Audio Analysis Is Most Consistent Two Shooters At Trump Rally

brian60221 · August 10, 2024, 5:05am

There are bullets that exist for Crooks’ gun that travel slower. It’s trivial to make one. I used to reload rounds with my dad all the time. If you don’t know that, you really have some kind of problem, and I have no idea why you’re being such a dick about it.

I agree with you. You should try hard to extract meaning from the words I write to you.

No it doesn’t. You just made that up and stuck it in my mouth for unknown reasons. I hope you’ll stop doing that, and I hope you’ll review the concept of skeptical reproduction of work in science.

sgt-raven · August 10, 2024, 6:04am

We may use a 55gr load as a estimate, but we don’t know what ammo was used. It could be .223 Remington or 5.56mm NATO. The bullets could be 40gr to 77gr. We don’t know for sure the barrel length.

At best we can make a SWAG (Scientific Wild Ass Guess).
Every one of these will affect the ballistic data we are working with.

sgt-raven · August 10, 2024, 6:17am

I don’t concider 75-77gr ammo weirdly slow and it runs in the 2700-2800 fps range. SOCOM has made good use of Black Hills Mk262 77gr OTM.

sgt-raven · August 10, 2024, 6:27am

We don’t know and neither do you, what ammo Crooks used that day. What any one place carries is all over the map.

brian60221 · August 10, 2024, 6:52am

So, I’ve been studying @cohler’s post here: Audio Analysis Is Most Consistent Two Shooters At Trump Rally - #220 by cohler

As I said already, I think it’s a good post. The algebra is good, and the computation seems correct to me. But I wanted to play around with it in my own way, so I wrote a dumb program to allow the variables to vary over (hopefully) absurd ranges.

For example, we don’t know the exact speed of sound that day. A good guess might be 1152 ft/s. I let it vary between 1135 and 1180. We also don’t know the speed of the round, but a good guess is probably 3000 ft/s. I let it vary between 2800 and 3500 ft/s. The crack-thump difference is thought to be 0.22 s, but I let that vary by one thousandth of a second.

If we run that kind of simulation, the best we can do is a tad over 450 ft, but that’s with a slow round (2800 ft/s) and a fast speed of sound (1180 ft/s) and if we undercounted the crack-thump diff by 0.001 s. So, we can see that to find a solution, it’s having to strain against the bounds.

But I think the best estimate for the distance of the shot is 454 ft (please correct me if I’m wrong). And there’s no way that the speed of sound on that day was 1180 ft/s – it was probably closer to 1160 ft/s. And most 5.56 rounds seem to have a speed of more like 3000 ft/s, maybe up to 3250 ft/s or so, not 2800.

End result: it seems like if the crack-thump diff is really 0.22 s, and the shot is really 454 ft, then it’s very unlikely that Crooks took the first shot unless he was using slow ammo.

Here are some examples I found of ammo that claims to be slow. I don’t know much about ammo, so I would appreciate any opinions:

Just more food for thought: if we allow the round to travel at 2650 ft/s, and for the crack-thump diff to be 0.2201s, then speed of sound could be 1160 ft/s, and the distance would be 454 ft.

kincses-zsolt · August 10, 2024, 7:09am

Note: he did not use the ladder he bought.
Why should we think he used those rounds?

sgt-raven · August 10, 2024, 7:14am

Assumption is the Mother of all Fuck Ups.

jmholford · August 10, 2024, 11:06am

I do not agree that “the math is simple.” The speed of a bullet is not constant. It slows considerably in the first few hundred meters.

Bullet_Slows

Analysis of Acoustic Signatures of small firearms
for Gun Shot Localization
Nimmy Pathrose, Raveendran Nair K, Murali R, Rajesh K R, Nimmy Mathew, Vishnu S

There are also non-math considerations:

barrel heating
bullet manufacturing tolerances
miss distance (between bullet trajectory and microphone)

wardenclyfferesearch · August 10, 2024, 2:33pm

The clean sources from Chris’ dropbox provide 3 sources to the West of the building that TMC was on, one to the South and the new (probably doctored; very theatrical IMO) police bodycam to the East.
I think this provides enough data to align these files with an absolute time. (I will try to do this)
If I can do this then we have a means for triangulation of the sounds with a 31 cm precision.

I forgot to click “send” yesterday… I think the audio “from Don’s mic” has to interpreted slightly different. The mic is not the one in front of Donald but a bit further back. This means that you hear the sonic snaps through the loudspeakers (=delayed!) and the gun report direct.
Today I didn’t have much time, more later.

elec_eng45 · August 10, 2024, 2:42pm

All,
I agree there are two shooters. There are several ways to prove two shooters but supersonic cracks to gun barrel reports is not the way to geolocate the different shooters. Shots 1-3 were accurate, 4-8, not so much. They were essentially scattered. I believe the supersonic cracks of shots 4-8 were from a bullet that may have long since passed. A Mach 2.5 object shock wave travels 40% the distance perpendicular to its forward path. So if a shot was 15 feet off, the bullet was 37 feet downstream of the microphone before the shock wave is detected. The shock wave snick time is not the time the bullet passed the microphone since the bullet is moving faster than the speed of sound. That is why shock detecting systems for geolocation of a shooter use multiple microphones (See Boomerang by BBN Raytheon). This only works for every bullet that passes very close to the microphone. But to range, we need to know the bullet and cartridge. I am not convinced shots 1-3 were .223 rounds at this point. So snick to boom time differences are not a reliable range solution.

commenter · August 10, 2024, 3:40pm

Greg, thank you for your great, careful, and transparent work.

To support you in checking your assumptions, I comment here on the assumption that the average velocity of bullet 1 was 3100, which is a reasonable initial velocity for a 223 round. Being so careful, I am sure that you have a good basis for this assumption.

I just want to point out that you can base your assumption as to the bullet’s average velocity on an implication from the snick-boom delay data that you so carefully got, for the recording at Trump’s mic for shoot 1.

Equations stating average velocity*time = distance:

Bullet: V*(T+s)=D
Sound: 1152*(T+s+.221)=D (per your Excel sheet)

Difference: (V-1152)(T+s) - 1152(.221) = 0 implies T+s=554.6/(V-1152)
Substituting back in top equation, D = V*254.6/(V-1152).
We can solve for V in terms of D:

V = 1152*D/(D-254.6)

Now we can impose an assumption as to distance between shot’s origin and Trump’s mic.
Say D = 420* ft.
This implies average velocity V = 2925.27.

*I have not seen an argument for this estimated distance.
With Google Earth I estimate closer to 467 ft, in which case average V = 3,252. Your 3100 is in between. In any case, I thought you might use this tight relation between V and D to tighten up your results.

offtheback · August 10, 2024, 4:31pm

So a big, big thanks to Greg Nichols for working so hard to align these sources and estimate report times. As a first matter, when I listen to/watch his slowed down podium video, I hear and see three distinct echoes at defined spacings after each of the first three shots. I know the second group of five shots step on each other, but these echoes were prominent in the first three shots and I would expect to see them in the later group, at least for shot 8. I don’t. This tells me there may be a fundamental difference in the sound source being more focused from the first three, causing prominent echoes. This is consistent with the sound having to pass through a relatively small aperture at some distance from the gun barrel. Hold onto that thought.

Second, these dudes aren’t complete idiots. They would try as best they could to place two shooters at the same distance, with the same bullet velocity, to the podium.

Now, taking Greg’s data, I worked from the assumption that shots 1-3, which sound qualitatively different, were from a different shooter than shots 4-8. So I averaged Greg’s arrival times for the first three shots, and I averaged Greg’s arrival times for the second five shots, for each audio source. I then differenced these from the same averaged podium times, as it is fair to assume from the bullet trajectories that both shooter’s reports took a direct line to the podium. Results? This is where it gets weird:

Podium: Shooters 1 and 2 at same distance and shooting towards podium (assumption)
Ross: Shooter 2 is 5-6 ft further
TMX: Shooter 2 is 15-18 ft closer
DJStew: Shooter 2 is 20-21 ft closer

This data can only be explained in two ways. 1) measurement error, i.e., sound recorders didn’t move like we thought, they ran towards Crooks after the first three shots or 2) sound path to TMX and DJStew for shots 1-3 was not a straight line.

So 2) works if Shooter 1 is roughly right UNDER patsy in eaves of building or in a room set back from a window, in either case set way back from edge of building. With a magnifying optic, he still doesn’t need a large hole to sight/shoot through. Note that he has no good way to estimate windage in this position, which could account for the miss. But to reach TMX and DJStew, the report would have to travel out the hole on a straight line to the podium, and then spread out towards these microphones, causing a dogleg sound path to the mics sideways in front of the bullet path:

This would account for shooter 1 appearing 15-20 feet further away from TMX and DJStew than the podium in the sound recordings, compared to shooter 2. To account for Ross, we would assume that there was some path (window?) for the muffled sound to travel out the near side of the building first, because the sound from the aperture would be even more delayed.

Oh, and if as some one observed one of the snipers was standing at the door to the building, he was freaking guarding it while this happened and the shooter 1 position was disassembled.

I don’t know how else to interpret this data. Thoughts?

kwaka · August 10, 2024, 5:04pm

Going to need to incorporate as many different microphone locations as possible +20 if we can. GPS does work with just a few satellites, works much better with many more. Get a stronger statistical average through all the data and limit the influence of any outliers for whatever reasons.

It is great to see all the audio analysis going on, I don’t fully understand all the math at this time, appreciate the principles and like the direction it is going.

bigtim · August 10, 2024, 6:19pm

I agree to a 2nd shooter theory because:

An inside job it obvious and the powers that be would not take this opportunity to rely on Crooks alone. Too much planning and risk invested, too much at stake (trillions of dollars, power, etc.).
Audio different gun shot sounds cannot be explained by 1 gun, but clearly 2 guns, two locations, and two shooters.
I have theories on 3 possible locations:

Arguments for the Vent location, with a concealed false ceiling.
• Excellent hiding spot to remain undetected.
• From many peoples calculations of around 425 feet based on sound, if that is accurate, the distance is perfect.
• Police were weirdly concerned with people standing at the fence on the grounds between the AGR #6 and the rally, telling people to clear away. Was this to get rid of witnesses to the forthcoming shooting? These people could presumably see or record gunshots coming from a vent.
• We have a suppressed and metallic sound/echo, consistent with firing a gun near metal, thru a tube, etc.
• Nearly identical trajectory as Crooks.
Arguments against the Vent location, or unknowns:
• This could be ruled in or ruled out depending on an interior building inspection of whether that is plausible.
• He would have to have some signal, perhaps a radio signal in an earpiece or a vibrating object on his person, or some other signal to when Crooks is in position so he can fire.
• We see no gunshots or evidence of gunshots thru those vents; no flashes of light, no obvious holes or signs of same.
• Literally no escape route. This assassin would have to be entirely confident that he would not be found and he’d have to wait there for some time and get the “all clear.”
• There were a lot of witnesses and several recordings and we do not see any evidence one might expect, no flashes from those vents, no damage to the vents which do not appear large enough for a person to look thru a scope and fire a rifle. One would need about a dinner plate sized hole or at least a saucer sized hole to look thru a scope and fire the rifle. In other words, more than just a gun barrel. I guess you could possibly use 2 perfectly aligned holes 5 or so inches apart but that gets somewhat complicated to adjust your view and firing solution.
• Do we know if the 7’ chain link fence or the bleachers or any other barriers would obstruct this firing position? Keep in mind Crooks apparently struck the victim in the bleachers and he was several feet higher (although Trump ducked down by then).

Another possibility is the 2 story AGR roof 300 feet behind Crooks. This is the location I would chose for myself on such a mission, in theory.
Arguments for 2 story AGR roof:
• Excellent vantage point and hiding location. It has vertical roof facades on two sides and a man wearing a white Tyvex suit could hide up there unseen for a long time. A white rifle and white bag could remain well hidden too.
• From this location, a pro assassin could remain hidden and observe Crooks getting into position, the crowds, law enforcement, the rally, and Trump taking the stage.
• Same linear trajectory.
• Effectively invisible from the ground.
• Excellent escape route after the shooting. Fire 3 shots, slide down, let Crooks fire off his burst of fire. Hide white tyvex suit and gun and 3 shell casings on the roof for later retrieval or take it all with him, sneak away off the back side of the building and blend in with the crowd (which was weirdly told to disperse after this crime), sneak away, or “law enforcement.”

Arguments against the 2 story AGR roof:
• Vertical trajectory and distance/sound may not line up with this position.

A 3rd possibility is a tree top but this seems remote, no ability to escape, shell casings would fall to the ground, etc.

offtheback · August 10, 2024, 6:29pm

Tim,

I like the position behind Crooks, but if Greg’s audio measurements are anywhere near accurate, it has to be ruled out.

So we are left with Crooks acting alone, which just doesn’t vibe with how almost every LEO failed to stop it or the sound differences, or

The vent theory. By my analysis of Greg’s data, the shooter would be well back from the vent, which could explain no one noting a flash? For this to be the answer, I think we also have to assume it was a couple of LEOs, so no need to hide as long as no one else is in the building. You just clean up and waltz out.

bigtim · August 10, 2024, 6:41pm

Just pulled this screen shot from a video.

Yes it’s terrible. And I’m not sure if that dark spot on the 1 story AGR #6 is Crooks but it’s awfully close. Notice the 2 story building behind him, see a white figure poking up over the roofline? If I were a pro, I’d be wearing a white Tyvex type suit, and positioned on that 2 story roof behind Crooks… Just saying. Perfect vantage, concealed from at least 3 counter sniper teams, excellent vantage of everything, Crooks getting into position, law enforcement, the rally, Trump on stage, with a excellent escape pathway.

It appears to line up with Crooks, over the window with the drainpipe, if you follow the roof ribs from the drainpipe to the dark shadow, and then follow that up higher behind Crooks there’s a unexplainable white pixelated figure - appears to be a silhouette the same size as Crooks…

It’s a wild theory, and does not line up with the audio, but perhaps the audio studies are playing tricks due to a silence rifle, different echo sources, snaps farther/closer to audio sources, etc.

Video source: https://youtu.be/UxoXzqg8aLwf

intolerance · August 10, 2024, 6:44pm

Idk if there is a second shooter, but I cannot be convinced Crooks acted alone.

He’s a 20 year old guy with no training or experience in any of this (from what we can tell by looking at his physique, I’m going to stick to that being true). Even if he got lucky, it would have to be the best luck an assassin could have.

If he was alone, we have to somehow believe that this random dude got soooooooooooooooooo goddamn lucky that he not only somehow managed to carry all of his gear (rifle, extra mags, “explosives”??) and he never got caught (questioned or detained). He gets caught with that shit in a backpack at a political rally and he’s going to prison, so he has some supreme confidence just doing this.

Next, according to bodycamera officer #122125, he informed Secret Service of the AGR building security flaw on Tuesday, which would have been July 9th. Crooks was somehow able to figure out, with like the 2 hours he spent flying his drone, that this giant flaw in the Secret Service security detail was not fixed. He somehow got sooooooooooo lucky that he knew the entire time Trump was going to be on that stage, no sniper or officer was going to be patrolling or engaging that rooftop position.

Instead I think he had help. Realistically I think at least two local cops and at least one government agent, not specifically Secret Service, maybe FBI or CIA, but someone who had insight from Tuesday that the security flaw on the rooftop was going to be exposed and never addressed. I don’t think any local cop would have been able to get that kind of intel.

edr77203 · August 10, 2024, 9:28pm

For any bullet fired, you can determine where the shot was fired if you have the following information:
The x, y, z coordinates of the cell phones, xn, yn, zn
The time of the rifle’s initial crack arriving at the cell phones, tn. The sound waveform of each shot will be marked by a transient followed by a lower waveform caused by the echoes. The rifle transient occurs at the time needed for the analysis.
From this, you can determine the position of the rifle and the exact time it was fired. Here are four equations with the four unknowns, x0, y0, z0, t0:
((x_1-x_0 )^2+(y_1-y_0 )^2+(z_1-z_0 )^2 )^(1/2)=〖Vel〗_sound x (t_1-t_0 )
((x_2-x_0 )^2+(y_2-y_0 )^2+(z_2-z_0 )^2 )^(1/2)=〖Vel〗_sound x (t_2-t_0 )
((x_3-x_0 )^2+(y_3-y_0 )^2+(z_3-z_0 )^2 )^(1/2)=〖Vel〗_sound x (t_3-t_0 )
((x_4-x_0 )^2+(y_4-y_0 )^2+(z_4-z_0 )^2 )^(1/2)=〖Vel〗_sound x (t_4-t_0 )
The rest of the information is taken from the cell phone data with the speed of sound at the given altitude.
t1, t2, t3, t4
The position information can be obtained from the person’s position in the stands.
This would be compared to the shell casings, the shot count, the shot that ended the kid’s life, and the gunshot residue in the barrel of the rifle.

brian60221 · August 10, 2024, 11:54pm

When “howdoiknowthisinfo” first started trolling the forum with this idea, I ended up testing it myself, since he wouldn’t do it.

TLDR: if somebody shot from the far roof with normal ammo, the crack-thump difference would be about 0.4s, whereas we see 0.22s in the actual data.

bigtim · August 11, 2024, 12:51am

Thank you for the analysis. Is there something, another variable, that might account for the .18 seconds crack-thump? Perhaps faster 5.56mm ammo, or larger caliber traveling 30% faster, or position/type of recording equipment, gust of wind, or a combination of variables?

If we are assuming Crooks was firing a 5.56mm traveling at ~3000 fps, then a assassin would need to be firing approximately 1.5x faster ammo to make up for the difference, or 4500 fps. I acknowledge that is unlikely to nearly impossible. Looking at even hot fast 5.56 loads into the 3500 fps range seems too slow. And larger bullets like 6.5 creedmore and 7mm magnum those are still well under 3500 fps. So this might very well preclude that location.

But frankly I’m not smart enough on acoustics and analyzing sound waves and speeds so I must defer to those more knowledgeable on this topic.

BTW, I’m not trolling at all if you read all my posts I’m legitimately interested in this topic. I personally think it is the best TACTICAL location for a assassin providing excellent overwatch of Crooks, law enforcement, crowd reactions, the rally, Trump on stage. Relatively easy shot with a scoped rifle. Excellent areas to hide, and escape toward the rear whilst everyone is focused on Crooks. Same bullet trajectory. Concealment from all 4 counter sniper teams too… Nobody was watching that rooftop.