The data in this image was built per:

[Using Sound of Target Impact for Acoustic Reconstructions of Shooting Events]
http://www.btgresearch.org/impactsound.pdf
http:// www.btgresearch.org/AcousticReconstruction02042012.pdf

"For example, if the microphone is adjacent to the victim (such as a 911 recording might be), the equation for determining the distance becomes:

t= tb - ts= d/Vb-d/Vs

If the muzzle blast duration obscures the sound of the bullet hitting the target, simple inspection of the sound waveform is insufficient. "

I took their formula, built a spreadsheet,  and plugged in 223 balistic data generated via shooterscalculator.com.

Important to note:

* Presently we don't have information regarding specificaly which weapons and amunition were used.  So the ballistic data was generated with a guestimate 223 configuration.
* My DAW (Sonar) doesn't appear to have the capability of capturing a sound spectrogram like the ones the authors of the study produced; but after reading their commentary on the blast noise obscuring impact noise, I filtered the crowd noise, and filtered/looked alternately for the report and then the high energy impact sounds - and I revised T1 and T2 accordingly.
* More accurate results could possibly be obtained if the corresponding burst sequence on the Taxi-Driver video is identified and aligned, as the taxi-driver's audio contains only the muzzle blast and echo.  It doesn't have the crowd and impact noise to obscure the muzzle events.

======================

The observed acoustic data, having a distance from the shooter of more than 1200ft, refutes and debunks the assertion made by this video:

[LAS VEGAS SHOOTERS DRESS AS COPS - VIDEO PROOF! 100%]
https://www.youtube.com/embed/qMxn7hpXmk4

The man dressed like a cop was NOT a shooter.