User:FluffyDunlop/Proposal: Difference between revisions

From XPUB & Lens-Based wiki
No edit summary
(Blanked the page)
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:


<center>'''Project Synopsis'''</center>
My research focuses on the possibilities that sonology & computer models can offer for the development of sculptures (2d or 3d) and give a different awareness of places and spaces. It will give the viewer an object to look at and perceive the room in a different way. The way the room is visualised is trough the use of reflecting sound frequencies. This idea comes from a method, called impulse response. It is a method that I frequently use in my sound design practice. With the information that the impulse response gives, a sculpture is created to visualize the acoustic properties of that space.<br><br> The sculpture will be calculated and constructed in a 3d model by a computer and could be made in a physical object. The way the transition of a 3d model to a physical one is conceived needs to be determent, there are more ways of doing this and the size of the sculptures has a huge role in this.<br><br>
The outcome of this research will be presented in the exposition gallery. The form of the presentation could be pictures, sculptures or both <br><br>
The research starts in two directions that need to be one in the end. The first direction is the place or space and the second is the method of visualisation. What I found in my prototyping research stage is that you can't finish one direction without another. So it is important to develop them both and see where the different directions merge.<br>
<center><gallery>
File:ExampleConv_1.jpg|Impression of the sculpture in a space
File:Example_2.jpg|Impression of the sculpture in a space
</gallery>
</center>
<center>'''Engagement'''</center>
With this project I’m aiming to surprise the viewer with a new perception of a space, and to connect the viewer to the space in a different way. The idea of engaging the viewer with the space comes from the idea that there are acoustics properties that you might not be aware of. Through this properties of a space I can visualize my research and let the viewer be engaged and connect to the space in different way<br><br> Within the early presentations of my proposal to my peers and pzw staff, I found out that the subject of impulse response is hard to grab. So the engagement with the viewer has some other aspects than just "presenting" objects or visualizations. I could choose to educate in the case that the impulse response technique plays a more narrative role. Or I could choose to leave the technical explanation out, this in case that the work is strong enough and doesn't need any technical details to explain the work. What I need to tell in a technical aspect is defined by the outcome of my research.<br><br>
The idea of engaging the viewer with the space comes from the idea that there are acoustics properties that you might not be aware of. Through this properties of a space I can visualize my research and let the viewer be engaged and connect to the space in different way<br><br>
<center>'''What is an impulse response?'''</center>
An impulse response is a method to calculate the reverberant properties of a certain space. Trough the playback and recording of frequencies ranging from 0hz to 50kHz the resonating properties and thus the reverberant character of the space is calculated. Imagine a space where at one end speakers are placed and on the other end a microphone. Trough the speakers every usable sound frequency is played back and simultaneously recorded through the microphone. The signal that has been recorded through the microphone is now very important, because it contains every frequency that we've played back and all the reflections of the room that came within the recording. If we let the computer subtract the original frequencies from the recorded frequencies trough a convolution method, we are left with the Impulse Response of that room<br><br>
Below are the basic sound examples of a Impulse Response technique.<br>
A clean sweep to play - [[http://www.audiocheck.net/Audio/audiocheck.net_sweep20-20klog.mp3 A sine sweep from 20Hz to 20kHz]]<br>
A sine sweep from 20Hz to 20kHz recorded trough a microphone in a room - [[File:R1.ogg]]<br>
Impulse Response through deconvolution method - [[File:Jar_Response.ogg]]<br>
These pictures show a representation of an impulse response. The way in witch the impulse response is shown is called a waterfall diagram.<br>
<br>
<center>
<gallery>
File:ReverbWaterfall_1.png|Waterfall diagram of a Impulse response from a kitchen (measured in dB and Frequency)
File:ReverbWaterfall_2.png|Waterfall diagram of a Impulse response from an auditorium (measured in dB and Frequency)
</gallery>
</center>
<center>'''Workflow, Prototype & Research'''</center>
The use of impulse response technique is one that I use daily in my work as a sound designer. Either to merge instruments together in a musical composition
or to use within sound for picture applications, like recreation of rooms, microphones and telephone systems or to be applied to automatic dialogue recording to fit a scene better. With my daily use of impulse responses I can say I'm quit familiar with them and the technique.<br>
I started by setting up speakers in a 90 degree order with a microphone in between. This system, if mirrored and pointed to the opposite side of the room, can provide me with 4 impulse responses (1 for each speaker repeated once) <br>
<center>
<gallery>
File:IMG_8276.JPG|Initial setup
</gallery>
</center>
I found out that when I put the outcome of these four measurements trough my visualization workflow, that the amount of measurements was not enough to give a detailed model of the room. So I quickly abandoned my 2 speaker setup with 4 measurements. And went for a single speaker setup with 8 measurements. More work to get the angle of the speaker right but much more information due to more measurements. The distance of the microphone to the speaker is equal to all measurements and gave some weird mic placements within the measurement technique. I decided to leave the room in the way as I found it and to not mess around with furniture between measurements.
<center>
<gallery>
File:IMG 8278.JPG|Second setup
File:IMG_8279.JPG|Measurement 5 gave me some interesting microphone placement
File:IMG_8345.JPG|Keeping equal distance
</gallery>
</center>
These are the recordings and the 8 measurements of that particular space (room 2.9 wdka)<br>
[[Media:R1.ogg|Measurement 1]]<br>
[[Media:R2.ogg|Measurement 2]]<br>
[[Media:R3.ogg|Measurement 3]]<br>
[[Media:R4.ogg|Measurement 4]]<br>
[[Media:R5.ogg|Measurement 5]]<br>
[[Media:R6.ogg|Measurement 6]]<br>
[[Media:R7.ogg|Measurement 7]]<br>
[[Media:R8.ogg|Measurement 8]]<br>
From here the visualization research started. I looked at different ways to get a detailed curve out of the wave files, and the best way that worked for me was to import the audio in After Effects and apply the Wave Form filter to a solid and link the filter to my measurements. This gave me a tweak able visualization, where I could change the amount of points, lines, height and offset of the audio file. After I exported a Tiff image with the waveform, I imported the image in Adobe Illustrator to trace the image into a spline. With the spline of each measurement (8 total) I created a sphere like object in autodesk maya. This method if done with only 4 measurements gives a less detailed object to work with, this was the main reason to go from 4 measurements to 8.
<center><gallery>
File:Spline_From_Illustrator.png|Splines as exported from Illustrator
File:Maya_1.gif|Splines imported into Autodesk Maya and shaped in a spherical setup
File:Maya_2.gif|The gaps between the splines are now filled and the visualization is complete
</gallery>
</center>
'''Founding.'''<br>
First thing that I found out is that a multiple speaker setup is quite hard to do, and makes thing unnecessary complex. The best way is to stick with multiple measurements trough the placement of 1 speaker and 1 microphone. I have measured on a  horizontal level. After experiencing the horizontal method I question it may also be a good idea to go in height directions, moving the speakers up and directing outwards in a 360 degrees spherical angle.<br>
The moment the wave form is analysed with the After Effects method, the measurements will loose the 1:1 accuracy. From here on out the model is not scientific accurate anymore. How This influences my research and what it does to the project I have yet to determine.<br>
'''Next Step.'''<br>
With the first prototype made in a 3d shape, the next step is to put this model into the physical world. For my other prototype I have been experimenting with visualising impulse response through 3e printing (see below) <br>
I could do the same and find out what kind of shape I'll get. Some questions came out the feedback session and it seems good to research them:<br>
'''What role the location of the measurement has in the project?'''<br>
'''How can I connect the object to lets say a room and what does it mean?'''<br>
'''Could similar sounds like impulse responses been made on a certain place?'''<br>
<center>'''Second Prototype'''</center>
During the tryouts with the above method, I experimented with different spaces and setups. One that worked well and had interesting results was the impulse response from a glass jar.
The jar was placed under a microphone and a speaker was placed inside the jar. One measurement was take and visualised through the same process as above. Through the 3d process a flower like object was generated.<br>
With the nature of the sound in mind and the flower like shape that was developed through the process, I decided to print the object in 3d and put it back in its "origin".
<center>
<gallery>
File:FullSizeRender_(2).jpg|Glass Jar with speaker
File:IMG_8341.JPG|Print process
IMG_8340.JPG|Final Product
</gallery>
</center>
Bibliography:<br><br>
Research<br>
[http://www.audioease.com/Pages/Altiverb/sampling.php Impulse Response as explained by AudioEase]<br>
[http://www.audiocheck.net/testtones_sinesweep20-20k.php Audio Check Test Tones]<br><br>
Similair Work<br>
[http://notnot.home.xs4all.nl/solidspaces/solidspaces.html Driessen & Verstappen Solid Spaces]<br><br>
Technical Tools<br>
[http://www.autodeskt.com Maya 3d software]<br>
[http://www.tamasoft.co.jp/pepakura-en/ Paper Mockup Software Pepakura]<br>

Latest revision as of 09:31, 8 September 2016