May 19, 2011

piezo transducers:
A piezoelectric sensor is a device that uses the piezoelectric effect to measure pressure, acceleration, strain or force by converting them to an electrical signal. thanks, wikipedia.
when a piezo transducer is in contact with a surface, when it vibrates it will cause the sensor to output sound the same way a microphone does.

to make the whole setup i needed 10 piezo transducers (one for each section of the wall as the vibrations couldn’t travel between as it was not a solid sheet of metal all across the wall. I had to buy 100m of 2-core cable to wire the sensors with enough length to sit in the middle of each panel to get the best spread of vibrations. These wires plugged into a box at the bottom of the wall to connect them in series, and then this box had an XLR microphone out as this is a standard connection for a film shoot and there was lots of xlr cable we could use to run it across the warehouse to where we had the computers set up. then i just needed an xlr->mini jack transformer so the signal could be run into my audio input on my laptop, which was running a max/MSP patch to detect attacks in the sound.
i had to buy:
100m 2 core cable
10x mini jack plugs (mono)
10x mini jack sockets (mono)
1x xlr socket
10x piezo transducer
10x 2-pin connector
1x circuit box
1x xlr cable
1x xlr – mini jack transformer

i had cut the wires to the lengths i calculated and then on the end of each one i soldered a 2-pin connector on the transducer end and a mini jack to the other end, i soldered a 2 pin connector to the transducer so it could easily be connected and disconnected from the wire. i then drilled 10 holes in the circuit box for 10 mini jack sockets and wired the sockets up in parallel-series, with 2 sets of 5 parallel connections, to give the least resistance.

i had been considering using arduino but it seemed overcomplicated for this task and i didnt have any experience with the arduino code, and i do have experience with electronics and max/MSP.

For the max patch, i used an object called “bonk” which detects attacks in the sound and prints an output whenever the signal is above specified parameters (how high the attack, how fast the decay), which i then added a crossover object so i could filter certain frequencies to make it more accurate. Then i made the print function connect to a flash server so the printed message is sent to flash and can be programmed in the game to signal a ‘hit’.

here is my max patch for download it will work with any sound (like a clap or a hit, something with fast decay) and also can work with a normal microphone, it requires the bonk object and flashserver dependencies which can be found online. the slider is gain to boost the low output of the transducers.

more content

May 12, 2011

a couple of videos wouldnt upload to vimeo properly because of the codecs we had to use for the media server, so here is a link to the download

Test shoots

May 12, 2011

heres the wall with my content while we were testing before the shoot

Lava scene

May 10, 2011

To make the scene to the specifications from the discussion of the storyboard, i had to use some fluid effects to make the lava. I used cinema 4d to make some 3d geometry that would be the cooled lava, and then i imported it into realflow and made a circular emmitter for the lava to erupt from like a geiser. There was some really heavy render time in simulating the fluid dynamics. it was also a steep learning curve as i had not done this before. Once the fluid shape was rendered it i imported it back into cinema 4d as a mesh to texture it and set the camera so that the lava would appear on the floor of the warehouse when it was projected and the perspective would match and look good.

i then sent lynne the PNG sequence so she could add some smoke in after effects as i had to work more on the game content, spaceships, and minor revisions for the client whilest i was in london


May 10, 2011

there was a lot of changing of the content, it was a lot different to working on my own project and having complete creative control, but the feedback was really helpful. here is the evolution of the aliens i made:

because of the talk about projection mapping and the importance of bringing the whole thing into something very three-dimensional, i thought the pixels of the alien should be 3d cubes, and i thought it would look good in the alien could appear to come more into the foreground when it is in the center of the wall and it could wrap around the wall (originally it was a curved wall). However this was not possible to implement in the flash game (which was already in progress). however i kept the idea of the 3d blocks

i designed a set of 6 aliens on darrens orders, they wanted it to be very colourful and because of the space invaders theme i used very atari-esque colours. I had a talk about copying the exact shape of the original space invaders but we were worried about copyright so we decided it would be best if we altered them slightly and made up new shapes of aliens in the same style

In the game they looked static so we added 1 frame of animation so it could flick between them like in the original game and this was effective, clients were pleased.

this is just a look at the complexity of the 3d models, although in the end i had to scale them down quite small it is not obvious, but there is a lot of detail. I though it would be best to work at a high level of detail/high resolution as i was not originally given the final size they would be in the game, or whether we would need the models for some other part of the game, we did not originally know the size the projected image would be.

projection mapping

May 10, 2011

Darren showed me some videos that he and his client liked for projection mapping ideas at the beginning of the project, we were looking at how to create the illusion of depth in the wall, and have it appear to change and have the space invaders feel like they were in 3d space and things. We talked in depth on the phone/skype about ideas and possibilities. in the end we went with taking the photo of the wall we had from the test shots and using it to create visuals so we would project the picture of the wall back onto the wall and it would look like it was changing. You can see my storyboard of ideas i came up with for this below. In order to create the visuals i had to take the photo of the wall and photoshop it to look flat and i also had to take out the light from the windows because they would be blocked up on the shoot. This was actually quite a drastic change to ‘shop in. before:


May 5, 2011

Time plan


Here is the feedback from the end client on the storyboard ideas:

Hi Jemma and Lukas

thanks for the last couple of days, had a really dark flu feeling better today,

can you both read though notes from yesterdays meeting and answer the parts that relate to elements asap,

Cheers Darren

Hello mate,

Thanks for all of this. Went down well. The clients particularly liked the moving lava stuff at the bottom. Lots of questions and comments we need to answer and address.

Can the Invaders come down on the screen like they do in the real game. In waves from side to side and then down?
Assume we don’t have time to see projected test until Thursday in the location.
How high is the tower going to be and could we put a camera on it to record a big wide shot?
Can the tower be drapped in black so it disappears.
How much light do you think we’ll get from the projections. Need to work out how much light we need to light the players and how much spill we will get from the projections.
Can’t remember from original emails but I believe that you are providing sound system for effects in game and what size system are you getting?
Also believe you are bringing in power. What size generator are you bringing in and we can see if we can share power that we need for lighting etc.
How difficult will it be and what are the implications of being able to project a live feed leaderboard on the wall somewhere that updates automatically as the players score.
We talked about how the game would work and the following was agreed by everyone but I need to get okay from you. Each player would do the first screen and hopefully get them all easily and therefore end up with the same score. They would then all play the 2nd screen, which would be faster and may separate out the scores a little. Then the third getting faster, fourth and finally the fiftyh which would be super fast. We agreed that if the invader gets to the bottom that the player would lose 100 points for each one but not keep counting down and down as it does in the test. We worry that it will make the players less up for playing if they see their score anihilated. This way we also then think we have 5 different rounds which get harder and harder and prolongs the excitement as they wonder what speed the next one will be. Also it gets more competitive if we have breaks between each round.
Are we still planning to have a big mother ship fly by at the top in every round?
The invaders will be in a block 3D construction. Will they throw shadows against the wall as they move?
Timing for the warehouse on Friday is now shooting 6pm till 8pm.
Still debating what happens at the end of the game but want the wall to seemingly collapse and reveal something behind.
They like the idea of the metal wall having holes ripped in it when the ball misses and we just need to confirm with them what will be behind it. Is there a limit to what they can or can’t do? Can it move or does it need to be static?

Think that’s it for now but give me a shout once you’ve had time to digest.



Final Product demo

February 1, 2011