Scaling Productions Up/Down

[Rodney]

Nothing particularly new here but thought I'd post on this recurring thought because it deals with production economy.

 

Chris got me thinking in his recent post about a potential project. He mentioned the idea of a series of 10 episodes that he thought he like to create.

 

My thought is that in the making of these 10 episodes there is surely a lot of processes that are repeated over and over again. This got me thinking about the economy of creating serial episodes using non-linear methodologies. This would be the A:M equivalent to filming 'The Lord of the Rings' trilogy on location and then using the resulting footage from those live action shoots to complete all the films and release them in their proper sequence.

 

So how does this apply to a 10 episode animated series?

 

Target those areas of production that overlap and repeat.

Develop a plan to and a budget to get from the first shot to the last with a macro view on all sequences.

Then create the 10 episodes non-sequentially.

Refine them (mostly texturing, lighting and the featuring of unique props and events) sequentially.

 

At any rate, I'm just working the thoughts through my brain but would be interested in everyone else related thoughts and ideas. Some projects obviously will lend themselves to scaling in this way more than others.

 

There is an old saying, "I can do three things better (that is more efficiently and faster) than I can do one." The key is to have a plan (and stick to the plan) that will take advantage of the common areas that continuously overlap and repeated.

[Darkwing]

That's definitely a thought worth pursuing. I've been personally racking my brain to figure out a way to streamline a large scale production in such a manner because to use another old saying, time is money!

[Rodney]

We are fortunate these days to be able to use a lot of limited animation and compositing techniques that we don't have to spend a lot of money for when we implement those ourselves. We don't have to worry too much about nasty chemical mixtures and superimposing images onto film and then shipping them off to be developed either.

 

Out of any given recurring episode I'd guess there is easily five or so minutes of repeat footage that will play repeatedly in each episode. At a conservative guess I'd say the major sets might be conservatively limited to 3 per episode with some well placed supersets to give the impression that all episodes are of the same complexity. (It's amazing the details our brain fills in when past events are recalled to memory) The starfields the ships move through should be fairly straightforward and repeatable with quickly placed overlays and adjustments in color, hue and blur allowing for a wide array of variations. Close up shots on primary crew members could be pre-planned religiously to minimize the need for extensive lipsync. I'd say the more the script stays away from talking heads the better unless you can work out some effective limited animation techniques for character dialogue. As many scenes may have the crew sitting that should aid in not having to animate a lot of feet. (All this assumes your project might be ST:RS or a series with a similar theme)

 

Of course the more you can invest in each episode (I'm talking of time more than money) the better the final quality can/will be.

[largento]

As someone currently doing a 10 episode webseries in a crazy short amount of time, I've certainly approached it with economy of time in mind.

 

However, many real world time savers simply aren't relevant in the digital world. I had considered shooting out of order, doing all of the scenes for one set or one camera setup at the same time, but the reason for doing that in the real world is that you don't want to have to tear down the lighting, set it up somewhere else and then tear it down over there and set it up again on the first set. That's not a factor with A:M. I have the ability to save and make copies of setups. If I like a camera angle, I can come back to it by just opening that choreography and doing a save as. I don't have the space limits of a soundstage that dictate how many sets I can have available to me at any given time. I don't have to take a crew on location. :-)

 

What's more important, I think is that there is a flow between shots. I'm finding it much better to edit as I go, seeing how one shot plays into the next. You're not going to be able to do that if you shoot non-linearly. Most real world filmmakers would rather shoot this way, too, but they have to cope with the real world and we don't. :-)

[robcat2075]

I think it's about tradeoffs. If you economize on one thing something else needs to shine to make up for it.

 

"All in the Family" was cheap on sets; practically the whole series was in one room but they had great scripts and actors to fill it.

 

A Michael Palin travelogue series has pretty minimal scripts and no actors but great locations to show you.

[Darkwing]

A particular element we're using on Red Squad, is some basic digital matte paintings. There's a couple sets that are seen for a total of a handful of shots and for that, don't deserve the amount of time it would take to build the entire set. For that reason, I'm building only certain key foreground elements for those sets and my brother is matte painting the background. We haven't done any yet, the only matte painting ready is a space backdrop for one of our star systems. But still, the point being that something like matte paintings can save a significant chunk of time on certain elements

[largento]

Set design is an area where I think we can get a lot of value from studying what they do in the real world. Since I've been re-creating existing sets, I've been marveling at the practicality and skill that went into designing them. It's clear that as much thought went into their functionality as into their visual interest/appeal. In 3D, there's this immediate thought that you can build the whole thing, but it's almost never necessary and even in 3D, you still need to be able to place your cameras and lights in the right place to get the shots you want.

 

I was going to build two sets for a couple of scenes, when I realized that I only needed to build part of the second room that was accessed through a doorway in the first set. So one set became two, saving me a lot of time.

 

The bridge set is another example of how aping the real world can be helpful. The bridge set is composed of 10 wedges. By making them individual models and piecing them together in a choreography, I can selectively turn off wedges ...or thought of the other way, I only need to turn on the ones I need for the shot.

 

Aesthetically, too, it's interesting to see how they designed the sets to provide texture and visual interest for the backgrounds.

 

Certainly there are elements that aren't relevant. For instance, some sets made use of forced perspective to make them seem larger. We don't have to do that. But all in all, I'm finding it very educational and I've definitely developed an even greater appreciation of their work.

[Gerry]

This is a good topic Rodney, especially since you talked about an efficient production pipeline over on the Nightcallers thread when I mentioned the flat shape you created for distant mountains. It was an "aha!" moment for me because as one guy I don't HAVE a production pipeline, but that's no reason not to think in those terms. So even though I've got all my models built I still want to develop efficient habits, something I've had the luxury of not having to think about.

 

In the scene I'm developing I initially thought to do it all in one chor with multiple cameras, but then I realized there's no reason to make the renderer even have to think about anything that's not actually in the shot. So as a first step, I did several "Save as..." chors, then went through each one and deleted anything that wasn't in each individual camera shot.

 

It's a small first step toward working as economically as possible.

[Rodney]

as a first step, I did several "Save as..." chors, then went through each one and deleted anything that wasn't in each individual camera shot.

 

It's a small first step toward working as economically as possible.

 

IMO, you are on the right track here Gerry.

 

What I need to learn myself is that sometimes just a simple flat plane behind a character can serve the purpose. Perhaps even better no plane at all and an Alpha Channel so that the imagery can be used and abused in artwork, advertisement, etc. etc. etc.

 

I'm not particularly trying to champion compositing here but I think that is an inevitable that we must consider. If using EXR format as our standard image format to render out to that also will give us a lot of options to work with. I am trying to move toward EXR but that can be a tough move when I'm not thinking long term. I'm too conditioned for the immediate gratification of rendering out to PNG. My goal is to have Netrender push out two images per (one PNG and one EXR). I might be able to drop that requirement just to EXR and then have a background process convert to PNG and any other required format (i.e. .SWF and .MOV)

 

The benefits to compositing are many as the assets used can then be reused and manipulated in many ways while retaining the original image also. Of course if we are well organized we'll always have the setup in A:M to refer back to also.

 

Having the assets is one thing. That doesn't help us a lot if we can't find them.

For my part... I've really got to work on organization.

I just saw a 3TB external drive that is calling my name...

need . to . appropriate.

 

The bridge set is composed of 10 wedges.

 

We need to get you some screen capture software so that you can record your sessions for posterity. Some day the fans of your shows will want to look back on the creation of these 'early' shows. This will especially be true in the year 2233 when the "Be Largento" campaign goes into full force!

 

The tendency to try to recreate whole worlds is very hard to resist. (But reality will reveal itself to us and teach us restraint in our efforts)

I think the earlier we can break away from this however the sooner we'll see success in our one person/small studio projects.

Save the whole worlds for when you have a contract signed to produce the game and can afford to put other people to work create those fully realized scenes for you.

[Darkwing]

On a side, how exactly does EXR work. Like I have a primitive idea of it, basically being able to adjust other things like lighting or something in post, but in what software do you do this? What is the range of EXR etc?

[Rodney]

I should let other answer here as you are no doubt getting tired of reading my thoughts... (What can I say... I'm infinitely interested in it all)

 

On a side, how exactly does EXR work.

 

First and foremost, even if you did nothing else other than just render to the EXR format you'd have access to a more robust and higher dynamic range image.

I don't know all of the specs... I need to learn them myself. EXR 2.0 is suppose to be out already, but I haven't seen any announcement. Trust me when I say you'll want to use EXR 2.0 format if we have access to that. Getting to know the current EXR format will give you a leg up on that.

 

 

basically being able to adjust other things like lighting

 

The EXR format stores the information required to alter the lighting in your image on-the-fly.

It's important to understand that you have to specify that a particular Buffer be created in order to be adjusted.

For instance, it should make sense that if you don't tell A:M to render out Light Buffers the EXR image will not store any information in a Light Buffer.

 

or something in post

 

Any image can be adjusted in post but because EXR can store customized variables it is better suited to compositing and manipulation in post than other formats. The problem is how to access that information...

 

but in what software do you do this?

The best answer I can give you (especially on a budget) is... use A:M's Compositor.

In current versions you can also leverage Composites as Rotoscopes, Decal, Patch Images and elsewhere where other image formats are used.

(Note: There may be some glitches here and there but that is because features that are seldom used tend not to have their bugs reported)

 

What is the range of EXR etc?

 

Rather than have me mess you up with my interpretation I'll send you to the source: http://www.openexr.com/

 

With George Lucas's Industrial Light and Magic behind the format (and Avatar tech being added in OpenEXR 2.0) I think it's the premiere image format. Well worth taking some time to explore. For those with the money to spend you can expect to see most digital cameras and software support it. The issue for most at this point is low end software for those that aren't use to investing money in technology. Those products (freeware programs and such) will be coming along at a slower pace but I think the rollout is mostly inevitable. The dude who gets a browser to support OpenEXR 2.0 will have accolades piling up at his door.

[Darkwing]

Oh fascinating! Oh...having ideas! Yes, this could be beneficial, maybe even save some time in certain ways, will have to play with it once I get AM again. Must admit, I'm beginning to go nuts without it and it's only been a few days!

 

And don't worry Rodney, you have expertise to offer so don't hold back because you feel like your posting more frequently than others! I value your opinion as I do everyone's here and this thread is already proving to be a very informative thread!

[Rodney]

Thanks for the words of encouragement. That means a lot to me.

 

For those wanting to explore the world of composting in A:M further:

 

As an intro to the world of Compositing and OpenEXR in A:M I suggest the following steps be taken immediately:

 

To start using A:M's Compositor I recommend using a simple setup.

Currently I tend to always start with a Mix node (for our purposes I'll use the term 'node' interchangeably with the term Post Effect because in the long term that may help you understand how the Compositor works better)

 

- Open new (empty) Project of your choosing to experiment with and Save it with a unique name.

 

- Open the Project Workspace (PWS) and Right Click on the Images container.

 

- Select New Composite (You should see a new Composite container appear near top of the PWS.

 

- Right Click on Post Effects and select "Create New Post Effect"

 

- Once this Post Effect (container) is created we actually need to assign it some properties so...

 

- Right Click on the Post Effect and select 'Change Post Effect to' and locate the listing of Post Effects available. While you can select any one you want I suggest beginning with a Mix. By using Mix nodes you can nest and add new nodes anywhere you desire in your workflow. Mix defaults to 100% for each of the images it is mixing so the resulting image will likely be too saturated. Consider lowering the input values to 50% for each image to get .

 

Here's a useful trick: Create a Mix with the same image for both inputs. Then add another Post Effect to one of those input nodes (Say... Film Grain or Blur). Now you can go back to your Mix node and adjust the amount of Film Grain/Blur (Input 2) relative to your original image (Input 1). This is a very effective way to alter rendered images and because we have access to so many different Post Effects . Considering these effects are animateable over time in A:M you may be tempted to stop using your other compositing programs altogether.

 

Now concerning the images you use... just use one image to begin with. It doesn't even need to be EXR.

 

Sorry.. I know I've only taken you part way here but Compositing is not a one post solution. Too many variables!

 

If you get stuck, just post a topic on it... if there is any interest I could be convinced to record a few videos.