V1 compared to V2 tests

[paulcurtis]

I've managed to grab some time and to do some tests. I'm working through the footage and will upload examples as i compare.

I set up a scene, with a macbeth, then went through and did various combinations. I had one set PPOFF and one with a typical PP

Cinetone1, Blackgamma High/-7, Knee 100%/-1, Cinetone 2/level 8, Colour Level -2, Phase -1, No Colour adjustments, WB set to RB but zero'd, Detail 0

I don't know why i put Colour Level -2 and Phase -1, weird. But no matter

I ran a series of exposures at constant 9db and f3.5 and changed the shutter speed. I did 3200 with a lamp, and 5600K with an LED source (Bi colour mixed around 56k), i also white balanced the 3200 tungsten source at 5600 (set on camera) because i reckon a lot of people warm up lights like that and i think it's a good way to create crappy highlights

I also did some cloud stuff but as they're moving it's a more difficult thing to compare. I haven't worked through that yet.

One version has the LED light moved in further and the other doesn't. It means that it's reflecting a little bounced light into one version but not the other
The frame is slightly different between v1 and v2 as i knocked the camera (doh) and zoom.

Right the good news: v2 menus seem a little faster and i like the iso read out

The bad news? I cannot tell any difference whatsoever between the v1 and v2 colour or highlight handling yet. Small variations can be put down to testing environment but frankly it looks like same to me. And i will also hold up my hand and say i never really had the kinds of issues that folks here were seeing either.



The two frames above show WB3200 with tungsten lamp, at 60th second, with the PP above. I don't find anything offensive in either version. Do you have a preference, i won't even say what is what. If i examine the luminance over the RGB channels they roll off pretty smoothly - no issues there. There is a perceptual magenta band but i have a feeling that's more my macbook pro LCD limit than the source.

I will collect and post more side by sides as i get to them.

If anyone has any questions or wants to see something in particular fire away.

cheers
Paul

 

[paulcurtis]

Here's a typical yellow band, which is 3200K and no PP, again no difference between the two firmware versions that i can see.

The blue channel blows out last and then quickly, compared to the RG, so the absence of blue creates the yellow halo



The image below shows a strip of clipping with the RG and B channels underneath and their levels condensed to compare more easily



You can see the blue channel having data right up until it blows, whereas the other two soft clip at the same time. This, i think, is something with the knee and is probably a separate conversation from v1 compared to v2. The wall itself is off-white and the WB of the camera is 3200 which, on the one with the PP, seems to match better. The PP Off looks warmer. So with no PP the camera white balance seems different.

cheers
Paul

 

[stevedocmaker]

Can you do the same test with skin and a red object in the scene?

 

[evilthought]

Can you do the same test with skin and a red object in the scene?

I doubt it

He already upgraded to FW2

 

[Michael Solomon]

I can do one next week if you'd guys like. Red/orange wall paper + a person + white card

 

[paulcurtis]

Can you do the same test with skin and a red object in the scene?

As pointed out i've done the upgrade now, there's no going back.

I will dig out some more comparisons today, the macbeth chart helps check the colour differences but so far the hue remains the same from version to version, so colour wise it seems pretty much identical.

I've yet to find any differences in the highlight or gamma handling. Of course i just compared one profile and no profile, that's not to say that sony have made changes to cinegamma 2 for example. But in light of all the other evidence and sony themselves i really don't think anything has changed with the curves

cheers
Paul