Took off the top part of the organ

Took off the front cover

Took out screw below front cover

Took off top under-panel

Slid out the key cover

Took out the stops bar

Took off the wooden bar that holds the keys in place

Took off several keys to take photos, numbered each key (1-61) with a pencil.

Made measurements:

From the measurements we took, based on the information from the MIDI 9 website it appears that the sensor strip will fit.

Removed 2 black wooden blocks from the left and right sides of the keyboard panel.

Took out 10 of 11 screws. The 11th screw was damaged. Help needed.

Took out screws from under the front of the organ.

Instrument builder, Harald Beckstrøm, helps out. Took out screws from the base of the organ.

Took out screws on knee levers.

Removed the front panel.

Repaired damaged foot pump canvas.

Replaced old valve covers (shown above) with new leather.

Replaced the felt under the keys.

Attended Making Sense VI physical computing workshop at Atelier Nord where I put together and tested a distance sensor with a midi interface from Interface-Z on the stop levers of the organ.

When I received the KS-61 MIDI sensor strip it became obvious that the distance between each "actuator" on the sensor strip compared to that of the organ keys was extensive enough to cause problems. I asked Harald Fetveit if he could help me to find a solution for installing the KS-61 MIDI sensor strip under the keys. The next images are of measuring and adapting the sound module to accomadate the sensor strip, and to find out how the keys could be modified so that each key corresponds to the appropriate "actuator".

Sensor strip (metal thing) upside down.

Measuring and marking where the keys have to be adjusted.

Cutting wood out from the centre of the sound module.

Cutting wood out from the sides of the sound module. The sensor strip now fits into the sound module, but the organ keys still have to be modified.

Cosmetic surgery:

Simon Steggell helps out with sanding down the body of the organ.

I treat the top part of the organ.

Key modification:

Harald Fetveit found a solution for modifying the keys so that the actuators on the KS-61 MIDI sensor strip are activated by the correct keys. Annesofie Norn, scenography student at the Norwegian Theatre Academy, helps out to see if the keys have been adapted correctly.

Carle Lange, my supervisor, visits my studio to see how the work is progressing. He discovers that the board which holds the bellows in place is not glued as previously thought, but screwed down. Good news, as it makes it possible to get a good look at the leaky bellows for the first time.

Carle removes 4 screws to release the footpumps which are attached to the bellows.

Inspecting the canvas on the bellows, we decide that it is best to replace it rather than trying to patch it up.

Finally installed the KS-61 Midi sensor strip with Harald Fetveit. A couple of the keys had to be modified and the sensor strip was removed from its metal bar and mounted onto a wooden strip to acheive the correct height in relation to the keys.

Tested the KS-61 Midi sensor strip with one of my messy max patches.

Made a protoype projection screen for the projected imagery that will eventually come out of the organ by extending the blades of a fan. When the fan spins fast, the image seems to hang in the air. I think the speed of the fan will be controlled by the knee levers that control the volume of the organ. It should stand max 3m behind the organ.

Took the leaky bellows to The Norwegian Organ Workshop in Snertingdal (ran by the Shrøder brothers) for repair.

Prototyped an analogue version of emitting aromas during a Motherboard theatre production called "Ikon". It is made up of an air compressor and 8 tubes connected to 8 airbrush guns. When the valves on the guns are opened, (in this case manually by the two women you see below) the aromas contained in their cups are released.

The system will eventually be automated by using push-type solenoids (image above right) to open the valves. The aromas (2x bacon, lemon, horseand stable and garden mint) were supplied by Dale Air Aroma Design. (I also made a Max patch for cotrolling 4 lamps via sound input with the assistance of Pjotr Pajchel that, regarding The Emotion Organ, might come in handy later.)

Collected the bellows. Re-assembled the organ to check the bellows - but there seems to be a leak between the wooden supporting board and the sound module. Worked with Harald Beckstrøm on cleaning and tuning the keys. Additionally we found that when we replaced the stops bar, several keys where being depressed unwantedly. We discovered that the two wooden blocks that seemed to be superficial/aesthetic parts of the organ (placed at each side of the keyboard) each have two screws underneath them for adjusting the height of the stop bar, which rests on top of these blocks. By adjusting the screws, the stops bar was lifted slightly higher over the keys and the problem was solved!

Received 6 reproduction tin phonograph horns which will eventually have small speakers installed inside for the electro-acoustic sound (manipulated organ sound). Bought 6 goose neck microphone holders that will be attached to the horns. There will be 2 stands, each with 3 horns attached on each side of the player. The player will be able to adjust the position of the horns as desired.

Bought a complete set of new screws for the organ. The old ones are getting worn out with all the assembling and dismantling.

Worked with Erich Berger and Simon Steggell at Atelier Nord, Oslo, to construct sensors/solenoids for the organ.

(Simon is my 17 year-old son and is studying electronics at Elevebakken school in Oslo. As part of this course is spending a week at Atelier Nord under the leadership of Erich to get practical work experience!)

TASK 1: AMANDA WORKS ON DISTANCE DETECTING SENSORS
1. Tested the distance detecting sensors (1) with the 8 analog/8 digital interface from Interface-z.

img: interface: 8 analog and 8 discreet inputs

2. Soldered wires together appropriate wires to allow for the installation of these sensors at the back of the stops bar on the organ.

img: distance detecting senosr and signal amplifier

3. Tested all eight sensors with the 8 analog/8digital Interface-z interface, and max adjusting the exampale patch from the Interface-z website.
4. Attempted to callibrate each sensor by adjusting the gain (sensitivity) and Zero point (scale controller) of each sensor on the signal amplifier (provided by IZ) that are placed between the sensor and the interface.

img: instructions from Francis and Zandrine at Interface-z

DISCOVERY
The distance detecting sensors produce different data results according to the refelctivity of the material used to activate the sensors. Skin works well for my purposes, but a non-shiney orange surface works best. The sensors cannot be callibrated properly until installed in the organ.

TASK 2: ERICH/SIMON WORK ON CUSTOM MADE BOX FOR CONTROLLING 8 PUSH TYPE MAGNET SOLENOIDS VIA LANBOX.

The solenoids for pushing down on the valves of the airbrush guns to release aromas will be contorolled via the 8 digital outputs on the Lanbox. Erich has designed a custom-made circuit board that will feed the signals via the Lanbox to the solenoids.

1. Order materials for building the special box that will make the Black Knights (push-type magnet solenoids) work with the digital output of the lanbox.
2. Construct the circuit board.
3. Buy the mateirals for the case.
4. Cut holes in the box for the power/37 pin connector.
6. Solder all cables/connections.
5. Test all connections.

img: Erich checks connections with Simon

TASK 3: ERICH/SIMON/AMANDA TEST THE SOLENOID BOX WITH LANBOX AND MAX.

img: Erich tests the digital outputs with a diode on the 37 pin connector on the Lanbox

Several tests and adjustments were needed before we got any results from the solenoids. Unsure of what data the digital outputs could receive, I emailed out to several email lists and was informed that the max patch example (2) I had previously used in Ikon to control lights could also be used for the digital outputs. We adjusted the max patch example (usb) found on the Lanbox website to test the solenoids. After a couple of rounds of re-soldering bad/incorrect connections, 7 of the solenoids worked perfectly:-) However, the 8th digital output of the Lanbox does not work:-(

TASK 4: AMANDA TESTS ACCELERATOR SENSORS FROM INTERFACE Z

These sensors are to be installed under the foot pedals of the pump organ to provide data about the relationship between speed and angle (slope) of the movement of the footpumps as the organ is played.

img: accelerometer

The cost for these sensors is 100 euros each. They are constructed of four potentiometers and provide data about horizontal and vertical tilting. (Actually, I only need one axis for the organ, but received the two axis version as result of French/English translation problems.) They have a range of -45 degrees to + 45 degrees. Francis from Interface z provided me with the following information about the sensors:

On the sensor, you have four potentiometers :
two for each axe:

- one for zero (zero in french)
- one for sensitivity (gain or sensibilité in french)
the one that is between the two connectors is a zero one. The two next are for the sensitivityof each axe,
the last one is the second zero.

Be carefull they don t turn the same side to increase or decrease.

A good way to adjust a first time is to put the accelerometer on the ground.

Then adjust the two zero so that signal is a the middle (value 64), and move the accelerometer in its maximum position and adjust gain to get the maximum of response . Do the same thing for the other side.

I tested the sensors by trying to simulate the movement of the foot pumps, and the results seemed okay. I am also thinking about getting two more for use on the knees levers. The problem with this method is that they will be exposed to the outside world rather than being built into the organ itself. While I have been trying to avoid this type of installation, having inspected the internal mechanics of the organ I can see no other option. The main reason for this is that the function of the knee levers (to open up vents that allow more air to flow over the reads and create greater volume) is also shared by a couple of the stops. If I install another type of sensor on the moving parts inside the organ the data derived will not be exclusively due to the movement of the knee levers. Conceptually I think this is a big problem, but will spend a bit more time thinking about it before I make a final decision.

Here are two attemps to visualize The Emotion Organ. The first was made at the start of my project in 2005. The second in 2006, which attempts to depict the placement of the organ in the Eidsvollsgalleriet, in the parliament building, Oslo. This image was made with the help of visual artist Ketil Nedgaard.

The phonograph horn sound system was prototyped in connection with an installation I made together with Per Platou. The "sound flower" was constructed by Aslak Nygren. It rotated while emitting seperate stereo channels out of the 6 horns. The rotation system was designed by Per.

The installation is called "IN DEATH VALLEY, EVERYWHERE WE LOOKED, GENTLY WAVING STANDS OF GOLDEN DESERT BLOSSOMS DANCED IN THE WIND, THEIR DAISY-LIKE FACES PUNCTUATED WITH VIBRANT ORANGE CENTRES.")

I also automated the intensity of 3 lamps and video control using Max, and got help from both Håkon Lindbäck (who also provided the rotating motor) and HC Gilje to make my patch more sophisticated than I could manage. The smell of an Egyptian Mummy was emitted via a dispenser care of Dale Air Aroma Design.

Worked with Carle's nephew Marcus (who is studying industrial design) to adapt both the airbrushes and push-type solenoids which are part of the organ's aroma emitting system. Marcus also built a frame/stand for securing the devices on the organ. Eight aromas can now be emitted (but I think 2 will be air freshening fluids!)

WIRING THE ORGAN FOR SMELL
Worked with Simon Steggell who helped with soldering wires to the 8 push-type solenoids to connect them to the relay/solenoid box he built with Erich Berger in February 06. Connected the cable from the relay/solenoid box to the lanbox, and then to the mac. Plugged in the the Midi 9 light sensor strip under the organ keys to the Midi 9 converter box. Plugged that into a midi interface and into the mac. Made a max patch that sends signals to the 8 digital outputs of the lanbox so that aromas are released when certain keys are depressed.

Click on the image below to see the details.

Results: 7 of the airbrush valves are depressed when the appropriate keys on the organ are pressed. The 8th solenoid did not respond.
Solution: trouble shoot the lanbox again via its own software (lc edit) to see if the lanbox is confirgured incorrectly.

WIRING THE ORGAN FOR (MORE) SOUND
Per platou helped out to put 2 contact microphones inside the organ and connect them to an analogue sound filter called Filter Factory.

Click on image below for details:

Though the mixer/amplifier/speakers shown in the image above will not be the ones used in the organ, the set-up gives me a chance to play the organ and experience the various ways both its music and the menachnics of producing sound can be modified to produce electro-acoustic effects. These will eventually come out of two phonograph horns placed on the left and right side of the player.

I found it difficult to remember the settings I made on the Filter Factory, so I'm documenting settings like this:

Aslak has built a large fan that will function as a projection surface. The speed of the fan will probably be controlled by the sensors that will be placed under the foot pedals of the organ - that detect the speed/angle of the player's "pedalling" - (partially) analogous to the volume of the sound. He has used the motor of a ceiling-mounted fan, and constructed propellor-like blades of light wood.

Here are the 2 "sound-and-light flowers" placed on each side of the player. There are 3 phonograph horns/blooms on each of the two stands/stems. The large and small horns have light bulbs installed in them, while the medium horns have Tivloi radio speakers for tranmitting the sound from the Filter factory to the player. The horns are on goose kneck microphone stnads so that the player can adjust the position of the horns as the wish.

Here you can see the spacial relationship between the player, the stands and the fan.

This image shows the installation of the distance sensors behind the stop bars, and the wooden boards attached to the stop bars for measuring how far they are pulled in and out.

The MIDI interface for the accelerometers installed under the footpumps.

This is the total equipment mass that shall be installed in the organ! What a mess.

HARDWARE

Click on this image to see the final hardware of the organ.

PROGRAMMING

1. Output patch: for lanbox control: proeller, lamps (dmx) and airbrush gun valves (digital outputs)

2. Input patch: for distance sensors (behind the stops) and accelerometers (under footpumps)

3.Input patch for the midi-fied keyboard: modification of Peter Elsea's chordcatcher patch that identifies chords as they are played (midi) combined with Matthew Mccabe's setcalc patch for identifying Forte names as they are played (midi).

Obviously I have tons of work to do! Not in the least is the matter of combining all patches into one main patch. And then there is the question of the mapping system .........

Piotr Pajchel helps out with programming, teaching me how to put my patches together, how to tune sensors and use video filters as data smoothers. We have a big problem porting my earlier max patch work to the new macmini intel, but get help form people on the max/msp developers list, amongst others, Oliver Larkin. Trond Lossius helps out too, but we still lack a USB control patch for the Lanbox, so we use my powerbook to send data from the lanbox to the macmini. It's not a perfect solution but it works for the time being.

During the dark hors we try out various video renderings on the propeller.

Here is the system of mapping all the elements of The Emotion organ:

There are 2 images stored in the macmini - a white circel on a black background and a mask of the dame image. Each of the 6 main and 2 sub-sets of Forte groups are asigned a colour on a chromatic scale of red to violet and white. When a Forte group is registered the hue of the circle changes. It takes 1 second to move across the chromatic colour scale to a new hue position, producing different effects.

Forte groups are also asigned a specific aroma. You have to play at least 3 unique tones in an octave to get a positive response. One finger play results in no colour and no aroma. If you play in one group longer than 10 seconds, you get an aroma from the appropriate airbrush gun.

The sound frequency from the filter factory (coming out of the horns and the sub-woofer) slides up and down depending on how many fingers you have on the keyboard.

The size of the projected imagery is controlled by the sound envelope. Different video effects are controlled by the 8 organ stops. Each time you pull out a stop the rotation of the projected circle changes direction. The more stops you have out the faster the projected circle spins.

The main speed of the fan is controlled by the footpump sensors. However, it is easy to get up speed with 8 stops out, and difficult with 1.

The velocity data from the keyboard controls the two light bulbs in the phonograph horns. They flash brightly on high values, and less brightly on low values. The sound envelope creates a smooth brightness effect in the bulbs in the 2 large horns.

There are 2 hidden keys on the keyboard that produce different effects. The to-p most F key causes the filter factory to change mode from an envelope follower to a random follower effect. A misbehaving key causes luminance and sound frequency to be reversed. It creates a deep sound (vibrations in the stool) and a circle with a hole in it!

It is impossible to capture the visual output of the organ, either with still camera or video. The effects are optical illusions that defy frame rate and shutter speed. It is truley live art!

After a couple of week's work Bjarne Kvinnsland is asked in to test the programming of the Forte group system. He is asked to play in specific groups and see if they correspond with their alotted colour. As neither I nor Piotr are mucians we have no way of finding out if the stored data is correct, so this was a great help. As far as Bjarne was concerned we had got it right!

Other testers were invited to open studio sessions, and as a result, adjustments to rpogramming were made.

It takes a while for people to find out how the organ works. Becuase the Forte groups are qualitive descriptors of chromatic scales most people first play in the most common harmonic range of western classical/pop/rock and folk music that produce a green and blue colour.

I am as yet undecided as to whether the technonolgy in the back of the organ should be left visible or hidden when the organ is presented in public in January. I think that this could vary for different occasions - you can't see the back of the organ from the front, so perhaps it is good to have it open.......