HPX250: Fast Motion AVCIntra-100 Footage - 1080/30pN - Football
- Thread starter David Saraceno
- Start date
Interesting, and thanks for the update Barry.
So at 1080/24pN, we optimally should shoot at shutter 1/48.
I shot some fast action 1080/24pN two weeks ago, and I thought the results was surprisingly soft, but I am thinking it actually blurry.
So at 1080/24pN, we optimally should shoot at shutter 1/48.
I shot some fast action 1080/24pN two weeks ago, and I thought the results was surprisingly soft, but I am thinking it actually blurry.
Last edited:
Dino
Veteran
David, you definitely want to avoid shooting 24fps with shutter off, unless there's no camera or subject motion. This is why the footage looked soft.
Coming from the 170, I now see that the 250 does require a different approach.
Sorry for this off-topic, but there's something I'm certainly missing:
I've read several times that skew effect won't be reduced by setting faster shutter speeds, some say this would even worsen things because of blur reduction (cleaner contours of leaning objects) and that's clear.
I'm trying to understand what exactly happens (on timing basis) during frames creation with CMOSes, but I found figuring it out pretty hard, plus, I have no CMOS camcorder to test.
Barry says that frame rate has a lot to do with Rolling Shutter effects, and that's where I get confused:
If "Rolling" speed and shutter speed are two independent things, is there a specific relationship between rolling speed and frame speed, then?
If rolling/scanning speed decreases at slower frame rates, some possible undercranking panning should be strictly forbidden...:undecided
In this Hpx300 demo video, at 03:33 they say: "In some instances, shorten exposure times may reduce the amount of skew".
Are they lying?
They show a 1/500 of second clip... With skew!
I've read several times that skew effect won't be reduced by setting faster shutter speeds, some say this would even worsen things because of blur reduction (cleaner contours of leaning objects) and that's clear.
I'm trying to understand what exactly happens (on timing basis) during frames creation with CMOSes, but I found figuring it out pretty hard, plus, I have no CMOS camcorder to test.
Barry says that frame rate has a lot to do with Rolling Shutter effects, and that's where I get confused:
If "Rolling" speed and shutter speed are two independent things, is there a specific relationship between rolling speed and frame speed, then?
If rolling/scanning speed decreases at slower frame rates, some possible undercranking panning should be strictly forbidden...:undecided
In this Hpx300 demo video, at 03:33 they say: "In some instances, shorten exposure times may reduce the amount of skew".
Are they lying?
They show a 1/500 of second clip... With skew!
Barry_Green
Moderator
Shortening the exposure time will not reduce the skew.
Some cameras have more skew at slower frame rates. The hpx300 is an example, it is MUCH more skewy at 1080/24 than it is at 1080/60i. The gh2 is another example, a lot more skew at 1080/24 than at 720/60.
The hpx250 doesn't care; it has the same minimal skew at all frame rates.
Some cameras have more skew at slower frame rates. The hpx300 is an example, it is MUCH more skewy at 1080/24 than it is at 1080/60i. The gh2 is another example, a lot more skew at 1080/24 than at 720/60.
The hpx250 doesn't care; it has the same minimal skew at all frame rates.
Thank you for your kind answer Barry, it is good to know that Hpx250 has the minimum amount of skew regardless of the frame rate.
So, back to the exposure time, what has been told in that famous demo video is definitely not true.
Do you know if there's a way to know (approximately) how long does a complete rolling scan last?
I may think that it should be a very short time, maybe quicker than the shortest exposure time available, but I'm not sure about this.
Could it be a fixed readout time, whatever the frame rate (and the shutter speed) is set at?
Hpx250 behaviour would seem to justify this.
Thanks in advance, I'd really like to better understand Rolling Shutter timing logic when it comes to build frames by reading out the electric values row by row...
So, back to the exposure time, what has been told in that famous demo video is definitely not true.
Do you know if there's a way to know (approximately) how long does a complete rolling scan last?
I may think that it should be a very short time, maybe quicker than the shortest exposure time available, but I'm not sure about this.
Could it be a fixed readout time, whatever the frame rate (and the shutter speed) is set at?
Hpx250 behaviour would seem to justify this.
Thanks in advance, I'd really like to better understand Rolling Shutter timing logic when it comes to build frames by reading out the electric values row by row...