Public Interest
NASA Jet Propulsion Laboratory — news release · 1998-03-26
NASA announcing the Cydonia re-imaging before it happened, with Carl Pilcher's statement of why. Establishes both that the agency did not avoid the question and that it framed the exercise as resolving speculation rather than as science — which is why this source is cited on both sides of the file.
S4Tier 1Read in full
Mars Orbiter Camera Views the "Face on Mars" — Comparison with Viking (PIA01442)
NASA Photojournal — Jet Propulsion Laboratory · 1998-04-05
The 1998 re-imaging and the only explicit resolution comparison either agency states — 4.3 m/pixel, "ten times higher resolution than the best previous image". Cited for its text only; the image is credited to MSSS and is not used.
S5Tier 1Read in full
MOC image E03-00824 — the "Face on Mars", 8 April 2001
Malin Space Science Systems — Mars Global Surveyor MOC release · 2001-04-08
The instrument team's own release for the 2001 image: the acquisition date and time, the approximately 2 metres per pixel scale, the frame ID, and the description of the landform as one of thousands of buttes and mesas in the transition zone. Text only; the imagery on this page is the material this file is not licensed to publish.
S6Tier 1Read in full
Popular Landform in Cydonia Region (PSP_003234_2210)
HiRISE — University of Arizona / NASA Mars Reconnaissance Orbiter · 2007-04-05
The most recent authoritative imaging located: 5 April 2007, an original scale of 29.9 cm/pixel, and the description of the subject as an eroded mesa. This is the source that answers the resolution check — the question did not stay where the story left it.
S7Tier 1Read in full
NASA Jet Propulsion Laboratory · 2026-08-02
The licence this file publishes its image under, and the clause that blocked the sharper alternatives. Establishes the default grant for JPL-owned imagery and the third-party exception for images owned by organisations other than JPL or NASA.
Checked for circularity: S1 and S2; S1 and S3; S1 and S4; S1 and S5; S1 and S6; S5 and S2; S6 and S2; S7 and S1; S7 and S2 were confirmed not to derive from one another. Two sources that trace back to the same origin are one source, and counting them twice is the most common way a false claim looks well-evidenced.
How we checked this
What we went looking for, and what we found
Four checks. Two changed the file.
First, against the easy version. The lazy form of this file is "it's pareidolia, next", and that would have been a worse file than the claim it corrects. The resemblance is real, the 1976 frame really does look like a face, and NASA's own caption says so. The file was written so that the correction lands on the suppression story, which is checkable, rather than on the reader who found the picture striking.
Second, and this is the check that mattered: the claim was restated at its strongest before being tested. "NASA hid the photograph" is trivially false and refuting it would have been shadow-boxing. The serious version is "NASA declined to investigate", and that version has a real history behind it — congressional correspondence in 1993, a decades-long campaign, and a gap of twenty-two years before any re-imaging. Testing that version is what produced the 1998 announcement as evidence and the twenty-two-year interval as contradicting evidence.
Third, against this file's own conclusion. A claim circulates that NASA's 1993 document to Congress asserted the face vanished in a second Viking frame under different lighting, could not produce that frame, and quietly revised the document. If true it would complicate the conclusion considerably. It was traced: a peer-reviewed media-studies article carries it, and that article cites one source, an advocacy book. No primary NASA document and no congressional record was found. It is recorded as an uncertainty and not as a finding, and the file's conclusion is scoped to what the 1976 and 1998 documents establish.
Fourth, on the images this file cannot use. The sharpest and most persuasive pictures of Cydonia are the ones it is not licensed to publish, and the temptation was to treat a nasa.gov address as permission. The credit lines were read instead, the third-party owner's own site was checked, and the file publishes the 640-pixel 1976 frame it is entitled to publish. That constraint turned out to serve the story: the low-resolution frame is the subject.
- We searched for evidence against our own conclusion
- We could not trace where the claim started
- The claim is stated in its strongest form, not a weak version
- The photograph was verified independently of the story
- We checked whether this was settled after the popular version froze
- Image rights were traced to a stated licence
- We checked what we were leaving out
- The verdict sat before publication, and was re-read cold
One check was not completed on this file. It is listed above rather than omitted, because a checklist that shows only its ticks is decoration.
- Confidence in the verdict
- high
- Verdict assigned
- Most recent authoritative source
- Researched by
- The Strange File
The verdict is False rather than Misleading because both halves of the claim fail rather than one. The landform is not artificial — the resolution chain from 43 metres to 30 centimetres per pixel settles that — and the suppression element is contradicted by the release that carried the photograph. What keeps the file honest about its own confidence is the twenty-two-year interval, which is recorded as evidence for the claim rather than explained away.
Part of a collection
The Mechanism Was MundaneLandscapes that look impossible, explained by something ordinary and knowable.
Also related
A 30-metre rock made a 1.2-kilometre hole, and a mineral made impact craters recognisable
The USGS puts the impactor at about 30 metres across and 100,000 tons, arriving at 12–20 km/sec with the force of over two million tons of TNT. The hole it left is 1.2 km wide and 180 m deep.
Grand Prismatic's rings are microbial. Its blue centre is physics, not sterility
Cyanobacteria top out near 73°C. The middle of the spring sits near boiling. No photosynthetic mat can live there — but the Park Service records abundant Archaea in it, and the blue is what deep water does to light.