Author Topic: Recording Frames per second and playback  (Read 2974 times)

rdemyan

  • Newbie
  • *
  • Posts: 44
    • View Profile
Recording Frames per second and playback
« on: August 02, 2022, 09:40:25 PM »
I am recording the movement of particles using the Chronos 2.1.  I need to determine the velocity.  I set the 2.1 to record at 640 x 96 which means the recording frame rate was 24,046 FPS.  But when I try to measure the movement of the particles frame by frame I'm getting velocities that I know are too high (there is a ruler in the frames).  I've recorded before at 24,046 FPS and not had a problem.

The only thing that I think I did differently was I reviewed the footage on the Chronos 2.1 before saving.  This time I set the playback FPS on the Chronos 2.1 to 25 instead of 60. I then saved the footage with the playback FPS still at 25.  Does that affect the time differential between frames?  The time differential between frames should be 1/24,046 or about 41.6 microseconds.  This time differential is in the denominator when calculating the velocity.  It is critical to calculating the correct results.

The only other thing I can think of is that the Chronos was on for a couple of hours including  an hour where I was off doing something else, but I left the camera on.  Does the FPS degrade if the Chronos gets hot or is on too long.  I don't think it was hot, but I didn't really check.  It didn't feel hot when my experiment was over and I put the Chronos away.

On the datasheets, why does Krontech list the FPS as "Max FPS"?
« Last Edit: August 02, 2022, 09:44:38 PM by rdemyan »

Nikon1

  • Hero Member
  • *****
  • Posts: 983
    • View Profile
Re: Recording Frames per second and playback
« Reply #1 on: August 03, 2022, 02:42:06 AM »
So, the Interval between Two Saved Frames will stay the Same, no matter what Framerate you save it at (more on that later down this post).
 Question is more, what you use to skip to the next frame, and if that software might have skipped a Frame, which could happen if Settings are wrong or the Software isnt very advanced.
 .
 Temperature and Runtime should not affect timing of the Camera by any kind of meaningful Amount that you would be able to notice at 24kfps.
 As Reference, this site here: https://www.electronicdesign.com/technologies/analog/article/21798809/minimize-frequency-drift-in-crystals
 Does Mention, Frequency Drift of Quartz Oscillators can drift up to 100 Parts Per Million from a Temperature Change from -40 to +85C; which is a Very Substantial Change in Temperature.
 In this assumed very Extreme Case, the Drift would mean a Difference of 2.4fps at a Framerate of 24046fps. Or in other words a 0,1% Change in Timing Intervals which include Frame Intervals and Exposure Times.
 To counter this Effect of Temperature Drift on Very High Precision Electronics, that need Higher or Much Higher Accuracy than that and cant allow such Drift, the Quartz Crystal is usually Heated (because it is easier to keep it at a generally higher, but highly Stable Temperature than the Environment is expected to ever be, than to figure out how to Heat AND Cool it Reliably to then be within a very small Margin of Temperature), See Crystal Oven: https://en.wikipedia.org/wiki/Crystal_oven
 Given the Chronos is made to run at its warmed up Temperature, and it itself Generating quite some Amount of Heat, it kind of works as Sort of a Low-Effort Crystal Oven itself when it has Reached a Stable Operating Temperature after something like 30 to 60 Minutes Runtime. This is generally recommended Anyways, as a Stable Temperature generally means the Sensor Noise Pattern Will not Change, resulting in A image with Less Noise usually (if Black Calibration is done once warm).
 So, if anything running the Camera for a while before Recording anything is a good thing. Should also be the most reliable way to get the Best timing Accuracy out of the Cameras Internal Timing.
 If you need anything much more Precise than the 0,1% Accuracy that is to be expected of a Typical Quartz Crystal Oscillator, you very much probably want to Trigger the Camera with an External Signal Anyways, i assume.
 .
 As for "Max FPS":
 You can very much run the Camera at a lower Resolution like 640 x 96 and run any Framerate down to 60fps, which is currently the Lowest officially Supported Framerate. You can Set lower FPS down to 1.0fps as an Experimental Setting, but it is not guaranteed to work properly or at all (it usually does not, but the Camera does allow you to do so anyways).
 The "Max FPS" just Literally means, that this FPS most People would usually want to shoot at, is the Highest Possible for the respective Resolution Setting.
 There are some Cases, where you would want Lower FPS or some very Specific FPS, like lets say exactly 24000fps, not 24046fps, so you could actually do that if you wanted to.
« Last Edit: August 03, 2022, 02:43:42 AM by Nikon1 »

rdemyan

  • Newbie
  • *
  • Posts: 44
    • View Profile
Re: Recording Frames per second and playback
« Reply #2 on: August 03, 2022, 11:29:43 AM »
You are correct.  The problem was with the editing software. I'm using AShampoo Video Optimizer Pro 2 (perhaps you know the software since the company is based in Germany).  I had unknowingly changed some of the settings and that caused the problem.  Problem solved. Thanks.