[McArdell]

I suspect you are looking to get multiple matte outputs from one action node? Not possible unless like you say you copy the node and choose a different layer to output on the copy.

If you just want one layer or multiples combined together that is possible, but there is only one matte output out of action.

Jeff

[McArdell]

Try defocusing both fill and matte thru two identical sparks, divide resulting fill by resulting matte before comping.

Jeff

[McArdell]

A lot of film flame work is done staying in log. This may seem odd as every algorithm in the system is linear based, but it is done all the time. Cineon files are imported directly, work is done and log files are exported. This requires you have a monitor lut that lets you see what your images will look like as you are working on them. This monitor lut is only used for viewing and usually is built by the facility to get close to the house that will do the film outs. So, in a facility that works with multiple film out facilities it will be common to have different monitor luts for different filmout places. Be aware that some things will not act the way you might think working this way and when you work with CG elements you will likely need to convert them to log space using a lut before comping.

In some cases you may need to convert log to linear to do the comp (comping fog, dust, smoke for example may be best done in linear space). In this case you will not use a monitor lut but rather will lut from log to linear at the head of the comp and apply an inverse lut that matches that original lut at the end to filmout log images. (When you create a lut in flame it automatically makes an inverse).

I have never used the luts that come with flame as I have always worked in places that had luts in house. I seem to recall issues with the provided ones though. Sorry, I don’t recall specifics.

Hope this helps, I know it is a big topic.

Jeff

[McArdell]

There are really three big events that come to mind in North America:

NAB in Las Vegas http://www.nab.org in April.
This is an equipment/software trade show. As a huge international show product introductions and updates are often announced. Less training and sharing, more inspriration and seeing newest stuff.

Siggraph http://www.siggraph.org location changes every year. More a 3D show than 2D but it is more of an education show than a trade show. It is a great show but may be less nitty gritty than it sounds like you want.

VES http://www.visualeffectssociety.com The Visual Effects Society hosts a yearly Festival of Visual Effects. Usually in LA or SF. Many sessions per day featuring effects crews giving breakdowns and “how we did that” talks. Again not training per se but you can pick up a bunch of info, tips, inspiration. I don’t see 2007 dates on the site which is odd because it is usually in July.

Jeff

[McArdell]

What is the material you are trying to track? Is it footage? Auto tracker will solve for the camera if it can… so there has to be a camera. What is this circle test you mention?

So no telecine repo, no processsing – just raw footage. I would also suggest tracking an image no bigger than 1k if you want it to succeed – use a proxy or resize a smaller same aspect image – track and scale.

If all of that has already been tried I would switch to manual tracking and try and help the solve by doing valid manual tracks. After that I would use external 3D tracking software.

Jeff

[McArdell]

paul_round wrote:

What material were you working on? I’m talking 2K12bit on Inferno5.5, Onyx2.

2k 12 bit feature film work on Inferno and Flame just recently moved to 2007 SP1 , 5.5.x before that. The errors you are talking about are memory ones… you have to break up batch setups into manageable sizes and do pre-renders. This seemed ludicrous to me at first but the more you get used to it you get a feeling for when to do it and it also starts to make sense in a setup as you prerender stuff that is not changing.

Also I find that usually I can work in proxy mode without breaking it up, even on burn and get things close before breaking it up.

Jeff

[McArdell]

I cannot even imagine working without burn. Even with only one machine (Flame, etc) I would have a small burn farm with all matching sparks. It is well worth the cost and while there are issues from time to time I use it day to day in a facility that has the largest burn farm in the world and we all rely on it with very few problems.

If you are having issues that would make you want to not use it, it is probably setup wrong or some part of the system is broken. It can and does work reliably day in and day out.

Jeff

[McArdell]

The automatic tracker works very well. I would advise looking at book/pdf because while it is simple there are specific steps (I cannot do from memory unless in front of machine).

One tip is to use a proxy size if you are working in 2K… it is exceptionaly slow otherwise. So make a 1k version, or even video rez – track that and then scale result.

Jeff

[McArdell]

Are you using an extended bicubic? Burn on 2007 SP1 will make these triangles if you use extended bucubic. Try turning shading on with no lights. This is reported to discreet.

[McArdell]

It simply allows you to put in one input and get out as many outputs as you need. It does nothing to the image.

I also find it useful if I have something in my batch that is far away – so rather than the line extending across the screen you can use a mux or muxes to change the direction (like make go outside the rest of tree).

I also like to use if I am breaking up a batch with pre-renders. I will often put a mux at the start after a pre-render because that resulting clip sometimes needs to feed several places.

Jeff

[McArdell]

I believe this was broken with SP1. I have not heard of a work around.

Jeff

[McArdell]

As I recall there was a technology demo at NAB users group showing a flame doing floating point… no date given as to when we might see that.

Jeff

[McArdell]

You will find that you have a lot more control comping these elements with logic ops in batch than in action because while action has some transfer modes they are often limited (like some have no opacity control). CG will give you these layers to finesse a comp and most of them are designed to be added, screened, multiplied, etc to the beauty. Logic ops in batch is great for this. Sit with your 3D person and find out how the elements were intended to be used.

There was a technology demo at NAB that showed the layers being generated in Maya and then building a comp automatically in Toxik using information passed from Maya as to how they got used. It just saved the manual part of building the tree and let you get right to tweaking.

We have also spent some time on this type of compositing in courses on fxphd.

Jeff

[McArdell]

rego wrote:

And what is better than a ferrari? 😀

A Ferrari with Burn? 🙂

Jeff

[McArdell]

Keyser_Soze wrote:

Pretty unstable. Wait is my advice.

This is pretty broad… do you have some specifics that would help people make an informed decision? My experience is that every upgrade has been worth it even with initial bumps. What exactly are you experiencing that would make you call it unstable?

Jeff

[Author]

Posts