FAILURE MAP
← Case archive

FA-79081 / Image orientation metadata / Open access

Flat-device detection compares squared magnitude to an unsquared threshold · case 01

Photos taken with the phone lying almost flat get a random orientation from sensor noise.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

The squared in-plane magnitude is compared with 300 rather than 300 squared, so the hold never triggers.

THE FAILURE

The squared in-plane magnitude is compared with 300 rather than 300 squared, so the hold never triggers.

Unsuccessful approach: An L1 norm threshold keeps diagonal readings that exceed 300 mg in Euclidean length only.

Case contract

Input [gx, gy, previous_rotation, camera] with gravity in milli-g in the portrait device frame (x right, y down). Device rotation: keep the previous bucket when the in-plane magnitude is below 300 mg or |gx| == |gy| (hysteresis); otherwise gy dominates -> 0 (gy > 0) or 180, gx dominates -> 90 (gx > 0) or 270. The back sensor is mounted at 90 degrees and needs (90 + r) mod 360 clockwise; the mirrored front sensor is mounted at 270 and needs (270 - r) mod 360 with the mirror bit. The tag comes from (quarter turns, mirror) via 1:(0,0) 2:(0,1) 3:(2,0) 4:(2,1) 5:(3,1) 6:(1,0) 7:(1,1) 8:(3,0). Return [tag, r].

Why this case matters

Camera, phone and scanner images carry an orientation hint separately from the stored pixels; galleries, thumbnailers, editors and upload pipelines must interpret it consistently or photos appear sideways, mirrored or doubly rotated.

1 / The failure

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(x):
    gx, gy, prev_r, camera = x
    if gx * gx + gy * gy < 300 or abs(gx) == abs(gy):
        r = prev_r
    elif abs(gy) > abs(gx):
        r = 0 if gy > 0 else 180
    else:
        r = 90 if gx > 0 else 270
    from_km = {(0, 0): 1, (0, 1): 2, (2, 0): 3, (2, 1): 4, (3, 1): 5, (1, 0): 6, (1, 1): 7, (3, 0): 8}
    if camera == 'front':
        total = (270 - r) % 360
        return [from_km[(total // 90, 1)], r]
    total = (90 + r) % 360
    return [from_km[(total // 90, 0)], r]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[[[247, -67, 270, 'back'], [1, 270]], [[-206, 207, 90, 'front'], [4, 90]], [[-382, -382, 180, 'back'], [8, 180]], [[50, -890, 180, 'back'], [8, 180]], [[298, 553, 180, 'front'], [5, 0]], [[53, -848, 270, 'back'], [8, 180]], [[743, -746, 270, 'back'], [8, 180]], [[201, -186, 180, 'back'], [8, 180]]], [[[-64, 236, 90, 'back'], [3, 90]], [[-230, 73, 180, 'back'], [8, 180]], [[-177, 195, 270, 'back'], [1, 270]], [[-926, 466, 0, 'back'], [1, 270]], [[110, -610, 0, 'back'], [8, 180]], [[-900, 900, 270, 'back'], [1, 270]], [[-869, 635, 270, 'back'], [1, 270]], [[-232, -166, 90, 'front'], [4, 90]]], [[[232, -182, 180, 'back'], [8, 180]], [[-237, -89, 90, 'back'], [3, 90]], [[153, 257, 270, 'front'], [2, 270]], [[741, -322, 270, 'back'], [3, 90]], [[486, 992, 270, 'back'], [6, 0]], [[54, 63, 0, 'back'], [6, 0]], [[-134, -182, 180, 'back'], [8, 180]], [[-73, 247, 90, 'back'], [3, 90]]], [[[137, 86, 180, 'back'], [8, 180]], [[-213, -186, 90, 'back'], [3, 90]], [[-186, 190, 270, 'back'], [1, 270]], [[230, 230, 270, 'back'], [1, 270]], [[66, -283, 180, 'front'], [7, 180]], [[116, 53, 90, 'back'], [3, 90]], [[-651, 121, 180, 'back'], [1, 270]], [[175, -26, 270, 'back'], [1, 270]]], [[[6, -76, 270, 'front'], [2, 270]], [[54, 289, 270, 'back'], [1, 270]], [[-107, 977, 90, 'back'], [6, 0]], [[816, 260, 0, 'back'], [3, 90]], [[994, 659, 180, 'front'], [4, 90]], [[-532, 686, 180, 'back'], [6, 0]], [[-22, 28, 0, 'front'], [5, 0]], [[-52, -248, 270, 'front'], [2, 270]]]]
labels = ["regression: flat-device magnitude threshold", "repair trap", "combined fault", "control", "control", "boundary", "boundary", "control"]
for i, (args, expected) in enumerate(fixtures[N-1]):
    check("%s %d" % (labels[i % len(labels)], i), solve(args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression: flat-device magnitude threshold 0[3, 90][1, 270]Failed
repair trap 1[5, 0][4, 90]Failed
combined fault 2[8, 180][8, 180]Passed
control 3[8, 180][8, 180]Passed
control 4[5, 0][5, 0]Passed
boundary 5[8, 180][8, 180]Passed
boundary 6[8, 180][8, 180]Passed
control 7[3, 90][8, 180]Failed

SHA-256 / 9753b5b7fcbac82f95806e409dd6a86ab5b7d590cdf2bda613ca5aac66df72bf

2 / The unsuccessful fix

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(x):
    gx, gy, prev_r, camera = x
    if abs(gx) + abs(gy) < 300 or abs(gx) == abs(gy):
        r = prev_r
    elif abs(gy) > abs(gx):
        r = 0 if gy > 0 else 180
    else:
        r = 90 if gx > 0 else 270
    from_km = {(0, 0): 1, (0, 1): 2, (2, 0): 3, (2, 1): 4, (3, 1): 5, (1, 0): 6, (1, 1): 7, (3, 0): 8}
    if camera == 'front':
        total = (270 - r) % 360
        return [from_km[(total // 90, 1)], r]
    total = (90 + r) % 360
    return [from_km[(total // 90, 0)], r]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[[[247, -67, 270, 'back'], [1, 270]], [[-206, 207, 90, 'front'], [4, 90]], [[-382, -382, 180, 'back'], [8, 180]], [[50, -890, 180, 'back'], [8, 180]], [[298, 553, 180, 'front'], [5, 0]], [[53, -848, 270, 'back'], [8, 180]], [[743, -746, 270, 'back'], [8, 180]], [[201, -186, 180, 'back'], [8, 180]]], [[[-64, 236, 90, 'back'], [3, 90]], [[-230, 73, 180, 'back'], [8, 180]], [[-177, 195, 270, 'back'], [1, 270]], [[-926, 466, 0, 'back'], [1, 270]], [[110, -610, 0, 'back'], [8, 180]], [[-900, 900, 270, 'back'], [1, 270]], [[-869, 635, 270, 'back'], [1, 270]], [[-232, -166, 90, 'front'], [4, 90]]], [[[232, -182, 180, 'back'], [8, 180]], [[-237, -89, 90, 'back'], [3, 90]], [[153, 257, 270, 'front'], [2, 270]], [[741, -322, 270, 'back'], [3, 90]], [[486, 992, 270, 'back'], [6, 0]], [[54, 63, 0, 'back'], [6, 0]], [[-134, -182, 180, 'back'], [8, 180]], [[-73, 247, 90, 'back'], [3, 90]]], [[[137, 86, 180, 'back'], [8, 180]], [[-213, -186, 90, 'back'], [3, 90]], [[-186, 190, 270, 'back'], [1, 270]], [[230, 230, 270, 'back'], [1, 270]], [[66, -283, 180, 'front'], [7, 180]], [[116, 53, 90, 'back'], [3, 90]], [[-651, 121, 180, 'back'], [1, 270]], [[175, -26, 270, 'back'], [1, 270]]], [[[6, -76, 270, 'front'], [2, 270]], [[54, 289, 270, 'back'], [1, 270]], [[-107, 977, 90, 'back'], [6, 0]], [[816, 260, 0, 'back'], [3, 90]], [[994, 659, 180, 'front'], [4, 90]], [[-532, 686, 180, 'back'], [6, 0]], [[-22, 28, 0, 'front'], [5, 0]], [[-52, -248, 270, 'front'], [2, 270]]]]
labels = ["regression: flat-device magnitude threshold", "repair trap", "combined fault", "control", "control", "boundary", "boundary", "control"]
for i, (args, expected) in enumerate(fixtures[N-1]):
    check("%s %d" % (labels[i % len(labels)], i), solve(args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression: flat-device magnitude threshold 0[3, 90][1, 270]Failed
repair trap 1[5, 0][4, 90]Failed
combined fault 2[8, 180][8, 180]Passed
control 3[8, 180][8, 180]Passed
control 4[5, 0][5, 0]Passed
boundary 5[8, 180][8, 180]Passed
boundary 6[8, 180][8, 180]Passed
control 7[3, 90][8, 180]Failed

SHA-256 / cff9f395f3bb8e921d253b890c946b5b8c1e088aa749845206db21bc1d5835a7

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 8 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Sign in to the archive ↗

Verification & scope

A deterministic bounded teaching model with a stipulated contract; it makes no claim of conformance to any published specification. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:49:41.160684+00:00.

Case digest / 6ab43105fbfcdd39201e0addbf0899ca115b92b76d051145a01f303621467b32