Great in-depth analyses. Importantly the results show issues/anomalies with the timing of the first 3 shots. Shots 4-8 timings across all mikes are remarkably consistent with being shot by Crooks at 1 location.
I assume you are using the gunshot/report times of arrival at each microphone source, as the speed of sound is constant and the gunshot/report radiated like ripples in a pond to each microphone.
Mics located close to the podium likely picked up the gunshot & crack sounds broadcast from the PA system.
You only mention the Ross source in 1 table but not in any other. Why?
âPuts tin foil hat onâ
Excepting the Podium reference source, all of the videos are not direct from the cameraman & USSS or FBI or ??? had access to them.
The first 3 shots sound track could have been manipulated by 1st shooters accomplicesâ from [heâs got a gunâŠto⊠just before shot 4]. This a pretty simple task to achieve by a expert. The timing adjustment does not need to be perfect, as the 1st shooters ammo likely had different bullet weights and powder charges to hide and confuse even more their origin.
There are also known audio timing issues within the videos, furthermore none have a solid visual timing que i.e. a muzzle flashâŠsnickâŠreport.
There were thousands of cameras rolling, ⊠how come we only have a few suspect videos of the action ???; and all seemingly telling/corroborating the official story???
The 1st 3 shots very well could come from the 3rd open window closest to Crooks in the two story building where the snipers nest was located.
There is a suspicious malodorous odour emanating from that very building and an obfuscation of the truth. Just because the provided audio does not point to this location does not mean it should be dismissed.
I have noted the members working on this have massaged some of the input data to help their conclusion coincide better with Crooks position.
Using the podium mike as a reference, very little needs to be done/adjusted to the other closest to Crooks audio tracks used, the 1st 3 shots report an snick times to make them originate from the 2nd floor window 3.
Sorry about the table confusion, I had updated my shorthand source names in my spreadsheet but missed a page. Fixed now.
Everything is report times. We wouldnât have a lot of sources if we suspect everything. I already suspect the police cruiser audio too much to use it. I am hoping with the others they thought they would pass and didnât tweak them.
I must apologise for contradicting you, but you are defining a point from a dust cloud, whereas the actual bullet mark is clearly visible in multiple videos, which were streamed live on the day of the shooting and posted shortly thereafter, on July 19th and 20th.
Anyway, Fico shooting has opened a new chapter in the history. (Crooks allegedly looked for it.)
People still discuss the JFK case. Several people would be more satisfied if there were a revealed second shooter, I guess. Why didnât they come out with a 2nd patsy and probably everyone would shut up.
In your video, you address the question of whether the shots could have been fired through the vent.
You then say that this is not possible because the results for shots 9 and 10 do not match this assumption.
Could you please explain in generally understandable terms why the localization of the first 8 shots should be dependent on shots 9 and 10?
Is it the speed of sound, which is always the same?
Can you rule out with your audio analysis that all 8 shots could have been fired through the vent without knowing which weapon and ammunition was used?
That gives a period of 0.00172s and frequency of 581Hz.
The distance travelled in that period is 0.603m. Now that is distance I know very well, itâs close the metal roof crest of 600mm. But that could be coincidence.
Now I counted the number of crests in the path from Crooks to source 4 at bullet 4 to 8. I counted 56 crests, close to the 57 waves. But that could be coincidence.
Now I looked at the line of sight from Crooks to source 4 cellphone elevation. At shot 10 he sees Crooks, so he is slightly elevated from the roof. At shot 3 he is below the roof line. Maybe he is very close to roof line at shoot 4 to 8. But that could be coincidence.
Now letâs hypothesize a possible cause of the reverberations:
The shock shockwave from the report would cross the roof crests and at each crest create a void and a reflection. Those reflections would not reach Source 4 when he is below the roof line because the sound reach that point by sound diffraction that could attenuate those roof reflections. And would not be heard in Ross audio because there is too much noise or the reflections are not at right angle.
This makes a lot of sense to me. The standing seam pattern could definitely reinforce a frequency that matches its spacing, when you are near perpendicular to it. As you move away from perpendicular, the fact that the wave is propagating radially and a reflection off one seam is going to be at an angle and bounce down the seam instead of reinforcing the next wave would seem to quash it quickly.
Whenever I can clearly discern the report time, I use the report time. When the report is being âstepped onâ by the crack, making it difficult to discern, I calculate the report time by using the formula that you provided.
Okay. Per your request, I added a new sheet called âcracksâ the the spreadsheet on the file server.
HI @vt1 Here is a report timetable that includes shots 9 and 10. I didnât include those originally as I feel they are fairly up for debate.
A few caveats.
shot 10 seems to be suppressed. That doesnât screw with snicks for mics in front of the rifle, but it really screws with the booms. It essentially spreads out the sharp pressure peaks at the beginning of the boom, so I am measuring mush. My confidence that I am nailing timing is low for reports and shot 10.
shot 9 can also be difficult due to low signal level and high noise in some cases. I think I did better than shot 10. The nice thing about the earlier shots 1-3 and 4-8 is, because there are similar repeated shots, it is easier to bootstrap the group into alignment using waveform similarities, and we donât have that here.
not all the timings here are absolute with respect to video time. Particularly so where I had to record off a playing YouTube video. For DJStew, they drop the minutes, but the times are also based on filtered data, which has a lag of roughly 0.7 ms from the real data, but I just canât see the reports accurately in the unfiltered data so we get what we get.
With this model, the localizations of the 10 shots are all inter-dependent. I have to choose one location for one shot as âground truthâ, and then the other 9 localizations fall out from there. So, the computed locations for shots 9 & 10 depend on what I choose as ground truth for shot 1. When I use the rooftop location, Shots 9 & 10 resolve to the ESU (SWAT) officer and the South Barn sniper. When I use the vent location, the location for shot 9 resolves to an implausible location near the light pole (shown below). We know this is not possible because of the police bodycam footage.
Yes, because I am using boom times for TMX, Podium, Cruiser, NTD and Donât Run for shots 1 - 9, so bullet velocity does not come into play. Therefore, I rule out the wall vent and all other alternate locations for shots 1-8. I am very sure of this.
For shot 10, I am using crack times for all recorders except one, so bullet velocity does come into play. But when I assume ground truth for shot 1 = wall_vent, the TDOA hyperbolas for shot 10 donât converge, and I donât think there is any muzzle velocity that would make them converge. So, that also rules out the wall vent.
Are there any other models?
Why are you sure that your model is the most suitable?
You have published the raw data.
Have your calculations been confirmed by other experts?
I still canât understand why the localization of the first 8 shots should be dependent on shots 9 and 10.
Letâs assume that shots 9 and 10 had not been fired. Then it should still be possible to calculate the origin of the first 8 shots using several audio recordings from different positions. Or is that wrong?
Imagine a scenario with static recorders and with bullet been fired from one position. The delta from the time of arrived from the first bullet and the time arrival of other bullets will be the same on every receiver. Independent from where they are. In this case the method used by Greg will only tell that all the bullets came from the same place. But could not tell where it came from.
Now if you know were one of the bullets came from, that allows to tell when the first bullet is fired on each receiver. Now every other bullet fired from any other place can be positioned. In this scenario if you only have one receiver you can tell that the location is within a constant distance from this receiver or a circumference. if you have 2 receivers the location will be where the two circumferences cross each other that gives 2 possible locations. If you have 3 receivers you can be certain of the location.
That why shot 9 and 10 are so important because we know where they came from.
The method (TDOA) is little more than that because it looks only at time differences and the geometric place were the difference between 2 distances is constant is a hyperbole. And by that you cross the hyperboles of each combination of two receivers to find the locations of the sources.
I would like to reiterate my request to the forum for the height of the bystanders and the height of Trumpâs ear, in relation to the rooftop from which Crooks was firing, which is positioned 15.4 feet above the AGR floor level.
Currently, the approximate figures are as follows: the bystanders are at 12 feet and Trumpâs ear is at 11 feet, with the sea level of the AGR building set to zero. This indicates that the bullet was ascending (rising bullet), which strongly suggests the possibility of two shooters.
Since you two are among the most proficient with audio manipulation/editing, I thought of asking for your help with an issue. Yesterday I finally decided to synchronize 5 bodycams and the PiperGrimley video on my own, in an attempt to better understand the senateâs report timeline and the undercover officers. While watching I was able to confirm cammo-shorts in a leadership position of sorts (he directs a trooper at one point, summons Greg, etc⊠on SERT teams, even a trooper or corporal can be the leader of a team, for instance, so it wouldnât necessarily mean a higher rank⊠maybe). Then, after the discussion yesterday with @vegaspatriot regarding doors, and watching the dashcam footage and seeing Nicol exiting door 9, I remembered DJStewartâs video, a section of it when Cammo-Shorts, who wants to stay at the corner og AGR6, asks Greg Nicol to come join him at the corner. On that moment, Iâve identified some verbal exchanges, but the audio is difficult to understand in full. Could you guys (and everyone else, of course) try to improve our understanding of this exchange?
Iâve tried to isolate vocals in Audacity, but so far I couldnât get much more than whatâs below.
Do observe that hand gestures may improve understanding.
Cammo Shorts: Top or down ? (contâd) unintelligible Nicol: I WAS-??-THERE (I was in there or I wasnât there?) Nicol: First burst of shooting âŠ(fence rattle sound)âŠwonât stop shooting
(wonât stop shooting, wonât start shooting or donât stop shooting?) Someone 1: (asks a questions), unintelligible Someone 2: Yeah, did that happen?
About 3 seconds of pops, cracks, fence rattling
Pause Someone 3: Hey Greg⊠unintelligible (asking a question) Greg (assumed): I did.(?)
