Blackmagic Forum

I'm looking for advice on selecting the right camera and settings for a work-related project. Our project involves capturing photos or videos of power line poles from a moving truck traveling at speeds of up to 90 km/h (~56 mph). The primary goal is to ensure that the captured images and footage are detailed enough to clearly read small tags and text on the poles.

Key Questions:

Motion Blur: What camera settings are crucial for eliminating motion blur in both video footage taken at this speed? Specifically:

What frame rate and shutter angle range should I aim for to freeze motion effectively without underexposing the frame?

Are there any additional settings or techniques that can help with clarity during motion capture?

Resolution and Detail: To capture fine details like small text and tags, what specs should I prioritize?

Is higher resolution (e.g., 12K) essential, or can a lower resolution like 6K or 8K suffice while maintaining readability?

How does sensor quality and pixel size factor into capturing minute details?

Dynamic Range and Lighting: Power line poles are often positioned in areas with uneven lighting (shadows, bright skies, etc.).

How important is dynamic range in this scenario, and what settings or techniques would you recommend to optimize image quality in variable lighting conditions?

Compression and File Management: Given the large volume of footage:

What Blackmagic RAW settings (or alternative formats) would you suggest to maintain image quality while managing storage efficiently?

Cameras available:

I have been looking at the BlackMagic URSA Mini Pro 12k. Would this camera be suitable for my purpose?

We need to ensure that the images and videos are sharp and clear enough to identify small labels and details on the poles while the vehicle is in motion. Any guidance on recommended camera models, specs, and practical setup tips would be greatly appreciated.

Thanks in advance for your insights!

You may need something faster, but 12K wouldn’t matter too much, since it already outresolves most lenses. What you are trying is really getting close to the technical limits. The most important points are shutter speed, and depth of field vs image size, which are mutually limiting.
IIRC, the 12K only goes to 1/800 of a second, while the Pockets can reach 1/5000.
I’ll get back with more thoughts later.

Detect_Colin_S wrote:Key Questions:

Motion Blur: What camera settings are crucial for eliminating motion blur in both video footage taken at this speed? Specifically:

What frame rate and shutter angle range should I aim for to freeze motion effectively without underexposing the frame?

That depends on the light level and the amount of noise you can still accept for readability. IIRC, the 12K can go to 1/800 of a second, while the Pocket models can do 1/5000.
I hope some owners chime in, since my tests date back quite a bit, and as a filmmaker you are normally not that interested in very short exposure times.

Detect_Colin_S wrote:Are there any additional settings or techniques that can help with clarity during motion capture?

Resolution and Detail: To capture fine details like small text and tags, what specs should I prioritize?

Shutter speed is the central question. Since most cameras can use a smaller window to get faster readout, you may need to get a balance between resolution for readability and speed to avoid motion blur.

Detect_Colin_S wrote:Is higher resolution (e.g., 12K) essential, or can a lower resolution like 6K or 8K suffice while maintaining readability?

You might be better off with such cameras, if they offer shorter exposure times (see above).
Plus, 12K vs 6K will not show that much finer detail with most lenses.

Detect_Colin_S wrote:How does sensor quality and pixel size factor into capturing minute details?

Again, mutually limiting factors: smaller photocells capture less light, which matters for those very short exposure times needed, but capture more detail.

Detect_Colin_S wrote:Dynamic Range and Lighting: Power line poles are often positioned in areas with uneven lighting (shadows, bright skies, etc.).

I doubt that would be a huge issue with modern cameras, since most of them can handle the contrast of daylight exterior pretty well, as long as as you avoid backlight and rather wait for an overcast sky.

Detect_Colin_S wrote:How important is dynamic range in this scenario, and what settings or techniques would you recommend to optimize image quality in variable lighting conditions?

While one would normally avoid that, in your case a camera with automatic exposure should be the better solution, for example a fast hybrid from Sony, Canon etc.

Detect_Colin_S wrote:Compression and File Management: Given the large volume of footage:

What Blackmagic RAW settings (or alternative formats) would you suggest to maintain image quality while managing storage efficiently?

Another point to balance: less compression would improve readability, but yield bigger files. That's true for all cameras, but those with GOP compression may have more issues with your fast driving.

Detect_Colin_S wrote:Cameras available:

I have been looking at the BlackMagic URSA Mini Pro 12k. Would this camera be suitable for my purpose?

Probably not, see above.
Another point to balance is focal length, speed, and depth of field. Not knowing how far or close these poles would be and which size are those text plates, it’s near impossible to advise. Better speed of the lens is good for short exposure, but more difficult to keep in focus.
Another point to consider: modern hybrids have excellent low-light capabilities and auto-exposure. They may show more rolling shutter, but for such linear motion like in your case, it should be easy to treat with specialised filters in post.

If nobody else with experience in a similar field chimes in, I think you should hire a camera and simply try, following the rules above.

If the BMCC 6K really does 1/5000, it's about as fast as a Sony A7IV. Its resolution is better, and the BRAW codec at low compression ratios too. It should have enough dynamic range. The larger file sizes should be the least concern, as long as you can read what you need. SSDs are not that expensive any more.

But your project is definitely not an easy task with that many conflicting parameters!

I’ll attach a quick test with an iPhone at 240 fps in HD. You can see that this is clearly not enough, even if we were only driving at about 30 km/h.

Like Uli has mentioned. However, also:

1) If the camera can be panned in sync as passing each pole, this could help. This may not be practical for large numbers of poles, but for a couple... maybe. (of course this may blur the background)

2) I believe the URSA Mini Pro 12K shutter will also go to 1/5000. At least it can be set to that... and I have shot wildlife with shutter speeds of 1/120, 1/240, 1/480, 1/500, 1/1000, 1/2000, 1/4000, 1/625, 1/1250, 1/2500, 1/5000 changing the aperture or ND appropriately. The resulting exposures track with the settings. My initial calibrations were done with the URSA Mini Pro (G1)... but I just checked and have many clips with the 12K at 1/2000 and 1/4000.

3) For such a specialized use, if would be important to test and practice... especially to figure your exposure settings.

4) longer lens for the tag size, but longer lens will show more motion blur than wide

5) With camera travel speeds the closer an object is the more it will be motion blurred.

6) There are cameras made specifically for high speed work. To get high speed and high resolution and high dynamic range might require a custom designed camera.

Thanks for chiming in on the shutter speeds, Jeffrey.

Jeffrey D Mathias wrote:4) longer lens for the tag size, but longer lens will show more motion blur than wide.

Sorry, but at the same distance the same size of a tag will show the same motion blur. OTOH a longer lens may be slower.
This is another test I did in traffic with the Sony A7IV with 1/2500 and a 50mm Rokkor PG, one of sharpest lenses ever made. Rolling shutter is not that bad, I think tags that are in focus would still be readable if detail suffices. It was already getting dark with a covered sky, around noon one could even close the lens down a bit, since at f1.4 the DoF is damn narrow. I don't know how well you can keep the distance constant. This is a smaller area from a UHD frame (no scaling). With some sharpening even the smallest text in the lower right should be decipherable (scroll):

Detect_Colin_S wrote:I
Dynamic Range and Lighting: Power line poles are often positioned in areas with uneven lighting (shadows, bright skies, etc.).

How important is dynamic range in this scenario, and what settings or techniques would you recommend to optimize image quality in variable lighting conditions?

That's really tricky.

If you're able to take photos manually (eg. somebody points a photo camera at each pole and snaps a photo), then the camera can adjust to the lighting fairly consistently.

Alternatively you can have multiple cameras recording in parallel with high/low light settings.

I think what's really important is:
How far away are the poles and the size of the tag?
Do they change distance during driving, or do they all pass at the same distance in parallel?

I think for your purpose you really need to test.

Also the way you set up the camera will greatly affect the outcome.

I think what you really want is the fastest possible shutter speed for reducing motion blur.

Higher frame rates don’t really help as much in this situation as much as shutter speed.

I think there are two steps you can take though that will help independent of the camera choice.

The way you MOUNT the camera.
The direction you point relative to travel.

If the camera isn’t locked into the vehicle there will be differential vibration. The camera is moving because of the car but also out of phase with the car adding to the motion blur and reducing the legibility.

You either want a mount that is locked down to three locations, grabbing the camera at the top and bottom to really make sure it only moves WITH the car.

You might also consider testing a gimbal to stabilise the camera which may also help. Sometimes gimbals works better with vibration isolators.

If you’re pointing the camera at 90 degrees from the car then the poles will move faster than if the camera is pointing forward or backwards from the cars direction of travel. Fast travel means the variables are even tighter and more difficult to resolve.



So…

Camera mounting and stability.
Camera direction relative to car.
Lens focal length.
Fast shutter speed (needs more light)
Higher frame rates (needs more light also)
Higher resolution.

These are all variable that matter most. Codec, DR etc too but less relevant.

For reference these kinds of rigs work well with gimbals.

https://customeasy.org/vibration-isolator-media

JB

I would think that you would get the best results with a global shutter camera that supports a very fast shutter; hopefully around 1/4000 sec. You might need to provide your own light source as well given the pole may be backlit; in that case you need a very high refresh rate on the high power digital light (that will exceed your shutter so you’ll always have light on the pole such as 5-10K).

I don't think a global shutter is needed. It'll loose dynamic range and the amount of distortion I can see at 1/2500 will not really destroy readability. The Sony is not great regarding RS.

Uli Plank wrote:I don't think a global shutter is needed. It'll loose dynamic range...

The loss of dynamic range is inconsequential if the shooter’s gear includes their own light source.

Of course. But we need more information about the whole situation.

If the truck is traveling at 90 km/hr, then the sensor is moving at 25 meters/sec or 5mm in 1/5000 second. So the effects of rolling shutter may be evident. As you mentioned, it’s important to know the distance to the poles and the height of the values of the information to be recorded for a given focal length of lens selected. An interesting question for your film class.

At an infinite distance the pole would not appear to move. Let’s pretend that ten kilometers is infinity. Assume the pole is 10 meters away at a minimum. Therefore would the pole would appear to move 0.005mm across the sensor?

That would mean the image would shift 2 or 3 photosites on the URSA Mini Pro 12K. Given that the camera only truly may resolve 4K, the image could look good.

But the focal length of the lens also plays a part in the movement of the pole. I’m not sure how to factor that into this scenario. A wider lens like 10mm will minimize the apparent movement, but it may be necessary to use a 100mm focal length to have the information on the tag readable.