Here is a little quiz for you math junkies. Pretend you are a forensic scientist.
Watch the video and estimate the speed of the vehicle to my left.
I can tell you that my cruise control was set at 59 mph and the video frame rate is 30 fps.

94 MPH (+/- 3pmh)

Using the time it takes you to travel between 2 light poles and your known speed, the distance between 2 light poles is 115yds (approximately)

Now using the know distance and the time it takes the car to pass between 2 light poles gives the speed.

using the same math the car on your right passed you at78 mph (aprox)

Without a stop watch it is tough, but I guestimmate 100+, using the rough method London808 did, looks like he is going about 2X the speed.

If this were an actual investigation I would get a video downloader software and grab the video file.
Then run it on a video editor so I could get a pretty accurate frame count between light poles.
One video frame is 1/30 sec.
Then go to the scene and measure the actual distance, if I could find the actual scene.
If that were not possible, I would ask the dept. Of transportation what is the typical distance between light poles.
-----------
another technique would be to look at the audio bump because that would tell you exactly when the car passed.
Then use that as a reference. see pic.

Anybody else wanna play CSI?

I can tell you from experience it takes a car 3 lanes to go around a bend on H2 going North traveling @ 164 mph uphill.

85mph.

By the Radford overpass, westbound.

100 yards to the 2nd light pole, 90 yards to the 3rd, 80 yards to the 4th.

My calculations disagree with london and zip. How you figured 85 zip?

I don't even know what to say about suka on the h2. 

Tried to time it again, hard to tell the distance and time.  I figure he goes past 8 reflectors for 280 feet in 1.9 to 2.1 seconds which is about 90 to 100mph.

85mph.

By the Radford overpass, westbound.

Using Google earth and the Radford exit tip, looks like the light poles average about 300' apart.   For the timing (after remembering my watch has a stopwatch function) I used the lightpoles right after the overpass which GE has at 304'.

Two timings of the camera car had it at 55-65 mph, so the average is right on.  The fast mover I got at 120 mph, over several timings. 

About the same as my very rough guesstimate of 2X speed just mentally counting seconds difference between the passing time of the two cars.

I would suspect the 30 fps is only useful of you have the original recording.  Dunno if frame rate is preserved on uploading.  The 59mph was good for mental estimate, and for double check of stopwatch/distance method.

Using Google earth and the Radford exit tip, looks like the light poles average about 300' apart.   For the timing (after remembering my watch has a stopwatch function) I used the lightpoles right after the overpass which GE has at 304'.

Two timings of the camera car had it at 55-65 mph, so the average is right on.  The fast mover I got at 120 mph, over several timings. 

About the same as my very rough guesstimate of 2X speed just mentally counting seconds difference between the passing time of the two cars.

I would suspect the 30 fps is only useful of you have the original recording.  Dunno if frame rate is preserved on uploading.  The 59mph was good for mental estimate, and for double check of stopwatch/distance method.

...
That's  good thinking. I never considered Google earth. That could be a useful tool.
Also, you had a good question about the frame rate being preserved when uploaded to youtube.
I can check that.....

Ok then
If nobody else wants to try, here is my analysis:

First of all youtube transcoded my original file to make it smaller. but the frame rate was the same.
It also retained the original 720p size.
So you can indeed download a video such as this and analyze it using common video editing software.
I used a cheap old version of Sony Vegas. I gives a readout in frames for timing accuracy of 1/30th of a second.

From the time that the car passed me to the first light post, his average speed was 103 mph.
Then from the first light post to the next light post his average speed was 125 mph.
The dude passed me and continued to accelerate at a brisk rate.

When timing the entire duration that the car was visible, his average speed was 115 mph.
So ROBTMC was pretty accurate with his estimate of 118-120 mph without even using software.
Well done sir.

Well done sir.

Olde farte engineer geeks are just drawn to this sort of  stuff. 

You never really forget your first physics courses where math suddenly can be used to predict how the real world behaves.

Thank God it did not require me to try and remember vector calculus.

He passed 5 lane markers in right about 2 seconds.

Each marker is 10 feet long and they are 30 feet apart, so 40 x 5= 200 ft traveled in 2 seconds. 100 feet per second to make it easy.

100 ft/second = 68 mph.

Then again, I wasn't super accurate about the time, but if someone could in slow motion and a better timer, that is the way to find out the speed.