Look here to create a linear ICC profile:
http://fnordware.blogspot.com/2008/05/making-linear-icc-profile.html. With
that set as your working space, you should be able to get correct blending.
If you only have a couple of layers, you could try adding a Levels effect
and set gamma to 0.4545 to
When I try to do this I usually set Photoshop to an sRGB working space,
then in Nuke before each merge I use a ColorSpace node to convert to sRGB,
then I do the Merge, and then I use another ColorSpace node to convert back
to Linear. I'm not in front of Nuke at the moment, but if memory serves
Well in Nuke sRGB is purely a gamma curve. Unless you are really using
the colorspace node to do more then apply the gamma curve.In
Photoshop its a color space always. Also regarless of what icc you
use. Even if its linear the blend modes are still your probably I
beat.
Randy S. Little
Thanks, Julik, Andy and Simon for your suggestions, ideas and questions. After trying a number of things and reading some more about the way ProEXR and Photoshop handles things, we're going to try to make our tweaks in Nuke and let Photoshop do whatever it does. It may take a few rounds of
Hey y'all
I was talking to my colleague, trying to figure out if there is a way
to see when transform nodes, colorgrades etc concatenate in nuke... we
both seem to rememeber it being possible in older versions...
anyone?
cheers
Peter
___
Nuke-users
Colour grades don't concatenate in Nuke 'cos it's all in float.
On Thu, Mar 8, 2012 at 10:11 AM, Peter Hartwig peter.hart...@gmail.comwrote:
Hey y'all
I was talking to my colleague, trying to figure out if there is a way
to see when transform nodes, colorgrades etc concatenate in nuke... we
Hello,
I´m not sure how many of you use Final Cut Pro X but if you do this might be
interesting:
we released an application called ClipExporter which is a simple workflow and
export tool for Final Cut Pro X. It opens exported Final Cut Pro X projects
(fcpxml) and exports each clip as a
Purple colored Transform nodes concatenate together, far as I remember.
Ron Ganbar
email: ron...@gmail.com
tel: +44 (0)7968 007 309 [UK]
+972 (0)54 255 9765 [Israel]
url: http://ronganbar.wordpress.com/
On 8 March 2012 12:14, James Etherington james.ethering...@gmail.comwrote:
Colour
Hi all,
i'm a newbie about Gizmo and Python ..i'm happy to be here...i hope to learn a
lot of things with your help, for become a big master ninja compositor! [Wink]
i'm trying to create a Gizmo for manage my multiEXR render... and i would like
to know if is possibile Hide a Tab if his node is
Very nice, Thomas.
I'm in!
A couple of questions regarding future releases (if you indeed plan to
update this periodically):
1. Are you planning to add speed change support?
2. Are you planning to add more software choices?
3. What about the round trip? Going back to FCPX from Nuke?
Thanks Ron,
Speed changes or retimed clips will not be supported in the near future. It´s
impossible to reconstruct the original (Normal) speed from the data that is
stored in the fcpxml. This is something that really bothers me.
Any suggestions for other software packages? After Effects is on
hey
this should get you started. For this exemple, I just created a
Blur and added a slider to it so that the User tabs knob is
created. Then I seected the node and run this script in the
script editor
s = nuke.selectedNode()
s['knobChanged'].setValue('''
n = nuke.thisNode()
k = nuke.thisKnob()
From what I know (most if not all) transform nodes concatenate as long
their filtering methods are the same. Color correction nodes don't, since
they are already calculated in 32bit linear float.
On Thu, Mar 8, 2012 at 7:36 AM, Ron Ganbar ron...@gmail.com wrote:
Purple colored Transform nodes
it would just be nice to have a visual indication that it's all
good... but thanks guys
On Thu, Mar 8, 2012 at 3:46 PM, Diogo Girondi diogogiro...@gmail.com wrote:
From what I know (most if not all) transform nodes concatenate as long
their filtering methods are the same. Color correction
Simon,Are you saying that the .4545 levels effect on the layer masks and the overall 2.2 gamma adjustment should be done before the 32 to 8 bit depth conversion or after...?Thanks,RichRich Bobo
Senior VFX Compositor
Email: richb...@mac.com
Mobile: 248.840.2665
Web: http://richbobo.comOn Mar
After, as it will make the layers blend in linear light instead of sRGB.
The overall gamma will be the same as the sRGB viewer lut you have in
Nuke. Although, you might introduce problems with banding as your really
bending the colors of 8-bit. Of course you will also clamp values above 1.
It
And the levels should be applied to the actual layers, not on masks. Unless
I'm misunderstanding you.
2012/3/8 Simon Björk si...@stillerstudios.se
After, as it will make the layers blend in linear light instead of sRGB.
The overall gamma will be the same as the sRGB viewer lut you have in
some notes to all the good info already suggested:
- if you do your gamma correction in photoshop, use the gamma in
exposure rather then levels. last time i checked, the one in
levels was not a mathematical gamma correction but has some
adobe visually pleasing stuff added.
- also, do your
Hiya,
Transform nodes in Nuke that concatenate include:
• Transform
• TransformMasked
• CameraShake
• Reformat
• CornerPin
• Tracker
• Stabilize
• Reconcile3D
• Card3D
• Transform3D (hidden under Other / T)
• MotionBlur2D (will receive a concatenated Transform, but doesn't pass it on)
However
And why a merge node block the concatenation like that :
set cut_paste_input [stack 0]
version 6.3 v4
CheckerBoard2 {
inputs 0
name CheckerBoard1
selected true
xpos -525
ypos -17
}
Transform {
scale 0.1
center {960 540}
name Transform1
selected true
xpos -525
ypos 77
}
Constant {
TransformMasked doesn't concatenate
Also the last node in line determines the filter type - not sure which how
motion blur is determined though.
The visual clue is the image is too soft ;)
Howard
From: mattdleonard nuke-users-re...@thefoundry.co.uk
To:
Simon,I made a linear ICC working space profile and I was able to successfully convert my 32 bit file to 8 bit with very little shifting of values. Yea! So, that was a "win" - thank you! The one wrinkle that I hadn't taken into account is that when I save out the 8 bit layered TIFF file from
22 matches
Mail list logo
