My three tips for machinima in SL

Hi everyone,

It's great to see all these videos coming out!

I've just put down my top three tips for filming machinima. If anyone wants to add a tip please free feel to do so.

#1 - Hide the UI/SL controls entirely.
You will need the debug menu active to do this (CTRL SHIFT ALT and D toggles this)
Pressing ALT CTRL 1 will toggle the UI/SL Controls on and off.

Note: If you don't want to see your HUD attachments you will need to also go to the View menu and uncheck Show HUD Attachments.

#2 - Try out using the new follow cam scripted objects. You can get some great shots by playing around with these new functions in v1.9

An example script using the follow cams can be found here:
https://secondlife.com/tiki/tiki-index.php?page=LLSetCameraParamsTest

Or just ask me for a copy of a example attachment when you are in world.

#3 - Run SL in a window and reduce to the window size, to the resolution you are capturing at (e.g. 640x480). Takes a while to get used to, but for capturing video will take a little strain off your graphics card, if you are using a program like Fraps for capture.

Anyway that’s my three tips... happy filming!

Harry

#1 allow people to refuse updates until their project is completed. If it's critical for something to work in a certain way for a certain project, and updates can change things at any time, then any serious developer needs the ability to keep the current abilities of the program. If folks are just messing around they'll only be frustrated, if investors are involved things can get very nasty.#2 #1 again because without that there's really no point in discussing it in the long term.

only thing i dont like with this, Is film makers have to learn how to script now...
yea no thanks lol, scripts do a good job at making gizmoz for camers to film with (you know who you are) so you guys - keep up the good work lol.

this is an example of the non user friendly sides of SL.
This is a GOOD example of team work and fellowship in SL though =)
so its really a good thing =)

Harry, one problem with disabling the GUI all together is that particles tend to persist on screen. I'm not sure if I'm the only one who's seen that effect.

I'd say the better option would be to shoot full frame and crop the GUI out of the video. I say this without know how the heck to do that.

Those are good tips...but it's Ctrl-Alt-F1 (not 1) to hide the UI,
and you can also use the keyboard shortcut: Alt-Shift-H to
hide any HUDs...

...and Ctrl-Alt-Shift-4 to hide your avatar. =)


- Infrared

Infrared,

Check the date on the posts. They're from April 2006. The turn off UI combination has changed since then.

Cheers.

My first thought was "Who the hell is Harry Linden and I think we're way past those incredibly basic suggestions now." Then I looked at the post date. =)

I know many of us have tried to start a thread, which we hoped would become sticky, to offer many ideas and concepts/techniques for machinima creation in SL, without success.

LL offered the wiki as a focal point, but I left that group after getting tons of emails spamming me daily. Many of which we just simple inserts such as 'I agree'.

I look forward to Havoc4, this will at least create more stable server sims in which we grab footage. Anyone else get the 10mtr jumps? Or the 'Go to Jail' (camera jumps to either an adjoining sim, or the sim edge at random times)?

There are no flat rules that suit everyone when grabbing footage. It is dependant on system, resolution, subject matter (portrait/landscape/square) etc. I have found that having just a fast GC and heavy weight CPU doesn't address all the requirements. A good HD array is important as there is no point grabbing footage if it cannot be saved quick enough. I hardly ever use codex when grabbing footage, I prefer to have uncompressed raw data. Re-framing and zooming with pan/crop into a high resolution section of footage, is preferable to having to continously reset my view. I always feel like I'm shooting through a letter box otherwise lol.

When I grab footage, it works out on the basic rule of thumb, 1 minute viewing time = 100gb raw data, this is post produced into a format resulting in around 5mb per minute viewing time, including music and sound effects, always trying to acheive less than 100kbps data rate, without using heavy codex application, as this simply defeats the object of getting the data to stream quickly, but become bottle necked by the client stalling trying to decompress the data. We aim for 25% or less loading on the client.