How to test motion blur for high speed moving objects

[adien]

Hello all,

I'm new here and this is my first post. I hope this topic hasn't already been covered elsewhere in the forum. If it has, my sincere apologies for cluttering the space.

I'm an electronics developer with roughly 10 years of experience in designing and developing various systems. Currently, I'm working on building a camera system that will capture detailed images of fast-moving objects—we're talking speeds of 45 m/s.

I’ve tried Googling to find a simple setup to test what shutter speeds would be optimal to eliminate motion blur at different velocities, but it seems I’d need to build a linear rig. At 45 m/s, that involves some serious kinetic energy.

Can I use any of the tests on TestUFO .com to evaluate motion blur and somehow relate that to real-world object speeds?

For reference, I have an Odyssey G7 monitor with a 144 Hz refresh rate, HDR support, and a 5070 Ti card.

Thx

[RealNC]

Moved to correct forum section.

[Chief Blur Buster]

Can I use any of the tests on TestUFO .com to evaluate motion blur and somehow relate that to real-world object speeds?

Short answer: Depends. It can be Yes.

It depends on exactly what you're trying to do, if you're comparing display motion blur versus real-world motion blur.

There's an equivalence between sample-and-hold refresh rate and real world objects.
But it assumes GtG=0 (e.g. OLED) to eliminate this error margin.

You need to get familiar with the known error margins, and configure your scientific variables correctly. Non-zero GtG is a big error margin at high Hz. Imagine the 480Hz LCD, whereupon refresh cycles are only ~2ms, but a LCD GtG of 1ms before/after blurs it up more, making it more equivalent to a 1/200sec photo instead of a 1/480sec panning-camera photo. To eliminate GtG-derived error margin, you must absolutely upgrade to OLED or some nigh-0ms-GtG display.

For accurate persistence motion blur that equates to camera shutter speed, you want an OLED (GtG=0). A non-zero pixel response means a pixel is fading from one color to the next -- just like a slowly-moving camera shutter.





www.testufo.com/photo and www.testufo.com/map (street label readability test etc)

• 120fps 120Hz OLED = panning image has same blur as 1/120sec shutter of a real-world panning camera photographing real life
• 240fps 240Hz OLED = panning image has same blur as 1/240sec shutter of a real-world panning camera photographing real life
• 480fps 480Hz OLED = panning image has same blur as 1/480sec shutter of a real-world panning camera photographing real life

While displays are not as big FOV as a real world scene, the motion blur equivalence is capable of becoming scientifically identical at GtG=0, so if you are going to involve displays in "realworld versus display" motionblur analysis, I recommend switching to an extreme-high-Hz OLED display.

Also, learning about pulsewidth blur -- same pixel visibility time matters -- at www.testufo.com/blackframes
This follows Blur Busters Law (where pixel visibility time equates to shutter time, whether it's a continuously displayed frame, or a pulsed frame, aka BFI).

Contractor Available For Hire

Short-term, 20/40/80 hour contracts, etc. Also, I am back from DisplayWeek and now available for services hire. See services.blurbusters.com if you need further help with the sciences or creating custom tests in TestUFO, etc.

[gogofly]

Hi, I read the peer-revewed paper about Manual Pursuit Camera.but I don't understand，Why the pursuit speed is correct，the images of the synchronization markers are aligned exactly vertically .It's way over my head.

[Chief Blur Buster]

Hi, I read the peer-revewed paper about Manual Pursuit Camera.but I don't understand，Why the pursuit speed is correct，the images of the synchronization markers are aligned exactly vertically .It's way over my head.

Nothing beats learning yourself with a smartphone handwave.

Check the easier one out:
viewtopic.php?f=23&t=4782

TL;DR Easy 1-2-3 try-yourself method:

1. Load https://www.testufo.com/ghosting
2. Start video-recording on your smartphone at a camera exposure setting of about 4x refresh cycles (e.g. 1/30sec exposure for 120Hz, 1/60sec exposure for 240Hz) <-- CORRECT EXPOSURE IS IMPORTANT
3. Wave your smartphone sideways along the display motion.
4. Observe how the tickmarks aligns or not on the screen of your smartphone

While not as professional as a rail, doing it is yourself is a fantastic way to realize how well it works.

Just like this video:

phpBB [video]