We got a little distracted from your comment:
It has since been clarified that you were not up to date with the latest findings. Recent first-hand video evidence shows that a flash of light inside the open vent 1 is clearly visible at the moment of shot 3. This flash, along with the matching rectangle appearing in the position of the vent at the same time of shot 3, makes it virtually impossible for it to be a random artifact over multiple frames— I would give it about a 1 in 100,000 chance.
That said, as you mention, we’ve reached the point where continuing to debate this topic is a waste of time.
Let’s return to the original question: the distinct echo patterns in Shots 1–3 versus Shots 4–8. We can define an echo footprint using Source 4 on the west side of the building:
Here is the source:
Source 4 - Shots 1 - 8 original.wav
- Shots 1–3: early reflection present but less consistent and generally weaker (0.077s).
- Shots 4–8: a consistent late echo around ~1.7 s.
And in Steward’s recording from the southeast of AGR Building 6:
Here is the source:
steward shots 1 - 8 original.wav
- Shots 1–3: clear ~0.25 s echos.
- Shots 4–8: no comparable echo detected.
Acoustic Report — Shot Origins ( Roof vs Vent)
This is a simple explanation of where the shots came from (vent or roof) using the two microphones (“Source 4” and “Steward”).
Files used:
- steward shots 1 - 8 original.wav
- Source 4 - Shots 1 - 8 original.wav
1) The basic idea (no math)
Loud sounds bounce off big walls and come back as echoes.
Each mic can “see” only one big wall:
Source 4 sees Wall 1 (to the northwest of the roof position) and not Wall 2.
Steward sees Wall 2 (to the south) and not Wall 1.
If a shot is on the roof (in the open), the sound reaches the big wall 1 and we get strong, clear echoes at the mic that can see them. However, the roof ridge blocks the path to Wall 2 and is in a higher position, redirecting energy toward Wall 1.
If a shot is coming from the vent (recessed), the vent blocks and muffles; and because Wall 1 is behind the building, only Wall 2 is visible to the shot.
2) What we saw in the audio
A) Source 4 (looks at Wall 1)
Shots 1–3: we see short echoes around 0.077–0.083 s. These are from a nearby surface (like the vent mouth). The big Wall‑1 echo is weak or missing.
Shots 4–8: no big Wall‑2 echo shows up. That fits a roof shot where the roof ridge blocks the path to Wall 2; another reason is that the roof position is higher than the vent, preventing sound from reaching Wall 2. However 4 - 8 shows the clear ~0.170 s Wall‑1 echo on each shot in this recording
B) Steward (looks at Wall 2)
Shots 1–3: we see a big echo, roughly a quarter to a third of a second after the shot:
Shot 1: ~0.243 s
Shot 2: ~0.244 s
Shot 3: ~0.331 s (mic was moving during shots 1–3, this explains the shot 3 differance) This tells us the sound went strongly toward Wall 2.
Shots 4–8: no big Wall‑2 echo shows up. That fits a roof shot where the roof ridge blocks the path to Wall 2.
C) Audio differences between the two shot groups
Loudness at Source 4: Shots 1–3 are noticeably quieter at Source 4, while shots 4–8 are much louder. This is because the mic is on the west side of the building with an almost direct line of sight to the roof; roof shots project strongly toward Source 4, but vent shots are partly blocked/muffled, so they arrive weaker.
Tone (high‑frequency content): Vent shots (1–3) sound duller (less top‑end above ~2–3 kHz). Roof shots (4–8) sound brighter/cleaner.
Early reflections: Vent shots show a short, nearby reflection at ~0.077–0.083 s (not the big wall). Roof shots show a consistent big Wall‑1 echo at ~0.170 s.
3) Putting it together (easy conclusion)
- Shots 1–3 came from the vent.
Why: Source 4 gets only a small, early echo (~0.08 s) instead of the big Wall‑1 echo, while Steward gets a strong, later echo from Wall 2. That’s exactly what you expect if the shot is recessed inside a vent, which points energy away from Wall 1 and more toward Wall 2. - Shots 4–8 came from the roof.
Why: Source 4 shows the clear ~0.170 s Wall‑1 echo on each shot, while Steward shows no Wall‑2 echo (the ridge blocks that direction). That’s the open‑air roof pattern.
C) Mobility constraint (prone on the roof)
The roof shooter was lying down (prone). In that posture, a ~2‑second relocation between roof and vent is not feasible . Other positions wouldn’t change the echo footprint.
Therefore the abrupt echo change between shots 3 → 4 is best explained by two fixed positions (vent then roof) or different operators, not motion between shots.
4) Quick reference tables
Source 4 (Wall 1):
| Shot group | Echo time(s) at Source 4 | What it means |
|---|---|---|
| 1–3 | ~0.077–0.083 s (short/near) | Not the big wall 1; consistent with vent shots |
| 4–8 | ~0.170 s (repeatable) | Classic Wall‑1 echo; consistent with roof shots |
Steward (Wall 2):
| Shot | Echo time at Steward | What it means |
|---|---|---|
| 1 | ~0.243 s | Big Wall‑2 echo → supports vent |
| 2 | ~0.244 s | Big Wall‑2 echo → supports vent |
| 3 | ~0.331 s (mic moving during shots 1–3;) | Big Wall‑2 echo → supports vent |
| 4–8 | none | Roof ridge blocks the path to Wall 2 and is generally in a higher position → roof |
Here are the two recordings compared at 25% speed (sounds like a Halloween ghost recording):
compare echoes shots 1 - 8 - source 4 and steward.wav
5) Plain‑English bottom line
The first three shots have the vent signature (small echos (from the vent mouth) at Source 4, big late echos from wall 2 at Steward).
The last five shots have the roof signature (strong ~0.17 s wall echo at Source 4, nothing big at Steward).
If we put it very straightforwardly: assuming all 8 shots were fired from the roof, then in the recording from Source 4 you should consistently see a 0.17-second echo from Wall 1 on all 8 shots, and in the recording from Steward you should consistently see a 0.25-second echo from Wall 2 on all 8 shots. The fact that shots 1–3 and shots 4–8 do not follow this echo footprint confirms—across two independent videos—that firing all 8 shots from the roof is completely inconsistent.
So: Shots 1–3 = VENT. Shots 4–8 = ROOF.
