STH vs LLO OLPF Black Shading

[Jamon Lewis]

I should have a separate black shade set for each OLPF, one set for STH & one for LLO, no? I can't post an .R3D right now but I went from the STH to the LLO and didn't have time to black shade and got a lot more noise than expected, hopefully I can blame it on the fact that I used the black shade for the STH OLPF.

 

[Phil Holland]

You do not need to Black Shade when switching between OLPFs.

 

[Jamon Lewis]

Thanks Phil, I wonder what the culprit is to the noise i'm seeing. Here's a .R3D of what I'm getting.


https://www.dropbox.com/s/v09wmy97jm9quy5/A001_C006_0620D8.RDC.zip?dl=0


LLO OLPF, 800 ISO, Sigma 18-35 @ f1.8, 5K ws, 23.98fps 1/48 shutter, my light meter is giving me a reading of f2.0 at these settings, just black shaded firmware 6.0.20 (just came off the release build with the same or worse results). Now I'm one who likes a little texture so it's no bog deal, i've just heard people getting "clean" results at 2000 ISO, I won't doubt it, I would just like to know how.

 

[Adam Montville]

I looked at it full debayer, and it looked very clean to me tbh. 2000 ISO isn't " glass clean" but more or less what you would get from ISO 800 from the MX sensor. This looks like normal 800 ISO LLO OLPF. The only hint of noise was in the blacks, and not on the subject, and even that, the noise is very fine.

 

[Phil Holland]

Looks about right for an ISO 800 REDCODE 8:1 image to me with that exposure.

 

[Jamon Lewis]

Thanks for checking it. Yeah, I looked at it at full debater and it is pretty clean, it's definitely not bad, but to the point, it looked (only) a bit cleaner than my MX at 800. Here's one that is more to the point of my post.

https://www.dropbox.com/s/k0a6sclnj4ia2ua/A001_C009_0617EH.RDC.zip?dl=0

LLO 5K 23.98fps 1/48 shutter Sigma 50 1.4 @1.4 1280 ISO (i bumped the R3D up to 2000 ISO). Granted I was in a theater with black walls, I didn't get a reading but to the eyes the walls weren't clipping.

 

[luigivaltulini]

Thanks for checking it. Yeah, I looked at it at full debater and it is pretty clean, it's definitely not bad, but to the point, it looked (only) a bit cleaner than my MX at 800. Here's one that is more to the point of my post.

https://www.dropbox.com/s/k0a6sclnj4ia2ua/A001_C009_0617EH.RDC.zip?dl=0

LLO 5K 23.98fps 1/48 shutter Sigma 50 1.4 @1.4 1280 ISO (i bumped the R3D up to 2000 ISO). Granted I was in a theater with black walls, I didn't get a reading but to the eyes the walls weren't clipping.

push to ISO 3000, quick pass in DaVinci.
you can extract more , no problem . .. Low light is much better than MX, especially under tungsten light.

[/URL][/IMG]

 

[Jamon Lewis]

That definitely cleaned it up, how much noise reduction did you use? Thanks for messing with it.

 

[luigivaltulini]

I used TNR in Davinci about 80%, 2 frame .
Clean Dragon is much easier to MX.

 

[Jamon Lewis]

Thanks luigivaltulini! I'd love to see a few of the cleaned frames in motion if you get a chance.

 

[Jamon Lewis]

Looks about right for an ISO 800 REDCODE 8:1 image to me with that exposure.

I find it hard to understand why there would be noise in areas that aren't under-exposed especially with the LLO. It's kinda strange to me. Ironically, I like it, it's an ascetic I love use for my personal stuff or paid work that I'd have creative liberty with but like I did with my MX ( I always felt properly exposed MX was to CLEAN for me) I would like to HAVE to push the sensor to get these type of results but other wise and "normally" have it smooth and clean.