Audio Analysis Is Most Consistent Two Shooters At Trump Rally

[greg_n]

@rough_country_gypsy

Is this continuing additional analysis changing anything? Do you still have Crooks shooting all 8 rounds? Hercules 1 from the south barn with the 10th and final shot? Has anything got moved to the windows or vents?

Since the last time I published results (5 weeks ago), I have added two new data sources:

1. NTD camera audio from the press riser
2. Vertical format cell phone “Don’t Run. Don’t Run.” (IMG 994)

These sources confirm the same results provided by the previous five audio sources.

The first 8 shots come from a 2.5 foot radium at the shooter’s location. In my observation, the data is so stable that it is simply impossible that any of the first 8 shots came from anyplace other than the established shooter position on the roof.

And I’m very sure that shots 9 and 10 came from the ESU officer on the lawn and the Hercules 1 sniper on the South Barn.

[offtheback]

@vt1,

The thing that still isn’t fitting for me is that source 4 doesn’t move between shots 4 and 8. I have looked and looked at my shot alignment and can’t see how I would move them to cure this. And when I look at the video during the time of interest (21.3628 to 22.1422) I swear it looks like he is walking.

None of my other sources show these offsets for shots 4-8. So are we sure he doesn’t move?

If I place a “0” at time/location for shot 4, I see the following anomalies for 5-8 at source 4 (positive means the shot reports are arriving too late, or placing the shooter too far away):

Shot Anomaly (ft)
4 0
5 1.6
6 2.7
7 2.9
8 4.0

[rough_country_gypsy]

It’s perplexing to say the least. However, even though I have a tough time swallowing it, I’m going to have to go with the evidence of where the 10th shot came from.

[screener]

This is clear. There was actually only one shooter, presumably Crooks. Do the rest of you geolocation guys agree?

[offtheback]

@rough_country_gypsy, @greg_n has done some very useful work in helping us understand the scene, and in generally locating the source area for shots 1-8, 9, and 10. But a shot-by-shot estimate of likely shooter location has large error bars on each, and Greg is mixing snick-estimated times and estimated-estimated times and police-fabricated data with measured boom times. I don’t agree with his timing numbers overall, but am waiting for confirmation on one more thing regarding source 4 before I wrap on my conclusions and publish numbers. I am using strictly measured report times, measured as accurately as I can. We can’t assume these guys were idiots, and if there were two shooters, effort would be made such that it would seem to a first approximation that there was one location for all shots.

What I have focused on is not where did each shot come from, but on the following:
a) did shots 5-8 come from the same location as each other? I am 99% sure this is so, based on inter-shot timing that agrees for all sources
b) is that location Crooks’ location or very nearly so? almost undoubtedy, as data from two moving sources, when corrected for movement relative to Crooks, supports this. @vt1 ‘s microphone movement analysis is critical to confirm this, and source 4 is the only moving source not yet converging to this.
b) did shots 1-3 come from the same location as each other? I am 95% sure this is so; stationary sources agree on inter-shot spacing for this group.
c) did shots 1-3 come from the same exact location as shots 4-8? absolutely not. Here are a few reasons:
-stationary sources show a time offset from each other between shot group A and shot group B, dependent on where they are on a compass rose
-moving sources won’t converge to a set offset when correcting for Crooks’ location, like they do for shots 4-8. This means the corrections are being made with respect to a wrong shot position

• for all shot locations that weren’t faked, the waveforms and echoes show consistency between shots 1-3 and between shots 4-8 that isn’t there between the groups
• three weeks after the fact, someone was compelled to fake police cruiser audio that would be inconsistent with all of the above
• weeks after the fact, someone was compelled to “re-release” source 3 video to make it look like Crooks took the first shot. I have a long debunk post of this that I can’t seem to find time to finish

So where did shots 1-3 come from? That is the next question. The data suggests a location a bit closer than Crooks, maybe near the wall of the building but probably set back, and generally on the same line to the podium, but I haven’t tried to numerically solve that yet. Perhaps someone else can jump on that based on my timing numbers when I publish them.

[howdoiknowthisinfo (howdoiknowthisinfo.com)]

@offtheback

The data suggests a location a bit closer than Crooks

I totally agree with you!

people will now argue that this is another tick in the column that shows my supposed confirmation bias, but it is simply a matter of common sense and critical thinking, keeping an open mind and exploring all options before making a decision

what makes you assume that the location must be a bit closer than crooks?

why not further away, say behind crooks, i.e. from the high roof?

so my question is: what is the data and what are the inputs to your analysis that make you assume that the shooter of the first 3 rounds had to come from a closer and not further away location?

people assume that the same rifle and ammunition must be used for all 8 rounds, and this would then lead to the specifications of crooks’ weapon, but the muzzle velocity, caliber and bullet weight of the first 3 rounds could be completely different than the next 5 rounds…

[offtheback]

@howdoiknowthisinfo the main thing that keeps the two rifles fairly close to each other is that stationary sources far off the line of the shots (like source 3) don’t show very big deltas between shot group A and shot group B, when compared to the podium. If there were more than 25 feet or so of difference in shooter location, these deltas would be huge. @greg_n isn’t far off, just off.

I personally really liked the behind-Crooks locations, but I think the data is driving us to somewhere in building 6.

Also, with respect to ammunition, rifle, direction of fire, etc., report times don’t care. As long as the speed of sound doesn’t change between shots, we don’t need to assume anything about rifles and bullets and such.

[greg_n]

@offtheback

But a shot-by-shot estimate of likely shooter location has large error bars on each

Not necessarily. The moving recorders add significant error, but if you focus on the stationary recorders I think you will see there is very tight agreement for the locations of shots 1-8. You don’t want to use the Cruiser video? That’s fine. Instead, use either the NTD press riser video or the “Don’t Run. Don’t Run.” video. If you use one of these, along with Ross and Podium, I’m confident that you will change your mind about shots 1-3 and 4-8 being from different locations. And I am talking about using boom times for these shots, not corrected crack times.

Using this data, the first 8 shots fall within a 4 foot radius:

I am now using leading-edge boom times as much as possible in my analysis, as shown by the color coding in the spreadsheet I placed on Brian’s file server:

For the Podium, I had to estimate the boom times for shots 6 and 7 because they were covered by cracks. With the audio waveform analysis that you do, you may be able to get better timestamps for those two shots and tighten up the group even more!

[offtheback]

@greg_n where can I get the two new sources you mention? I have no pride in a particular result, and will include them and see.

The stationary source deltas are not big. That is why I have put so much effort into measuring arrival times, to make sure I am seeing something other than measurement error.

[offtheback]

I assume this is the NTD position:
NTD 586691.16 4523403.54
what is the Don’t Run position?

[vt1 (VT)]

@offtheback

@vt1 ‘s microphone movement analysis is critical to confirm this, and source 4 is the only moving source not yet converging to this.

Working on that!

[greg_n]

The NTD video is here
Geolocation by @vt1 is 586691.16 4523403.54

The “Don’t Run.” video is here
Geolocation by @vt1 is 586654.15 4523414.58

[undisputed-facts]

I notice there’s a debate in this forum, similar to the one on X (formerly Twitter), regarding whether there was one or two shooters.

In fact, there’s a straightforward way to determine this: by checking whether the bullet of the first shot was ascending or descending. Based on what I’ve seen on Twitter, the point of impact near the bystanders is approximately 12 feet high, while Trump’s ear is about 11 feet, using the AGR building at sea level set to zero.

This suggests a ascending bullet, which rules out the possibility of Crooks on the roof being the shooter of the first shot.

[vt1 (VT)]

My latest version for bullet 4 and 8 for source 4.

reference	longitude	latitude
S4 bullet 4	586684.73	4523536.29
S4 bullet 8	586684.29	4523537.00

The difference is 0.84m (2.74ft).

[howdoiknowthisinfo (howdoiknowthisinfo.com)]

and where would you place the shooter of the first bullets?

[offtheback]

@greg_n @cmartenson @vt1 @brian60221 @howdoiknowthisinfo

Here is my report timing and a few things I think we can conclude. A few of the tracks that I recorded off the Internet have a random offset and DJStew has minutes dropped, none of which should matter as we are syncing everything below to shot 4 timing anyway. I added NTD and S7 “Don’t Run” today on @greg_n 's suggestion, and they confirm what I was seeing.

Report Timing, seconds, first 8 shots

Source	1	2	3	4	5	6	7	8
S2 Fence	14.7458	15.6044	16.2792	19.0714	19.3274	19.511	19.6881	19.847
S4 Mover	17.025	17.882	18.56	21.3628	21.62	21.8047	21.9815	22.1422
S3 TMZ	3.6907	4.5445	5.2177	7.9982	8.2535	8.4369	8.613	8.7727
S6 DJStew	55.0078	55.8632	56.5392	59.3238	59.5792	59.7627	59.939	60.0985
S5 Root	2.599	3.453	4.1263	6.9085	7.164	7.3474	7.5234	7.6833
S1 Podium	2.8105	3.6645	4.33771	7.1204	7.3758	7.5592	7.7354	7.8952
S7 Don’t Run	27.6693	28.5232	29.1964	31.9787	32.234	32.4176	32.5938	32.7535
NTD	40.2103	41.0642	41.7375	44.5199	44.7752	44.9587	45.1348	45.2946

Here are my delta distances from each source to Crooks, as compared to shot 4, based on @vt1 geolocation data.

Delta Distance to Crooks From Shot 4 (feet)
Source	1	2	3	4	5	6	7	8
S2 Fence	-8.8	-7.0	-6.4	0.0	0.0	0.0	0.0	0.0
S4 Mover	-28.1	-24.3	-20.2	0.0	0.4	0.8	1.2	1.6
S3 TMZ 3	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
S6 DJStew	-0.8	0.1	-0.8	0.0	0.0	0.0	0.0	0.0
S5 Root	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
S7 Don’t Run	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
NTD	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

I am doing this so I can postulate whether shots came from the same location as shot 4. I now use these as adjustments to shot timing relative to shot 4, and then compute an anomaly by comparing this to relative shot timing for the podium. And convert these from seconds to feet using the speed of sound.

Delta Distance from Expected Arrival (feet)
Source	1	2	3	4	5	6	7	8
S2 Fence	-9.1	-5.7	-4.5	0.0	0.7	0.9	1.9	0.9
S4 Mover	-3.8	-4.1	-2.8	0.0	1.6	2.7	3.0	3.6
S3 TMZ	2.7	2.5	2.5	0.0	-0.1	-0.1	-0.2	-0.3
S6 DJStew	-4.6	-4.8	-1.3	0.0	0.0	0.1	0.2	-0.1
S5 Root	0.5	0.5	0.6	0.0	0.1	0.1	-0.1	0.0
S7 Don’t Run	0.6	0.5	0.4	0.0	-0.1	0.1	0.1	0.0
NTD	0.3	0.2	0.3	0.0	-0.1	0.0	-0.1	-0.1

If the shot being compared to shot 4 came from the same location, the anomaly should be zero. I consider a few tenths of a millisecond, or about 3 inches in distance, to be achievable in resolution. For stationary sources and decent waveform SNR, it seems relatively easy to get half that resolution. Here is a bar graph for the stationary source anomalies:

A few observations. First, shots 4-8 definitely cluster around zero anomaly, with the largest deviation being 4 inches. Second, shots 1-3 cluster for each source, but with a definite shift. The shift is 3-4 inches for NTD, which is closest to being on a line to the podium from Crooks, so we should expect a small shift. Source S5 Root and S7 Don’t Run were shifted a bit more around the compass rose from the podium, but close to each other, and show a roughly 6-inch shift. Source S3, which was almost orthogonal to the line of fire, shows a 2.5-foot shift in the data.

You may say “these are tiny, so what?”. But they shouldn’t be there, shots 1-3 would have anomalies like shots 5-8 otherwise.

You may say “what if Crooks shifted his gun 6 inches or whatever?”. That isn’t what these anomalies mean. For a source like NTD for which the cosine of the angle between the direction to it and the direction to the podium is like 0.975, you have to move the gun barrel towards the podium almost 15 feet to register a 4-inch shift in delta arrival distance. So yes, these mean something.

Now let’s turn to the moving microphones. I got decent agreement for all but source 4 that shots 4-8 all came from the same location as Crooks, impressive work @vt1 ! For source 4, we are a bit off but orthogonal to Crooks so it is sensitive to errors in position, but these are still decent.

The beautiful thing about moving sources is that if the correction I am applying for the vector to Crooks is wrong because Crooks’ position isn’t the position of the shots, these anomalies go whack. But get better the closer we are to the microphone position for shot 4. Which is exactly what we see for shots 1-3.

So what I haven’t done, is instead of correcting shots 1-3 for Crooks’ location, correct for a matrix of other locations (which I would suggest are south and east of Crooks) and try to minimize the anomalies from the podium timing, and you should have your first shooter.

[undisputed-facts]

To prevent any confusion, I would, for the time being, keep it very simple:

If the first shot is ascending, it indicates the presence of two shooters

[howdoiknowthisinfo (howdoiknowthisinfo.com)]

@undisputed-facts

If the first shot is descending, it indicates the presence of two shooters

yes, I know.

form my point of view, the first shots were fired from the high roof behind crooks, i.e., from the highest AGR building at their parking.

if crooks fired any shots, these would have been shots 4-8, if he did not fire any rounds, they would have originated from that highest roof too…

so, where would you put the shooter of the first rounds?

[undisputed-facts]

Apologies for the typographical error.

With the bystanders positioned at 12 feet and Trump’s ear at 11 feet, the shot must be ascending (indicating a rising bullet).

Therefore, your assumption that the shooter is even higher than Crooks is unfounded.

I have amended my post accordingly to prevent any further confusion.

[khunter]

I certainly am not sure of an actual bullet mark, but I tried to work out the position of the bullet impact indirectly, by taking the midpoint of the dust cloud on the 1st frame that it appeared. I get it as being about 4 inches from the top of the back railing.