@greg_n, do you have any other info for Ross and TMX locations other than their on footage? Did someone identified them on other footages?
Based on this photo I found this location for the cruizer (586793.39 m E;4523525.12 m N):
Edit: Also, this source is subject to sound diffraction, by my calculations the apparent distance increase from 26.34m to 26.60m.
Edit2:
Could Ross be one of these 4? This is 1 sec before 1st shoot.
Been having a look as well. One problem when looking at file properties/details is a comment: āCreate videos with https://clipchamp.com/en/video-editor - free online video editor, video compressor, video converter.ā. The video frame rate has also been set to 30 fps. With the video files I have gotten from TMZ so far, the frame rate is set to 29.97. One thing that can be causing this frame offset problem is in how clipchamp has transcoded from 29.97 to 30 fps.
i am seeing some big differences in what frames are being used. It looks like the source file was set at 60fps with different frames dropped between the two different versions you are using. There are significant differences with what frame image is present as the man moves in front of the camera.
I am picking up a few frames of the thrown case in both those videos. They are not clear at a lower resolution. The more visible case moments is consistent with the enhanced version.
Hi Rough_County_Gypsy,
I just wanted to let you know that you donāt have to get the drone overview if you risk some kind of illegal action. Donāt do anything illegal, itās not worth it.
However, if you find legit way, then make sure you get high enough with your drone. Maybe you have noticed when you are on the side of a picture taken by your cell phone you look extremely fat. The people in the middle look great. This is most likely due to the wide-angle mode.
So, if you could get a shot (or video crossing the shot line of Crooks) at a very high altitude without wide angle modus activated, that would be the best and you would be the hero!
Good luck! 
Thanks for the feedback.
Yes, like I explained, I cut the video with clipchamp and then saved it in HD. If you have the feeling the frames got messed up by doing so, then I think it would be best to grab the screenshots directly from the two originals and start to compare each frame by frame? I think that should be possible with VCL media player.
Before I waist a lot of time, do you agree with my approach?
Interesting language. How do you know that the suspect was voluntarily there in order to kill people?
Itās weird that it looks like the dark clump explodes mid air.
As if there is an impact of force propelling it up and then a subsequent explosion/burst.
Yes. Going back to the video file before clipchamp touched it is a good start. I recommend using Open Shot to export the image frames, It is free open source video editing software. Easier to export video frames to image sequences than Shot Cut.
I already started with VLC and got many more frames from the originals and there is no frame mismatch. Thanks for the tip!
At a first glance:
Frame 67, 68 and 69 perfect match
Frame 252 (almost at the end in the trees) perfect match
And I couldnāt help from peaking (spoiler alert) Frame 177:
Here are the files:
If Ross is the guy behind the tree than this is his location (586683.91m E;4523542.97m N).
A cross reference the footage with line of sights of both footages to find Ross location. Here is the footages project on the ground matching the line of sights of the footages:
Edit: By the same method here is TMX position (586741.77 m E;4523497.44 m N):
Audacity seems to do a good job with its filters and keeping the data centered.
But loop delay is no good. On one of our programs, we ran all kinds of signal processing to figure out where we should be headed including feedback filters, but at the end, we had to go estimate where we thought that spot was in the raw, undelayed signal input, as guidance couldnāt tolerate even the smallest lagāfor them, better a noisy, timely signal than a smooth delayed one.
Thank you so much Roger. I have downloaded your materials and will start there.
@VT, is the person you are calling Ross also what is called Source 4 in the data repository here? And they are stationary during the entire first 8 shots? Because there is a huge discrepancy (30 ft) between Source 4 and the podium when you compare arrival times of the first shot group and the second shot group. I had not really worked hard to chase this discrepancy down because I had assumed Source 4 was moving.
All the U.S. Reps. are there this week investigating, so I decided to stay away. Iāll be joining a retired Pennsylvania State Police investigator in Butler on September 9th. Iām hoping to gather additional information on the AGR building. I should be able to drone then.
I sure wish we had floor plans for buildings 1 (AGR 6) and building 2. Is your magic wand fully charged for the visit? 
Speaking of building numbers, it would be nice to know the numbers AGR assigns to all of their buildings.
Ross is source 3.
Wow, I really respect your efforts!
Since you will be there, maybe you can ask around and find out who this guy is?
He has all 8 shots on video (or maybe only 4 shots? 
)
His T-shirt shows UPMC Rehabilitation Institute in PA
Could you elaborate more on this, maybe by circling the white dot you mention? As I understood, weāre talking about that video recorded from the bleachers. On that video, when thereās a first audible sound from the 2nd shot, you see a white dot where Crooks was, and it disappears 13 frames after? Did I get it right? What was the duration of that white appearing dot? Did it last for the whole 12 frame difference before it disappeared or was it just a blip? Did it appear on the crack or on the thump?
Weāve been brainstorming with complete strangers over the internet for over a month, and youāre scared of AI bias? ā¦Skynet didnāt agree with you on HighRoof, did it? 
(sorry, I couldnāt resist it).
If you have updated elevations, I would like to check them with the ballistic curve, same as I did with Rogerās Vent and Crooks scenario on post 1123.
Thanks, @vt1 . Yes. Analysis I did of āRoss!ā (Michael Difrischia) and Grimley videos indicates that the man you point to with the arrow is Michael Difrischia, the man who recorded the video purchased with exclusive rights by TMZ.
Iām not totally following how you used the perspective lines to locate the TMZ (Michael D.) and TMX locations for shot 1, but you certainly seem confident in your results, so I will use them for my next iteration.
Compared to what I used in my previous iteration, the differences will be:
Ross: 1.97 m
TMX: 0.50 m (very close, so Iām not going to change my points for that one)
DJStew: 4.54 m
Cruiser: 1.19 m
Podium: 5.02 m
Those are some pretty significant differences, so Iāll be surprised if they resolve to something that makes sense for all 10 shots, but weāll see what happens.
Also, Iām thinking about including the audio from NTDās coverage in my next analysis, but Iām having trouble discerning whether the audio is:
At this point, Iām thinking #3, which would make it difficult to use because how do I discern the crack/boom times? Has anyone else spent time analyzing the NTD video and has insight to provide?
@vt1 , thank you. I am just trying to refine my TOA numbers now with known positions of the microphones. I saw you also gave a position for TMXāI wanted to double check what video that corresponds to and if they changed locations during the gunfire volleys.
In essence, for raw arrival times NONE of the sources except for the fake popo audio, when compared to the podium, point to a common location for shots 1-3 as compared to shots 4-8. A final step, though, is to account for distance changes in microphone location as a possible source of some (or all) of the discrepancy. @greg_n and I vary in our estimates of how many shooters there are by one.
You seem to be doing a wonderful job at triangulating from multiple views where these microphones are.
Once I have fairly accurate movement numbers for the sources and correct for those, I will publish my time of arrival data in full.
Here are source 4 locations:
From 4 to 9 its very close to differentiate.
| ref | E | N |
|---|---|---|
| source4@1shoot | 586693 | 4523528.92 |
| source4@2shoot | 586691.84 | 4523529.6 |
| source4@3shoot | 586690.6 | 4523530.21 |
| source4@4shoot | 586683.23 | 4523537.98 |
| source4@9shoot | 586682.84 | 4523539.08 |
| source4@10shoot | 586683.83 | 4523544.57 |
