C500: C500 Mark II: Observations from a C300MkII point of view

[puredrifting]

With the electronic image stabilisation in the cameras, it's very important to understand how they work. They're using gyroscopes inside the camera to register the motion of the sensor and then the warp adjustments are based on that information. This gives it a precision that software-based warp stabilisers can only dream of - but it also comes with caveats.

If you bought in to the forerunner of this tech on kickstarter, the SteadXP (as I foolishly did), you'll know that a short shutter angle is ABSOLUTELY KEY to getting good results. The motion blur from a 180 degree shutter angle is just too extreme to not see significant warping issues as the software tries to warp motion-blurred edges back into shape.

At a shorter shutter angle (90 degrees is okay, 45 degrees is better) the software has sharp edges to work with, and becomes much more effective.

That's interesting. You're the first DP I have ever read bringing this up. Makes sense.

 

[Ben_Ericson]

With the electronic image stabilisation in the cameras, it's very important to understand how they work. They're using gyroscopes inside the camera to register the motion of the sensor and then the warp adjustments are based on that information. This gives it a precision that software-based warp stabilisers can only dream of - but it also comes with caveats.

If you bought in to the forerunner of this tech on kickstarter, the SteadXP (as I foolishly did), you'll know that a short shutter angle is ABSOLUTELY KEY to getting good results. The motion blur from a 180 degree shutter angle is just too extreme to not see significant warping issues as the software tries to warp motion-blurred edges back into shape.

At a shorter shutter angle (90 degrees is okay, 45 degrees is better) the software has sharp edges to work with, and becomes much more effective.

Yes, definitely. The shutter angle huge. Maybe it is the wide lens, but those samples look exactly like what I see form warp stabilizer. Jello looking camera sway and dancing lens flares. Does the gyroscopic stabilization work better when it knew what lens was in use, (electronic lens.) I can't imagine it doing the same processing on a 8mm and a 105mm.

I've noticed that warp stabilizer can work incredibly well with film, even at 24p. The rolling shutter inherently has more issues. I guess this would be the same with a digital bolex or F35.

 

[ericsanpablo]

I haven't tried it yet, but theoretically I like the Sony FX9 implementation of stabilization better... similar to SteadXP where you get the gyro data and can adjust the stabilization amount in post. SteadXP software and support was awful/tedious though. I'm curious to see how the FX9's staibilzation works, but haven't really looked into it too much as I like a lot more things on the Canon C series vs Sony (in camera stabilization isn't a make it or break it feature for me, but cool to have if it works).

 

[mantas]

I haven't tried it yet, but theoretically I like the Sony FX9 implementation of stabilization better... similar to SteadXP where you get the gyro data and can adjust the stabilization amount in post. SteadXP software and support was awful/tedious though. I'm curious to see how the FX9's staibilzation works, but haven't really looked into it too much as I like a lot more things on the Canon C series vs Sony (in camera stabilization isn't a make it or break it feature for me, but cool to have if it works).

yea me too, i like the post stabilization more and it works very good with a 90° shutter:

https://vimeo.com/382197335/6f85fc5d54

 

[dipstik]

yea me too, i like the post stabilization more and it works very good with a 90° shutter:

https://vimeo.com/382197335/6f85fc5d54

The problem with the FX9 stabilization is that it only works with a couple of Sony's lenses, and only if the lens IS is turned OFF while shooting. So if you're handing off the footage to a client, what are the odds they're going to download Sony software, and apply stabilization to each individual clip? Canon's implementation gives the control to the camera operator, which I like a lot better.

The other issue with the FX9 is that the stabilization doesn't work with cine lenses, arguably where it would be the most useful. The Canon does, so it opens up the creative doors to using cine primes handheld in creative new ways.

 

[Barry G]

yea me too, i like the post stabilization more and it works very good with a 90° shutter:

https://vimeo.com/382197335/6f85fc5d54

I really have no opinion on the matter (one camera's approach vs the other), having never used the Sony, however I think Matt Porwoll's walking test of the C500II EIS looks a lot like the Sony "post" version you posted...

https://vimeo.com/382689277

For me the Canon's strength is stabilizing the camera when you are trying to hold it still or for doing simple pans, tilts etc, and for that it works really well.

 

[mantas]

The problem with the FX9 stabilization is that it only works with a couple of Sony's lenses, and only if the lens IS is turned OFF while shooting. So if you're handing off the footage to a client, what are the odds they're going to download Sony software, and apply stabilization to each individual clip? Canon's implementation gives the control to the camera operator, which I like a lot better.

The other issue with the FX9 is that the stabilization doesn't work with cine lenses, arguably where it would be the most useful. The Canon does, so it opens up the creative doors to using cine primes handheld in creative new ways.

yes, Catalyst Browses workflow is s*%t, but maybe the NLEs gonna intergrate it.

And it works with every lens, which gives the focal lenght to the cam. I did it with several Canon and Sigma lenses.

I really have no opinion on the matter (one camera's approach vs the other), having never used the Sony, however I think Matt Porwoll's walking test of the C500II EIS looks a lot like the Sony "post" version you posted...

https://vimeo.com/382689277

yea it's not bad. I saw some footage, where it applys the stabilization not to the camera movement, but to the person in front of the camera. that's strange

For me the Canon's strength is stabilizing the camera when you are trying to hold it still or for doing simple pans, tilts etc, and for that it works really well.

yes, that's nice

 

[robotfist]

Since the C300 III is shooting with a 4k sensor, I imagine that when you enable the image stabilization in that camera, you will lose resolution? The camera has to scale the image up to give itself that extra edge around the frame. I wonder how this will impact the C300 III’s image when stabilization is turned on.

 

[cpreston]

Since the C300 III is shooting with a 4k sensor, I imagine that when you enable the image stabilization in that camera, you will lose resolution? The camera has to scale the image up to give itself that extra edge around the frame. I wonder how this will impact the C300 III’s image when stabilization is turned on.

It will be a bit like the EOS R. Pixel peepers looking at a chart will be able to notice it, but you won't be able to notice a difference if you are cutting between stabilized and non-stabilized footage.