Can the Ursa Mini Pro create HDR Video?

Hi I’be been doing some research into Creating HDR video and I’ve ran into more questions and answers.


Can the Footage shot from a Ursa Mini Pro be turned into HDR video?

Also I know I would need a HDR tv or monitor in order to grade and edit correctly.

I currently have a Shogun Inferno as well as a Eizo ColorEdge CG248-4K 23.8" monitor. Would any of these monitors be ok for HDR editing? I know the Inferno has a HDR function and the Enzo has a Rec2020 color space option.

I also have the studio version of Resolve.

Any modern camera can be used for HDR if it delivers RAW or a flat image in a good codec. All of them capture considerably more DR than Rec709 can present to rhe audience.

Regarding monitors, I can only speak for the Inferno. Yes, it can deliver pretty nice HDR and can even be calibrated.

I'm more curious about exactly what DR characteristics are need for HDR? You can shove 20 stops into a 709 signal if you want to, but HDR formats are about how it is displayed on screen. In a SDR screen those 20 stops for in the 100nit plus range. On a 1000 nit screen you may notice banding, so they use 10 bits, but you get a better appreciation of an out door scene as it was (plus grading). However, there must be some DR requirement otherwise without stretching. There will be a lot of burned out highlights. Maybe somebody here knows?

Setting a reference monitor that matches one (or more) of the HDR specs for distribution, is not (yet) a cheap or simple process. Some sources of information that I've found useful in learning about HDR have been this blog post by Alexis Van Hurkman http://vanhurkman.com/wordpress/?p=3548, reading the HDR section of Resolve manual, and also the articles covering HDR on MixingLight.com https://mixinglight.com/tutorial-category/hdr/

How you squeeze 14 or 15 stops into Rec709 or even HDR is an aesthetic decision, not purely technical. That's what colorists do when taking a flat source and re-designing it for Rec709, P3 or HDR (whatever standard of it).

Plus, you have to understand that 100 nit for TV is pretty outdated, modern ones do about 300-400. A HDR screen with 1.600 nit would only have two more stops to display and 2.000 a quarter step more.

Uli. I mentioned a 100 nit, as a standard, because if I said 300-400 some little nuisance was likely to come by and correct me saying that 100 was the standard. But I'm having a hard time trying to get how your post relates to my post again?

10 bits reduces banding, but Dolby Digital testing determined 12 bits was more appropriate. However, its a simple information availability question, as to his much DR from the camera is needed to produce good results on a HDR display without looking unnatural.

As Alexis states in the link in my post above, any camera that can record around 13-15 stops has enough dynamic range for HDR. Since the Ursa Mini 4.6K and Ursa Mini Pro sensors have that dynamic range and can record in 12bit color, they should work well as a source for HDR.

As to whether HDR looks natural or unnatural, that's a subjective judgement up to the eye of the viewer. I've yet to see an HDR image (including some Dolby Vision theater screenings) that didn't appear unnaturally oversharpened to my film loving eyes.

He, he, not my meaning. You want quality, you can match a natural characteristic more closely (16.5 stop intra frame eyesight). Trying to put too few stops into a delivery is going to stretch it out or require early clipping etc.

So, the minimum is 13 stops then? (Official spec?) Sort of makes sense will all the lower quality the industry is used to doing.

Thanks for the links Jamie.

It's correct that SDR calibration is still standardized to 100 nit, but most consumers are looking at 300-400 nit TVs today. What I really tried to explain – based on the initial question – is what we can show to an audience vs. the capability of modern cameras.

If all we can reproduce in the home is about 2 1/4 stops more vs. the 6-7 stops of conventional TV (in a very dark living room), any camera with more than 10 stops should be OK without grading. To have some room for stylistic grading, we'll want 12 or 13, and the UMP46 is definitely capable of that much.

To put it into another perspective: if you look straight into the sun in bright daylight for a while, you'll go blind!
Even a very expensive Dolby Vision HDR device will not damage your eyes when showing an image of the sun…

Ursa mini 4.6 is approved for HDR content creation on Netflix.

And Resolve already supports Rec.2020 12bit using BMD CinemaDNG.


Sent from my iPhone using Tapatalk

Uli. In respect of our guesses, I'd rather hear what the definite science of top end formats worked out as minimums for different levels and bit depth. I can see potential for troubles.

Now, maybe you didn't pick up I said I know that TV sets are that bright, but the spec is the other, so I mentioned the standard to avoid being hassled for not mentioning the safest option.

Rick. I'm interested in what coverage the camera has if rec2020. Any charts of it out there?

None I’ve seen Wayne.

Since I don’t have a monitor that displays Rec.2020, I haven’t given it much thought. I recently used it on my 10bit P3 monitor, but I’m sure that’s only an approximation.


Sent from my iPhone using Tapatalk

It's worth mentioning that there is as of yet no commercially available display that can produce 100% of REC2020.

Even though some camera manufactures have record/output modes that fit within the REC2020 spec, I've not seen camera manufacturers state exactly what part of any particular gamut that the sensors capture. This article from cinematographer Art Adams sheds some light on why that might be:

https://www.provideocoalition.com/camera-may-color-gamut-musings-camera-color-pipelines/

Let me guess, rec2020 is an aspirational standard for manufacturers to strive towards, and a processing and display format standard to accurately map this progression into? After they finish fleeting us with old DCI tech? Am I close?

Quantum dots can nearly map the whole space, to a point the consumers can barely tell a difference (manufacturers were proposing to reduce the rec2020 space because of this). Most of the greater rec2020 coverage sets I have heard about, mysteriously didn't turn up with that coverage. I think, can't remember the brand, they have them in America, turned up with between 90-93% I think. But this is far down from the over 97% that QD can do. Next year I hope the WF color filters will finally stop being held back and come out. Those should get close to the full range QF can do. There was a problem using them on LCD that required LCD's to be redesigned, but that was LCD, not OLED. So what has stopped high not HDR oled screens with these filters? I don't know why, but the oled had to be a blue range for the dots. But since the announcement these WD color filters were to be released by BASF (I think) maybe 3-4 years ago I have been eagerly awaiting them. I take colour seriously.

This might be helpful too:
https://digitalfilms.wordpress.com/2018 ... ed-part-1/