Christopher Adams
Forum Replies Created
-
Yes I would agree that keeping it to the big panel at this point is silly!
CJ -
Can’t say too much. But so far this kat is fast! faster then lion was.
CJ -
One thing you can try is to add a node with higher saturation and key it then after keying reverse the saturation boost. We are using 32bit float here after all.
I find sometimes it locks on to things better. Also. what kind of footage was it? I find that that will totally effect what you can and can’t key well. Another thing you can try is to do a noise reduction prior in the chain from doing your isolations. Or hell if its a bad clip run it though something like MB-denoiser or my favorite Neat video plugin. Then try keying. I suspect that it will help.
Just a few ideas. NOTE DLSR footage is some of the worst! to key on. ITs lossy 8bit H.264 breaks down quickly.
CJ -
Though if this is being used to work the display could this then not be used for cuda?
-
Keep in mindtyat is for version 8 He version 9 one is coming soon. Keep an eye on http://fxphd.com
For the new version. Also my mentor over at taoofcolor.com they will be re-working for version 9 as well. Lots of options out there.
CJ -
fxphd.com will be putting out a Resolve 9 Fast forward class soon I believe. I think Warren is doing it again soon.
The one i took for version 8 was very well done. He knows his stuff.
CJ Adams -
This has been happening to me off and on for some time now. Not sure what causes it.. I take a still store of the grade then delete the nodes for that shot and paste the correction back.
-
Actually. i think what you guys are doing is trying to trick the system and might cause a ton of other issues with other programs.. If your calibrating for a 1.8 gamma curve then the other programs other then FCP will then be incorrect. Far as I know calibrate your monitors as 2.2 period. Lion is now correct for the system wide gamma of 2.2. all programs *modern* will adhere to this. FCP is legacy and this more the likely will not be fixed. Not sure what is up if your getting this in fcpx though. I know the QT players and New 64bit VLC are correct as well as Adobe premiere. This is isolated to FCP at this point. *ie it lies* Do not trust the preview pane. Just go by your scopes *you are looking at those right?* They won’t lie to you about where your white and black points are. If you are crushing stuff they will show you that with a bunched up trace at the bottom. I suspect you should ignore FCP preview window and just move on and judge it in something more modern that doesn’t have this bug. I always look at my stuff on a THX calibrated external reference monitor with HD-SDi. and YES FCP is way way too dark compared to this or to davinci resolve. setting your systems to 1.8 i would think is a mistake. it will only make other things then tcp look wrong. If your a one program guy then maybe it is fine. Otherwise i would say don’t do it.
CJ Adams
colorist
simplexity digital post -
I do wonder if one of the eliete books would would. Those have an option for a Dreamcolor display at high res. and a quadro card and desecrete graphics. I would assume that you would need to swap to the internal crappy graphics to free up the quadro from being connected actively to the display for resolve to work.. anyone go that route testing this yet?
CJ -
I would tend to agree.. Maybe a setup&configure group would be useful. I would love to get back to discussions on theory and craft. We don’t seem to talk about that very much anymore.
Sadly!CJ