Shutter type without camcorder info

[nji]

I've some videos where there is no camcorder info embedded,
and I where I need to know (for DeShaker, see thread above)
if they are done by rolling or by global shutter.
Is there a reliable way to determine that?
(I could do Deshaker with and without rolling shutter option
and look at the result ... if "wobbling" at quick camera motions.
Which reminds my on a Calvin & Hobbes strip, where Calvin asks his dad
how they know how much weight each bridges tolerate.
"Well they drive heavier and heavier trucks above, until ..." :)

[poisondeathray]

If it's a modern camera in the last 10 years, 99.99% likely it will have CMOS sensor with rolling shutter . Very few newer cameras use CCD, even fewer have CMOS with global shutter

If you move the camera around and there is "jello" in the shot, wobbling, very likely CMOS and rolling shutter

[nji]

Thank you.

Unfortunately some of the videos are surely older than 10 years (20, 30, ...), and I don't know anything about the used camera.
Some of the movies may even be digitalized ... (what about shutter type then?).

I wonder if it's possible to determine shutter type only by the movie itself?

[poisondeathray]

Thank you.

Unfortunately some of the videos are surely older than 10 years (20, 30, ...), and I don't know anything about the used camera.
Some of the movies may even be digitalized ... (what about shutter type then?).

I wonder if it's possible to determine shutter type only by the movie itself?


You can often tell just by looking at it.

But there are often clues, just based on what the "video" is about, or the format.

Is it really a "movie", as in theatrical movie ? Or "home video"

Older home videos, maybe 10-25 years ago which were shot on DV, interlaced almost always used a CCD sensor.

A "movie" that is digitalized does not necessarily change the shutter characteristics . But a film that is scanned can have different types of artifacts (warping) similar to RS artifacts depending on the process used. Or a projected film that is captured by a camera with CMOS partial scan sensor can have RS artifacts (eg . partial exposures, but you're probably not physically moving the camera in that setup)

[nji]

The date of the home movies is approx know, but not the used camera then.

And some of them - as said - seem to be moved from analogue. Where I do know even less.

I had hoped there would be some signs from which to determine
(Like progressive vs interlaced in here
https://video.stackexchange.com/questions/12980/how-to-see-if-a-video-file-is-progressive-or-interlaced)

But it seems much more difficult ...

[poisondeathray]

If it's multigenerational - there are potentially many issues, but if you want to post a sample , someone will probably take a look and help you out

[nji]

It's not about one or two movies, but a whole lot I have here from different times in the last about 30 years.
So I asked myself if it would be possible to get a hint from the content of the video if it's done by roller or global.
Something like the "test" on interlaced/ progressive I posted above.
Maybe there are some characteristics of rolling shutter when you have a look on a scene with quick motion, and compare adjacent frames 1:1?

(BTW The day before I used DeShaker on a clip where I thought it was taken with rolling shutter (as done 8 years ago).
But DeShaker's result was much better with global shutter setting.
But it's very uncomfortable to generate both DeShakings, then identify quick scenes, then compare).

[shekh]

Did you know, CCD may overflow in bright light, like here (notice vertical pink stripes).
Well, I'm not sure, maybe this specific artifact is caused by something else, but this was shot on CCD :)

https://s14.postimg.cc/lo5fv0x29/IMGP0046.png

[poisondeathray]

Did you know, CCD may overflow in bright light, like here (notice vertical pink stripes).
Well, I'm not sure, maybe this specific artifact is caused by something else, but this was shot on CCD :)

<snip>

Yes, this is caused by CCD. It is very common - known as "CCD smear" .



It's not about one or two movies, but a whole lot I have here from different times in the last about 30 years.
So I asked myself if it would be possible to get a hint from the content of the video if it's done by roller or global.
Something like the "test" on interlaced/ progressive I posted above.
Maybe there are some characteristics of rolling shutter when you have a look on a scene with quick motion, and compare adjacent frames 1:1?

(BTW The day before I used DeShaker on a clip where I thought it was taken with rolling shutter (as done 8 years ago).
But DeShaker's result was much better with global shutter setting.
But it's very uncomfortable to generate both DeShakings, then identify quick scenes, then compare).

The older the content, the less likely it has rolling shutter issues. (But film transfer artifacts , warping can sometimes mimic similar issues)

The classic test is checking a building or tree (ie. vertical lines) during a pan . If it had RS artifacts, that's also how you'd measure the skew and calculate percentages to estimate scan rates

The other classic example is shaking or vibrations - you should see "jello" or vibrations in different directions.

[nji]

That's all new to me. :thanks:

Actually I had a closer view at the mentioned video - and at special high lighted scenes there is that CCD smear too!

It seems that for an experienced eye it's quite easy to differ roller from global.
But unfortunately I am not experienced in that at all.

I thought there maybe an easy way to differ the both without "looking experience". Especially to assess that jello effect is quite uncomfortable.

(Even more comfortable would be a filter, determing roller (% for DeShaker) vs. global, progressive or interlaced, ...)