Shaft Flex as Seen on Camera (Photos or Video) - Rolling Shutter Illusion

[Golfingdad]

1 hour ago, billchao said:

Very cool way to show how rolling shutter effect works.

Short answer is no, that's not what's happening in his swing.

I'm pretty sure in order to hit the ball while holding the "lag" or wrist angle or whatever, you'd have to drop your upper rotational center since you're basically shortening the radius of your swing arc from your address position. The only pro that comes to my mind with that kind of move is Tommy Gainey.

Right.  Unless he crouched down awkwardly at impact, he'd whiff by several inches with the lag still intact like that.  Frustrating that a guy like Breed, with that platform, would promote something like that.  I'd like to think somebody with his knowledge would "see" that and then question it because he should know that ain't right.

[RFKFREAK]

3 hours ago, Golfingdad said:

Right.  Unless he crouched down awkwardly at impact, he'd whiff by several inches with the lag still intact like that.  Frustrating that a guy like Breed, with that platform, would promote something like that.  I'd like to think somebody with his knowledge would "see" that and then question it because he should know that ain't right.

Breed is an asshat, IMO.

[Golfingdad]

23 minutes ago, RFKFREAK said:

Breed is an asshat, IMO.

I dont really have an opinion because I haven't met him but I tend to think you're probably wrong, partially because of the experience that @kpaulhus and @GolfLug recently had with him in Florida.

But in this case, what he was saying was so glaringly wrong to me, a very much NON golf swing expert, that it's troubling that he wouldn't do a double take before reporting on it.

[Hardspoon]

21 minutes ago, Golfingdad said:

But in this case, what he was saying was so glaringly wrong to me, a very much NON golf swing expert, that it's troubling that he wouldn't do a double take before reporting on it.

Yeah, it seems weird that someone who has presumably seen hundreds of face-on swing videos/snapshots of impact would get that wrong.

Maybe I'm being too generous, but maybe he was just trying to make a point and didn't worry too much about the screen capture reflecting precisely the impact position?  When he drew the line, it was along the back leg, clearly not to a location where the ball ever was...and he had to know that wasn't an actual impact position (right)?  He clearly advanced a frame and could see the ball disappear (I'll admit that weird comment about "the ball is still there" is problematic).

[GolfLug]

11 hours ago, RFKFREAK said:

Breed is an asshat, IMO.

 

10 hours ago, Golfingdad said:

I dont really have an opinion because I haven't met him but I tend to think you're probably wrong, partially because of the experience that @kpaulhus and @GolfLug recently had with him in Florida.

But in this case, what he was saying was so glaringly wrong to me, a very much NON golf swing expert, that it's troubling that he wouldn't do a double take before reporting on it.

Sorry I know this is OT for the discussion. Yupp, MB is absolutely NOT an ass-hat. He was really nice when some of our group guys walked up to him to say hi while he was chipping on the practice green. He is exactly how he is on TV - high energy, smiling, engaging and seems to almost always indulge in full on discussion about a golf swing. Spent a full hour to us talking about full swing, putting, our experiences, his own crazy plane trotting life between Orlando and his home up in NY area, etc.  

[RFKFREAK]

11 hours ago, Golfingdad said:

I dont really have an opinion because I haven't met him but I tend to think you're probably wrong, partially because of the experience that @kpaulhus and @GolfLug recently had with him in Florida.

But in this case, what he was saying was so glaringly wrong to me, a very much NON golf swing expert, that it's troubling that he wouldn't do a double take before reporting on it.

 

55 minutes ago, GolfLug said:

 

Sorry I know this is OT for the discussion. Yupp, MB is absolutely NOT an ass-hat. He was really nice when some of our group guys walked up to him to say hi while he was chipping on the practice green. He is exactly how he is on TV - high energy, smiling, engaging and seems to almost always indulge in full on discussion about a golf swing. Spent a full hour to us talking about full swing, putting, our experiences, his own crazy plane trotting life between Orlando and his home up in NY area, etc.  

I don't mean personally because I haven't met him.

I say he's an asshat because on more than one occasion he's said something on his show that was really suspect, such as you described in your post, @Golfingdad.  In the same episode he was discussing Spieth's second shot on 12 and he was talking about how the grass was cut and how, at least how I took it, it was a contributing factor as to why Spieth's shot went in the water.  I just had to roll my eyes at that.  The way the grass was cut there had minimal impact why Spieth's shot went into the water.  I get you may want to bring up a point about how a blade of grass can impact a shot but if you were talking about Spieth's shot there, I feel there were a few other things he could have discussed that were more relevant.

[GolfLug]

20 minutes ago, RFKFREAK said:

 

I don't mean personally because I haven't met him.

I say he's an asshat because on more than one occasion he's said something on his show that was really suspect, such as you described in your post, @Golfingdad.  In the same episode he was discussing Spieth's second shot on 12 and he was talking about how the grass was cut and how, at least how I took it, it was a contributing factor as to why Spieth's shot went in the water.  I just had to roll my eyes at that.  The way the grass was cut there had minimal impact why Spieth's shot went into the water.  I get you may want to bring up a point about how a blade of grass can impact a shot but if you were talking about Spieth's shot there, I feel there were a few other things he could have discussed that were more relevant.

Oh, ok. 2nd shot from the drop zone? If so, yeah, that does sound a bit suspect.

[Golfingdad]

16 minutes ago, RFKFREAK said:

 

I don't mean personally because I haven't met him.

I say he's an asshat because on more than one occasion he's said something on his show that was really suspect, such as you described in your post, @Golfingdad.  In the same episode he was discussing Spieth's second shot on 12 and he was talking about how the grass was cut and how, at least how I took it, it was a contributing factor as to why Spieth's shot went in the water.  I just had to roll my eyes at that.  The way the grass was cut there had minimal impact why Spieth's shot went into the water.  I get you may want to bring up a point about how a blade of grass can impact a shot but if you were talking about Spieth's shot there, I feel there were a few other things he could have discussed that were more relevant.

That's funny because I watched an event earlier in the year (European tour I think but don't remember) where the announcers went on and on about something very similar.  How the grass affected the shot, specifically how the direction it was mowed would affect the strike.  Oh boy.

(of course now I'm getting way off topic) :P

[RandallT]

Back on topic, I'm trying to think this through more and I'm confusing myself. Here's my thought process, so feel free to laugh if I've got things all jumbled up.

I'd imagine that the speed of motion of the object being photographed and the speed of the upward-moving "black bar" (a property of the camera shutter speed?) dictate the warping that occurs. Change either of those speeds by a little bit, and the resulting picture can probably be quite different. For the propellor, you can change various parameters here and see different warping: https://www.desmos.com/calculator/yc9znckbcg.

With that in mind, in the pic below near impact, I would think that the "black bar" similar to the gif above is traveling from the top to the bottom, right? So as the black bar moves down, the club head gets further and further ahead, making the club look curved forward. That's how I visualize it, anyway.

 

But doesn't that conflict with this picture below? The club is horizontal in this pic at A6, so if this camera was going top to bottom, the club would be captured essentially flat as it overtook the club parallel to the ground, right? 

So maybe this camera has a "black bar" that goes from right to left so that the club head on the left continues traveling downward as the bar moves across to left, making it appear bowed down. Am I overthinking it? 

 

And then to totally overthink it and confuse things, there doesn't look like any bowing of the club below (I can't see the club actually!), so what would make the club lag so far behind his hands in this case (from a camera perspective, I mean)? Wouldn't this camera be essentially using a "black bar" that went from bottom to top, so that the club head is captured first, and then the rest of the body as it moved forward?

The effect below is nearly opposite of the pic above showing a near-impact position. Above, the club head is more forward than expected at impact. And here, the club head is more to the rear. So that makes me wonder what causes that difference.

Maybe I'm just confused right now, but it's an interesting thing to puzzle out. Someone out there has to be thinking more clearly than me on this.

Bonus questions I have: Do cameras document their specs as far as the speed and direction of their "black bar"? What is the effect of the "black bar" even called?! The site i linked to calls it "speed of shutter."  Do some go up/down, some right/left? 

 

 

[mchepp]

@RandallT There are two main components to sensor read outs. Exposure and read out speed. Exposure is how long the pixel is exposed to light. Read out is how long it takes to read out every line (this is the black bar you are referring too).  So to capture motion without blur the exposure needs to be 1/2000 of a second, or even faster. Better is 1/3000s. Then the camera need to read out each "line" of pixels. The camera read out speed is dependent upon how fast the sensors can read out the given line of pixels. So in a high speed camera like the iPhones that can do 250 fps (Frames per second) that is how fast it takes the sensor to complete one line of read out (1/250sec). These two factors do not need to equal each other, this is why Casio cameras that allow you to set the exposure speed are so valuable. You can disconnect the exposure and read out speeds. In low and medium light the exposure and read out speeds equal each other. If they don't your picture will be black. When they equal each other is when the effect begins to show up.

The trouble starts when capturing lower speed video (typically 30fps or 60fps, 30 is the worst offender though). In video settings the camera is doing its best to maximize the exposure so exposure equals read out speed. So now to read out every line of pixels it is set at 1/30sec. The golf swing is MUCH shorter than that. So we the club head is moving in time and has changed position by the time the next line is read out. Here is a picture of Hogan with the black bar moving down. This is not a CMOS sensor, but a top down shutter used in the 50s but ironically the effect is the same. So you get the propeller example you refer too. 

Read outs are going to be different for each camera. For example sometimes the iPhone rotates its cameras in different applications so the "black bar" as you describe it can move either up or down. Not usually left to right though. It could, but less likely.

Lastly, the only way to solve this is something called a Global Shutter. This means all pixels capture the image at exactly the same time, the data is stored in the pixel (every pixel has a small capacitor inside) and then read out at the normal read out speed. There is a trade off for this, as you will lose pixel area to store the extra charge, but in really high speed cameras like the Phantom this is necessary. In some applications rolling shutter affect must be removed because it causes errors. One example is certain bar code reading applications. (Think about the long bars of a barcode) Cell phone cameras need to be small so space trade offs are not possible.

[iacas]

That's not entirely true @mchepp. Shutter speed and line read speed aren't necessarily linked, even for cameras that cost well less than the Phantoms. Some cameras expose the sensor all at once, even at 1/30th exposure. It depends on things like buffer size and the inherent speed of various parts of the system.

[GolfLug]

1 hour ago, mchepp said:

@RandallT There are two main components to sensor read outs. Exposure and read out speed. Exposure is how long the pixel is exposed to light. Read out is how long it takes to read out every line (this is the black bar you are referring too).  So to capture motion without blur the exposure needs to be 1/2000 of a second, or even faster. Better is 1/3000s. Then the camera need to read out each "line" of pixels. The camera read out speed is dependent upon how fast the sensors can read out the given line of pixels. So in a high speed camera like the iPhones that can do 250 fps (Frames per second) that is how fast it takes the sensor to complete one line of read out (1/250sec). These two factors do not need to equal each other, this is why Casio cameras that allow you to set the exposure speed are so valuable. You can disconnect the exposure and read out speeds. In low and medium light the exposure and read out speeds equal each other. If they don't your picture will be black. When they equal each other is when the effect begins to show up.

The trouble starts when capturing lower speed video (typically 30fps or 60fps, 30 is the worst offender though). In video settings the camera is doing its best to maximize the exposure so exposure equals read out speed. So now to read out every line of pixels it is set at 1/30sec. The golf swing is MUCH shorter than that. So we the club head is moving in time and has changed position by the time the next line is read out. Here is a picture of Hogan with the black bar moving down. This is not a CMOS sensor, but a top down shutter used in the 50s but ironically the effect is the same. So you get the propeller example you refer too. 

Read outs are going to be different for each camera. For example sometimes the iPhone rotates its cameras in different applications so the "black bar" as you describe it can move either up or down. Not usually left to right though. It could, but less likely.

Lastly, the only way to solve this is something called a Global Shutter. This means all pixels capture the image at exactly the same time, the data is stored in the pixel (every pixel has a small capacitor inside) and then read out at the normal read out speed. There is a trade off for this, as you will lose pixel area to store the extra charge, but in really high speed cameras like the Phantom this is necessary. In some applications rolling shutter affect must be removed because it causes errors. One example is certain bar code reading applications. (Think about the long bars of a barcode) Cell phone cameras need to be small so space trade offs are not possible.

@mchepp, don't you mean that is the speed for the entire frame readout and not just each line of pixels? If not, I am assuming 30 pixel lines per second wouldn't even complete a frame for the exposure, yes?

I don't know, I always thought a full frame read out completion was necessary for each shutter opening.

47 minutes ago, iacas said:

That's not entirely true @mchepp. Shutter speed and line read speed aren't necessarily linked, even for cameras that cost well less than the Phantoms. Some cameras expose the sensor all at once, even at 1/30th exposure. It depends on things like buffer size and the inherent speed of various parts of the system.

 So, there is not an actual 'line' speed in that case to be even linked to the shutter speed, right? It is one big collective read.

[mchepp]

46 minutes ago, iacas said:

That's not entirely true @mchepp. Shutter speed and line read speed aren't necessarily linked, even for cameras that cost well less than the Phantoms. Some cameras expose the sensor all at once, even at 1/30th exposure. It depends on things like buffer size and the inherent speed of various parts of the system.

Yes, you are correct. I only gave examples from two ends of the spectrum, cell phone and phantom. As you said, there are options in the middle.

2 minutes ago, GolfLug said:

@mchepp, don't you mean that is the speed for the entire frame readout and not just each line of pixels? If not, I am assuming 30 pixel lines per second wouldn't even complete a frame for the exposure, yes?

I don't know, I always thought a full frame read out completion was necessary for each shutter opening.

Yes, you are correct, all of the lines of pixels will be read out in 1/30 seconds. So if the camera is 1280x720 (commonly called 720p) then the readout of line number 1 to line number 720 is done by 1/30 seconds. But remember that it is a "rolling" shutter so it is continuing to read out over and over again.

There is no actual "shutter opening". Digital cameras in video mode use a rolling electronic shutter and do not use a manual shutter. There is a electronic shutter opening, but you probably shouldn't mix that with the thought of a manual shutter opening. 

[LukeDaniel]

On 4/24/2010 at 10:10 AM, steviesouthpaw said:

You're positive about that?

It is an illusion. If you look at a video of airplane propellers spinning, you will be a distortion of the true shape of the propeller so that they looked curved.

 

Luke