Foundry Buys Nuke
Saturday, March 10, 2007 at 10:32PM The Foundry has acquired Nuke.
Update: The fxguide article now features an interview with The Foundry's Bill Collis.
Nuke is impressive to say the least, but it's a bit pasty from being behind closed doors for so long. Maybe a handsome Brit to escort it to its coming out party is just what it needs?
The world of compositing software is confusedly annoying right now. Shake has voluntarily succeeded the throne, only to watch Fusion stumble and fall on its face in an attempt to take the seat. After Effects, while still the best place to be creative with images, added 32-bit support to an aging architecture, effectively putting gold rims on the hoopty. Toxik offers you the option to composite using Russian politics.
Meanwhile, Nuke is production-proven, has great kung-fu under the hood, and an "interface" that makes Kodak's Cineon look luxurious. With a fresh take on how it might be bundled, dressed-up, and marketed, Nuke might just pull out ahead in the race to suck the least in the world of desktop compositing.
Reader Comments (17)
I love your last statement. "...to suck the least..." but so true...desktop compositing seems to be all about knowing "the work-arounds" of any app to get what you want or at least close to what you want...
lol. Completely agreed on the various packages...as a studio we've moved our main comp package from After Effects (which we still use quite a bit) to Fusion (which stability-wise almost killed us on a couple of projects) to Nuke. and great statement re Toxic...the autodesk guys came to try and sell it to us and I could barely understand the licensing scheme, nevermind that it was mandatory to buy X amount of licenses.
We're absolutely loving Nuke, so here's hoping things continue in a positive direction.
I'll have to admit, I'm standing in line to take a serious look at Nuke from this point on.
We're currently looking at different packages as well, and I was wondering in what ways Fusion "stumbles and falls", so to speak.
Like Adam said, Fusion's biggest issue is stability, especially with large/complex comps (the kind that are SOP for film composites).
I hate to sound like a Fusion fanboy but i believe the phrase "Fusion's biggest issue with large/complex comps is stability" is totaly unrealistic.
I have done realy huge comps without having the slightest problem.
Also, a recent article about using Fusion for Frank Miller's 300 had a similar comment about big comps in Fusion.
http://www.vfxtalk.com/forum/showthread.php?t=9094
I know there are well-respected facilities that use Fusion heavily, and honestly I don't know how they do it. An artist I know and respect well suggested that Fusion should ship with a slot machine arm to attach to the side of your monitor, so unpredictable was his experience with the app.
The pattern that seems apparent to me is that folks who have been using Fusion since the DF4 days seem to be using and loving 5.1, but few new users are adopting it. I've seen this pattern before with other apps and it points to a user base that has learned to work around an app's particular funkiness. But if you're used to the stability of After Effects or Shake, moving to Fusion can be a scary process from what I've seen at The O and other places. I wish it wasn't so, as I like many things about Fusion. But speaking from purely personal experience, I find a new bug almost every time I use it.
I last used Fusion on "Sin City" at CafeFX, and while it didn't seem to have any issues with the size/# of layers of the comp, the first time I started using the integrated paint, everyone insisted that I stop and find another way to do what I was doing. They seemed scared to death of the stability of the paint node.
Nuke's interface has always turned me off so much I have yet to try it.
Sean, was that DF4 or Fu5?
DF4, i believe. Just as I left there, they started testing DF5.
Well, one of the things added in Fusion 5.1 was multistroke paint. It would get a bit tough with every individual paint stroke being a seperate spline, so now you have the option of doing a more 'raster'-type of paint on the image.
It's funny, but at Frantic we haven't been having any major problems with Fusion 5 (adopted it studio-wide just before the release of 5.1). This is probably partially a direct result of having used it since version 2 and internalizing workarounds as being regular workflow. There are issues that we're currently ironing out with development, but it pretty much works the way I expect it to (and it's on a project that's rather... complex).
It would probably benefit eyeon to look at what we're doing differently at our respective studios and focus on improving those parts of the app. Even with all my complaints about some of the aspects of Fusion, it still works really well for us.
-sean konrad (a different sean)
Hi Other Sean, :)
I think you're perfectly illustrating my theory. Fusion 5, for us, was unusable. But we didn't expect it to work like DF4 because we hadn't really used DF4 much. But in our comping software we used expressions a lot. So we tried to use the cool new "simple expressions" feature in 5, only to discover that it just plain did not work. A DF4 veteran would probably have just used the old publish/link method and been fine with it, but we got hosed because we didn't know what to trust and what not to. Result was that we had a rather miserable go of it in 5.0. Things got better in 5.1, but not enough to keep us from taking a serious look at Nuke.
Same story with us, we are mostly a Shake shop, stable and tested feature after feature. We have a few Fusion guys hanging around, and its not rare to hear them yelling from across the room. I've mostly been a Nuke guy cause i like the geek factor it has built in (never was a fan of pretty GUI comp apps). But its hard to compete with a $500 dollar app with free render nodes.
Exactly the same with us, Stu. We basically made the move to Fusion with 5, and just didn't have the background to know what worked and what didn't. Especially with network rendering - we had to babysit every render, and most of them crashed at some point.
If we had known what to avoid we could have built a workflow around issues, but coming into Fusion without this foreknowledge led us directly to Nuke.
Yeah, expressions were a weak point before 5.1, and even then it's still a bit weird in points. Taking a Shake artist through the system usually causes a bit of a headache with lots of "Really?"s :). I use them in macros semi-frequently and they work, so that's something.
I've used Nuke a bit and I found issues with it beyond the aesthetics/usability of the interface, some of which would prevent us from using it, or at least make our lives painful if we didn't vastly change the way we work. Obviously it's used successfully, so I won't pretend it's not a usable or even an awesome option -- just not for us at this time.
-other sean
Just another personal perspective. Fusion 5 was the first node-based comping app I ever touched coming from a combustion background, and I just picked it up and ran with it (this was 14 months or so ago). I couldn't believe how much easier it was to deal with than a layer-based system, and how quickly things just... came together. To this day, I have yet to see it crash. I'm just now up to 5.1, and I've been running what I (in my limited experience) would consider pretty complex 2K+ comps, and it's been an absolute breeze. I firmly believe that there are different apps for different chaps, so to speak, but I personally love the minimalistic interface, the fluidity of the shortcuts and node controls, and the ridiculously speedy workflow. The network rendering has a very "wham-bam-thank-you-ma'am" feeling to it as well.
Meanwhile, I'm starting to poke around at shake when I feel like booting into Linux, since everyone and both their second cousins seems to use it, and I have to say it actually comes pretty easily. Nuke has always looked a bit...strange to me from the safe distance I keep between us (for now), and I can't get AE to do the same backflips as combustion, so I've parked that train. But I'll still be very interested to see how Nuke evolves into its next iteration, and maybe that will be reason enough to grab it by the horns.