Paintball

[drg]

Oh so suddenly, I've sped up the video, and my integrity is in question, even though it shows essentially the same thing as the two other videos? Gimme a freakin' break.

This is taking the 'nothing you post will be believed' thing to a new level. I give you a better video, at a higher resolution, showing much more and suddenly it's not believable. GEE.

Do you NOT need to verify backblock movement in the video, you shouldn't anyway. Use the SOUND data as its sampling rate is MUCH higher than the video sampling rate.

[drg]

Originally Posted by manike

We once took a gun (years ago now) that knowingly added shots based upon a "debounce" setting and asked people when they shot it if they felt it was semi. We could get to the point where we added 0.5 shots per pull after they achieved 5bps or more and people still believed it was all themselves.

It was actually hard not to laugh when people said that at higher settings they felt the gun was missing inputs, and at the lower settings they felt the gun was more in line with what their fingers were doing.

Well if you weren't so busy being smug, you'd realize why this is. The brain can sense minute disturbances in rhythm, and it can tell when the shots are not corresponding to the finger motions. It might be firing too fast to make out that it is actually getting EXTRA shots but the brain had indeed sensed that one pull is not translating into one shot.

BTW a 50% speedup in firing rate would be VERY easy to detect on 30fps video.

[manike]

drg. What time frame on that new video would you like us to analyze? At what time frame are you really happy with the speed? Actually on both videos would be good so that we are all looking in depth at the right areas.

I'm trying to use tools that are available online for everyone so that the analysis can be duplicated and verified by all. Independent verification is always best.

The trouble with any video that gets posted, and mine included is no-one knows what's been done to it. That's why posting the best video that is the best resolution and data is the best way to be believed. The more information and data the better. That's always the case when you are trying to "prove" something beyond all doubt. The slow frame rates make things look choppy and don't look as genuine. I'm sorry but that last video looked bad, but it may be on my machine. It looked very choppy. The first video you posted was smoother.

hmmm maybe I should make some videos and speed some up and slow others down and see if anyone can guess which is which... That would be another interesting psychology test.

Not enough hours in the day to do everything I want to with this camera.

Originally Posted by drg

BTW a 50% speedup in firing rate would be VERY easy to detect on 30fps video.

That's true, only I was talking about in person. Humans couldn't detect it, and didn't want to believe it was anything but themselves. If you re-read what I posted, I said the people couldn't detect the extra shots. They thought the extra shots were genuine, and the real count was missing...

[manike]

Originally Posted by drg

I'm sorry, but YOUR new video shows nothing of consequence.

Really, so then why say this?

Originally Posted by drg

So what if that happens? Does it result in one shot or two or more? If one shot, then it wouldn't be counted in sound-based measurements, so it's insignificant. If it results in 2 shots, then each finger motion is a legitimate trigger pull even though the trigger doesn't appear to come up (as long as it crossed the triggering point). If it results in more than 2 shots, THEN we have significant bounce. That would be detectable even at a MUCH MUCH slower framerate.

Sounds like a lot of 'consequences' to consider.

Is each movement a legitimate pull even if the trigger doesn't move back forwards?

If the gun detects that as two shots, what does that mean for other times when this doesn't happen?

[manike]

Originally Posted by drg

Do you NOT need to verify backblock movement in the video, you shouldn't anyway. Use the SOUND data as its sampling rate is MUCH higher than the video sampling rate.

Finally! Now you see our point. This is what we have been getting at all along.

The sound data is good to analyse ROF, but not finger inputs.

The visual data is NOT good enough to analyse ROF, and in this case it's really not even good enough to analyse finger inputs and gun cycles.

I guess we could close this there, but I have more data I've been working on to back this up too.

[cockerpunk]

Originally Posted by drg

Well if you weren't so busy being smug, you'd realize why this is. The brain can sense minute disturbances in rhythm, and it can tell when the shots are not corresponding to the finger motions. It might be firing too fast to make out that it is actually getting EXTRA shots but the brain had indeed sensed that one pull is not translating into one shot.

now thats not true at all.

i can stick my WAS board on the lowest debounce and rock the trigger hard, and it still feels like im shooting it every time it goes off.

then i put it back to legal and im ashamed of how slow it is.

the brain might be good at picking up rhythms but it is terrible at judging rates. you can hear the rhythm and its easy to hear a difference in speed, and to tell how consistent it is, but somthing on the order of "perfect pitch" would be needed to be a human pac timer of both the gun and your fingers. i call it perfect pitch for lack of a better term, but again, its that the human mind is good at judging differences and comparisons, but terrible at pulling out numbers.

[drg]

Originally Posted by manike

drg. What time frame on that new video would you like us to analyze? At what time frame are you really happy with the speed? Actually on both videos would be good so that we are all looking in depth at the right areas.

I'm trying to use tools that are available online for everyone so that the analysis can be duplicated and verified by all. Independent verification is always best.

The trouble with any video that gets posted, and mine included is no-one knows what's been done to it. That's why posting the best video that is the best resolution and data is the best way to be believed. The more information and data the better. That's always the case when you are trying to "prove" something beyond all doubt. The slow frame rates make things look choppy and don't look as genuine. I'm sorry but that last video looked bad, but it may be on my machine. It looked very choppy. The first video you posted was smoother.

hmmm maybe I should make some videos and speed some up and slow others down and see if anyone can guess which is which... That would be another interesting psychology test.

Not enough hours in the day to do everything I want to with this camera.

There is no way to verify this, something I've stipulated to before in this thread. There has to be a baseline of trust in order to do ANY analysis. Scientists are not immune from moral lapses, and we have seen many scientists' research turn out to be fake.

The reason this video looks different from the other is it was taken with a different camera. The both film at ~30 fps, it's possible that the second one films at 30p rather than 60i. The difference in 'look' is the lack of stabilization on the second camera.

All "full speed" videos are unedited other than trimming.

Originally Posted by manike

That's true, only I was talking about in person. Humans couldn't detect it, and didn't want to believe it was anything but themselves. If you re-read what I posted, I said the people couldn't detect the extra shots. They thought the extra shots were genuine, and the real count was missing...

I don't see where what you posted says that:

It was actually hard not to laugh when people said that at higher settings they felt the gun was missing inputs, and at the lower settings they felt the gun was more in line with what their fingers were doing.

Do you know whether the gun was indeed translating inputs properly? Inputs meaning not just electronic inputs but that motions of their fingers were being translated into shots?

[Shane-O]

Manike

Something interesting I seen in that Mini video, the second finger has a little bounce off the trigger to it after it hits the trigger, that could cause extra switch noise, and add a shot couldn't it?

[drg]

Originally Posted by manike

Finally! Now you see our point. This is what we have been getting at all along.

The sound data is good to analyse ROF, but not finger inputs.

The visual data is NOT good enough to analyse ROF, and in this case it's really not even good enough to analyse finger inputs and gun cycles.

I guess we could close this there, but I have more data I've been working on to back this up too.

No, no, no, that's not a 'gotcha' at all. The sound data just renders watching the backblock (which is traveling far faster than my fingers) needless and can be used to judge the cycling of the marker with great accuracy.

[Shane-O]

Originally Posted by manike

We once took a gun (years ago now) that knowingly added shots based upon a "debounce" setting and asked people when they shot it if they felt it was semi. We could get to the point where we added 0.5 shots per pull after they achieved 5bps or more and people still believed it was all themselves.

It was actually hard not to laugh when people said that at higher settings they felt the gun was missing inputs, and at the lower settings they felt the gun was more in line with what their fingers were doing.

Part of the problem is that people want to believe they are that fast, therefore they are.

drg:
I highlighted the important parts for you.................