BM pocket camera test charts

The Blackmagic Pocket Compact Camera is a nice new kid on the block. After shooting and testing with it for a few days and reading a lot of other peoples articles I decided there was room for some technical tests. So don't expect pretty pictures, this is for Cinematographers and Graders that want to use the material for their own test.
I will put my remarks between the lines too.

 

Used in this test a BMPCC with the 1.4 firmware (will update today and see if there is any difference)
Lens was a Panasonic 12-35mm F2.8 MF lens.

Shot in daylight on a cloudy day, as you can see my meter gave it a 4500K reading so the camera was set there too. Looking at the scopes that was pretty close.

 ColormeterLigthmeter

I measured with a spot and flatsphere meter but my two greycards give slightly different results and so do my sportmeter and incandence meter. Normally I use my Transvideo 15"UB for waveform monitoring but I am still waiting for those micro HDMI cables. 

Looking in Resolve on the waveform, my choice seems pretty accurate.

waveformwaveform.

 

What did I shoot:

Standard:

Clip 1 - 800 ASA F-stop set to (external) meter at F5.6 - FILM / LOG mode.

Clip 2 - 800 ASA Same as above in REC709 / VIDEO mode to compare LUT's against.

 

Sensitivity and grain (from here on everything is LOG /FILM mode)

Clip 3 - 1600 ASA so Fstop was at F8

Clip 4 - 200 ASA

Clip 5 - 800 ASA 1 stop under exposure

Clip 6 - 800 ASA 2 stop under exposure

Clip 7 - 800 ASA 3 stop under exposure

Clip 8 - 800 ASA 1 stop over exposure

Clip 9 - 800 ASA 2 stop over exposure

Clip 10 - 800 ASA normal exposure on a Kodak greycard

Clip 11 - This happened when I pressed the IRIS button to let the camera decide the exposure. It opened all the way up to F2.8. - Putting a white sheet in front of it and pressing the IRIS button gave another bad result and it overshot its setting in the other directon.

Reading through the docs I understood that the camera helps exposing 'to the right'. It should look at the brightest point and expose for that.

None of the choice made any sense to me so I tried the same on REC709 and that gave a much better reading around what I expected. Funny enough I had not a single problem shooting real life scenes using the build in meter.

 

Next a quick keying test:

Clip 13 - 800 ASA setting on LOG / FILM 

Clip 14 - 800 ASA setting on REC709 / VIDEO 

Clip 15 - Shot with a Canon 7D at 800ASA for comparison. 

 

My test was in Resolve and Adobe Premiere CC. In the last one I dropped the Ultra keyer on the timeline and clicked only once at the same place. The results are very nice already. The better sharpness gives way less alliasing.

Here is an example in H264 / 720P which does not do any justice to the material. So please download the material yourself and test, test.

 

You must have the Adobe Flash Player installed to view this player.

 

All clipnames are download links too. I have only put short original shots online for testing.

 

Postworkflow

Working in Resolve I found all LUT's pretty bad. Crazy enough one of my Alexa LUTs seems to do a much better job than the build in BM one.

We are creating some new ones based on this material and will put them online.

More to come....

Please leave your comments here below en share your own experiences en results.

Enjoy.

 

Bastiaan Houtkooper NSC