Forum Replies Created
-
AuthorPosts
-
guillem ramisaParticipant
Hello filmsoup and thank you for your reply. You are correct when you assume that i am a tablet guy and used to flames direct way of dragging values. It sure is smooth and hard to switch to other ways of working. Nuke is powerful but sometimes you can sense it somewhat technical background/history. But it is getting better though.
I already have found the way of alt-dragging values, thats the way i’ve been working but its not perfect as you describe it. I mean its backwards and slow that you have to adjust in such small values to get a usable result. I will follow your advice and explore the grade node some more but i still miss something more telecine like (color warper or similar please). It would speed things up. You can feel that the software was built for power and long term projects
and not quicker client attended sessions, well now the purpose for nuke is probably not and never will be in advertising and such but i do wish for and hope that thefoundry will not only develop and fill nuke with complicated and cool new “wow” technology but also remember to improve and and come up with new creative ways when it comes to interface design and workflow. It’s those small repetetive everday tasks that you do again and again that counts. Not if you have a fancy 3D tracker included. Well it sure is nice to have one but i would prefer to have a super robust 2d tracker instead of any 3D tracker. Of course its nice with technical progression but dont forget to improve the things that already are there. They are not perfect. I’m looking forward to the new paint and roto 🙂guillem ramisaParticipantHello,
Nuke is not my first choice for doing editing stuff, it can get pretty messy fast. But it is possible though using the various nodes under the “Time” menu. Here is one way to do it but you could probably do it in several other ways as well. Cut and paste the script below:set cut_paste_input [stack 0]
version 5.1500
Constant {
inputs 0
channels rgb
color {0 0.67 1 0}
name Constant2
selected true
xpos -206
ypos -200
}
Text {
message “Clip B”
Font 1
font C:/WINDOWS/Fonts/arial.ttf
size 32.5
translate {406 522}
name Text3
selected true
xpos -206
ypos -126
}
Text {
font C:/WINDOWS/Fonts/arial.ttf
translate {480 286}
name Text4
selected true
xpos -206
ypos -99
}
Retime {
input.first_lock true
input.last_lock true
output.first 30
output.first_lock true
output.last 129
name Retime3
selected true
xpos -206
ypos -72
}
Bezier {
output alpha
invert true
points {{
{359.8 579}
{360.8 -4}
{-2.5 -7.5 0 2.37}
{-2.5 577.5 0 0.78}
}}
shape {}
center {360 288}
name Bezier1
selected true
xpos -206
ypos -45
}
Premult {
name Premult1
selected true
xpos -206
ypos -6
}
Dot {
name Dot1
selected true
xpos -172
ypos 51
}
Constant {
inputs 0
channels rgb
color {1 0 0 0}
name Constant1
selected true
xpos -33
ypos -200
}
Text {
message “Clip A”
Font 1
font C:/WINDOWS/Fonts/arial.ttf
size 32.5
translate {16 522}
name Text2
selected true
xpos -33
ypos -126
}
Text {
font C:/WINDOWS/Fonts/arial.ttf
translate {93 286}
name Text1
selected true
xpos -33
ypos -99
}
set N5b62348 [stack 0]
Retime {
input.first 30
input.first_lock true
input.last_lock true
output.first 50
output.first_lock true
output.last 120
output.last_lock true
name Retime2
selected true
xpos 7
ypos -62
}
push $N5b62348
Retime {
input.first_lock true
input.last 30
input.last_lock true
output.last 30
name Retime1
selected true
xpos -90
ypos -63
}
Switch {
inputs 2
which {{curve i x30 0 1}}
name Switch1
selected true
xpos -40
ypos -23
}
Merge2 {
inputs 2
mix {{curve i x29 0 1}}
name Merge1
selected true
xpos -40
ypos 47
}
Viewer {
name Viewer1
selected true
xpos -40
ypos 74
}guillem ramisaParticipant@N8 24849 wrote:
I just installed Toxik today and boy is it slow. As soon as I add a reaction node it goes to pot. Any thoughts? I added a new SATA drive and installed the software on there. That is also where my mediacache is. I also moved the footage I was working with to the same drive. I upped my VM as well to like 4 gigs. I have a P4 2.8ghz with 4 gigs of ram, and I am using a Quadro FX 3450. Any thoughts?
Sounds weird. I’ve tested it on my latptop (asus a8js) and it was pretty responsive for that kind of setup. One thing I can say though is that if you really want the best performance possible you should get a raid0 and point the mediachache to that raid. The way Toxik works is that it will automatically cache all the media that you link to in your project and that cache should be on the fastest drive possible.
The Quadro card should be fine but the reaction thing might sound like a graphics card issue. You should check that the driver version you use is supported. I’ve also read (probably not your particular problem but anyways) that to get the smoothest playback the screen frequency should be a multiply of the fps setting in your project.guillem ramisaParticipant@VOODOO 24775 wrote:
I’ve tried the deform node and can not achieve the look I want.
If you have aftereffects cs3 you might have a go at the puppet-tool. Not sure if it will work though it might look a bit squeezed if you need to bend it that far.
Or maybe it is time to collect that favor from your best friend the 3d-dude. He could probably fix your little problem in a minute or so.guillem ramisaParticipant@Alfafa 24768 wrote:
Hi everybody,
(I’m using Flame 2007 on Linux)
Does anyone knows how the camera projection works ?
I had to remove some wires on a shot with a huge nodal pan, so a lot a parallax…
I 3D tracked it fine, then the idea was a to clean a frame and project it in scene, but I could NOT get it to work at all.
If someone can explain me the use of that Projector thing that would be terrific and make my day !
Thank you youyou.
Alfafa
In action add a (white) layer that you want to project your image on. It could be a simple plane or an extended bicubic etc. But it can also be an imported 3d object (.obj format).
Add another layer with the image that you want to project on the plane/object. Highlight that layer and choose add projector. I you have a 3d camera track you should copy the camera parameters from the first frame (or the frame that you painted) but without the animation. If your object is in the right place in z-space it should stick, check where the the nodes from the 3d-track are positioned. If it is a nodal pan, they should be “sphere-like”.guillem ramisaParticipant– Why not try to team up with carpenters and miniature model makers? Would be nice if your studio could to be able to offer that kind of service.
– Service of renting and/or giving advice on the right type of camera for your effects material.
– Maybe on-set service with an operator+desktop where you can previz your effects with a live feed from the camera.
– Motion control rig. Expensive yes, but nice.
– Motion capture equipment
cheers,
kJanuary 20, 2008 at 8:17 pm in reply to: newb question: Where can I put the field interpretation? #216496guillem ramisaParticipant@filmsoup 24640 wrote:
Yes, you can remove the fields in nuke, Under the “Time” menu “remove 3:2 pulldown”
That would remove pulldown but not fields or am I missing something?
January 19, 2008 at 12:54 pm in reply to: newb question: Where can I put the field interpretation? #216495guillem ramisaParticipant@ragornette 24637 wrote:
Hello all
I have a footage in field (upper first)
where should I put the fact it is upper first or lower first in nuke
I suppose it’s in READ but can’t find anything about and didn’t find
in documentation. Am I blind?Thx:rolleyes:
As Nuke is mainly a film compositor I don’t think there are any tools for dealing with fielded footage. You should remove your fields in another application before entering Nuke.
Cheers,
K.guillem ramisaParticipant@rueter 24612 wrote:
You can output vectors for objects in the Render node but I’m not sure if that works properly yet.
Your best bet is using multi sample motion blue (either in the Render node or a subsequent TimeBlur node)Yeah, the vector-output from the rendernode doesn’t seem to work properly. And motion sample blur isn’t optimal either. I guess thefoundry has something more to fix…
guillem ramisaParticipantYou need a motionblur2d node after your transform, it will generate uv channels that the vector blur can use. So you’ll put a vector blur after the motionblur2d and set it to use motion uv channels.
guillem ramisaParticipant@CheGuevarra 24607 wrote:
Thanks so much! Works like a charm. I should possibly brush up on my photoshop skills now that I learned you can do this, great.
You could do it in shake as well but it wouldn’t be that responsive thought. If you paint with white on a black frame and pipe that to a switchmatte you’ll get your alpha that way. If you wanted to see a red overlay it would just be a matter of adding a color correct and a layer node and watch that output while painting. But as I said it wouldn’t be that responsive so if it’s just one frame I’d definitely do it in photoshop.
K.
guillem ramisaParticipant@CheGuevarra 24603 wrote:
Okay, so I have been using the paint module in Flame and it’s great. I can paint onto a frame and make a mask by using the black frame as the back, and painting using the “Matte Button”, it turns my stokes pink making it stand out as well.
I don’t have much access to a flame, is there another software which will also allow me to paint using a “Matte Button” as in the flame module. Say I want to draw a matte out of some grass, in Flame I simply enter the paint module, and use the “Matte Button” and trace the grass as desired, then export the matte.
Can any other software allow me to do this? I tried in Shake quickpaint node and had no luck, maybe I am doing something wrong. I know in Photoshop this is impossible right. I dont want to use combustion at all, so If someone can recommend a tool or plugin that would be great, thanks.
This is perfectly possible to do in photoshop in a number of ways. I’d use a layer mask. Just paint on the layer mask with a black or white brush and you’ll get your alpha. If you don’t want to see the actual transparency and just the pink overlay, disable the layermask (right click on the mask layer and disable) and go to your channels and click the eye for the alpha layer.
K.
guillem ramisaParticipantAfter Effects supports import of maya project files – with some restrictions – read more about that here:
http://livedocs.adobe.com/en_US/AfterEffects/8.0/help.html?content=WS3878526689cb91655866c1103906c6dea-7f25.htmlI know that boujou 4 has maya scene export that after effects “likes”. I don’t know if pftrack has that kind of support.
K.
guillem ramisaParticipant@Imhotep397 24541 wrote:
If for whatever reason you’re still stuck on Fusion you’d be better off using VMWare Fusion under OSX and you should be able to run Eyeon Fusion in Windows inside of a Window, 64-bit if necessary, with no performance hit and when you get the blue screen of death or everything freezes you can just close the window and continue working on OS X.
Running Fusion using VMWare… you’ve got to be kidding.
– – –
And for answering jimmybee’s original question, why not use Shake? I mean it’s the most commonly used compositor in the film business and I don’t think there is anything “inferior” about it’s toolset. You’re using OSX and Shake is also extremely cheap.
Maybe it’s getting a bit old and don’t have all the latest technology under the hood but it will do it’s task and do it good. I love the workflow and feeling of Shake. I guess there is a reason people are still using the 2.5 verion on PC. And as mentioned above I wouldn’t sit around waiting for anything new from Apple, at least not software targeted at any serious compositing.
If you wanted to try the latest I’d bet my money on Nuke, I like it a lot and it’s getting better and better for each version. Latest 4.7v5 is really stable and I’m looking forward to the Foundry’s rework of the user interface that will arrive soon. It’s on OSX too. That way you don’t have to dual boot.
The Foundry is also really innovative as a company and I hope and believe that a lot of their technology will be incorporated within Nuke. In addition to that they seem to really listen to us users… a thing that is surprisingly uncommon in this business.guillem ramisaParticipantYes, I think it’s pretty weird too. It should be enough with the settings in the read node for Nuke to understand what format and aspect the image has and carry that information downards the node tree to operators that needs this kind of data, like masks and such.
I’m not that familiar with the inner workings of Nuke so I can’t answer why the reformat node is needed for this to happen I just know that following this rule solves the problem. So if you have a feeling that you’ll need to change formats later on in the process make sure to throw in the reformat node or you’ll be in trouble.
But (and I hope you read this theFoundry) it would be great if the read node could do that job saving us the hassle of adding the reformat node to every read node. -
AuthorPosts
