StarSound in Action

The StarSound installation is finished for presentation and was set up in Blackrock Castle Observatory for three days so people could interact with it. All the visitors gave very positive feedback about the experience and really seemed to enjoy the installation.

Despite a few minor issue that will be addressed during the next days, no major problems came up and the installation works just fine. The dates for the final presentation are the 6th and the 7th of June.

Installation Setup @BCO

The installation itself has been set up for the first time. It is running fairly well, however a lot of adjustments still have to be  addressed. The star field is now being projected onto the castle wall, which makes it a little blurry, but also adds a certain style.

The technology such as the computers and cables still has to be hidden somehow. Still missing is also the fake data visualization as the interaction supporting visuals from the eMac on the floor. Both will hopefully be finished by tomorrow.

Exhibition Demo

The Demo application for the exhibition is slowly coming towards the end. Despite the connection to the telescope, the sketch controlling Stellarium and changing the visuals for the data stream is finished. The last thing to do is to adjust the resolution of the visualization sketch to the resolution of the screen that will be used during the exhibition.

For the telescope connection, all adjustments are made on the software side and it should work just fine, however the suitable adapter is still missing. The Meade #507 cable seems to be hard to get, as it would have to be shipped from abroad and probably would not arrive in time.

MEADEN ETX70 connection

After having set up the MEADEN ETX70 telescope, first difficulties came up quickly. The motor and the control via the MEADEN remote called Autostar works just fine, however, when trying to connect the telescope to a computer, one has to realize that consumer telescope technology is not really using latest standards and is far behind in terms of connection plugs. A special adapter has to be purchased for the ETX70, to be able to connect it via a serial port.

This adapter alone makes sure to come out of the telescopes remote plug with a serial plug. Additionally then, a second adapter is necessary from serial to USB, as no machine has serial ports anymore. Without those adapters, there is no way to control the telescope through the computer. Also, the motor of the telescope is fairly loud and can be quite annoying over time. The use of this device for demonstration on the final year exhibition starts to become questionable.

Telescope Presentation

For the Final Year Exhibition in the Cork School of Music, a small space will be set up to show and demonstrate the possibilities and the functionality of the installation at Blackrock Castle Observatory. For this purpose, a small optical telescope with motor control will be connected to a computer and possibly the Microsoft Kinect, to show the idea behind the Kinect-Telescope interaction. The telescope used will be a MEADE ETX 70.

Since the exhibition is inside, it won't be possible to actually look through the telescope. However, the telescope is almost constantly moving anyway and only suits a demonstration purpose. The motor runs on batteries, so sufficient batteries have to  be purchased before the start of the exhibition.

Projection test at BCO

First testing where made today, to find possible spots inside BCO to project on. A possible room could be a large round one at the end the castle. After experimenting with the projector, the possibility of a ceiling projection came up.
Unfortunately, the claw for the ceiling montage of a projector is currently not available, so an alternative has to be found. However, those issues will be addressed after the final place for installation inside the castle is found.

The first test show, that with the projector available, it will be very hard to create an immersive space, as the lens ratio of the projector is casual and the ceilings in Blackrock Castle fairly low.

Equipment updates

A new MacPro was gained for the project. Ubuntu 12 was installed and Stellarium and Processing runs excellent. Two video outputs are available, so enough computer resources should now be available to set up the installation.
Additionally two old eMacs could be gained from Crawford College, as they where going to dump them anyway. The screens can be used for possible instructions, webcam feeds or data visualizations.

For the projections however, only one casual projector could be gained. An Epson wide angle lens projector from Crawford College was not available. With only one projector, the preferred way of visualizing the telescope data will not be possible, so quite some work on visualization sketches in Processing cannot be used now for the project.

Equipment List

Here's the equipment list of what is needed to implement the installation in Blackrock Castle Observatory and to set up a space at Sullivan's Quay. Detailed updates of what of which of those components will be available and detailed descriptions of technical specs and brands will follow when the available equipment is received in a few weeks.

a Mac Pro

possibly 3x Projectors

• Epson EB410We Projector + cables (from Crawford College of Design)
• 3M 62x Projector + cables
• any small projector + cables for Sullivan's Quay

16:9 Widescreen Monitor + cables (like the Dell ones from B120)

any iMac for Sullivan's Quay

5.1 Surround Sound System for BCO and any Stereo Sound System for Sullivan's Quay + all required cables

probably 2x RGB to DVI adapter (or vice versa)

Router (with wireless if possible)

3x Ethernet Cables

2x Power plug Multicore

Allocating calculation resources

As the project won't run on any single machines I have at the moment alone, different programs and tasks have to allocated to the different machines available.
To do so, CPU and RAM usage of the different programs on different machines and operating system have been review, to get the best combination and allocation of computer resources for the tasks, that have to be performed. The machines, available at the moment are a Dell workstation (Linux and Windows XP) and a Sony Vaio sub-notebook (Windows 7) (also see here).
The results are the following:
When it comes to Stellarium, it obviously runs best on the Dell workstation and a lot faster on Linux than XP, which is not really surprising since Stellarium is a Linux project. Dell and Linux would be the preferred setup for executing Stellarium.
The processing sketch for visualizations that will be projected onto Stellarium run quite fluent next to Stellarium on the Dell workstation with Linux, however it quite drives it to its limits. But the complete graphical output of the floor projection could be fluently executed by the Dell workstation, it should be its only responsibility though.

The processing sketch for accessing the Microsoft Kinect runs with the openNI drivers on Windows at the moment (Linux will be probably be tested in the future as well). It doesn't require any graphical calculation power and runs just fine on the Sony Vaio notebook. However, body scanning like e.g. searching for the user's hand can take more time, when less resources are available. A fluent interaction is vital for the project, so when the Vaio notebook will be responsible for the Kinect interaction, it should possibly not be given any other tasks.

The animated poster processing sketches run on all the machines and but require a lot of resources. It won't be possible to run that sketch next to Stellarium in the Linux Dell machine and neither on the Sony Vaio notebook together with the Kinect sketch, without having either a non-fluent visual output or interaction.

The animated poster has to run on a separate machine, any OS would be fine. It does run fluent on all the iMacs in the CIT multimedia labs when executed alone. Getting one of those to display the poster is an option.
After reviewing all the CPU and RAM usage, it can be concluded that basically one more machine is required to perfectly execute the project. A perfect addition to the current machines would be a single high spec desktop machine, like on of the Mac Pros from the Multimedia Department of CIT. The Dell Workstation and a fast Mac Pro with both running Linux should provide sufficient calculation power to run the complete project.
The Sony Vaio Notebook could be used for working on the run or necessarily for the Sullivan's Quay exhibition. As it is just a sub-notebook, it is probably better not to use for a large museum installation.

A full equipment list of what is needed for the installation at Blackrock Castle Observatory and the Final Year's Exhibition at Sullivan's Quay will be posted soon.

Possible Projector

After review a long list of different projections, CIT and Crawford College have in their stock, one certain projector stood out for having a very suitable lens ratio for close projections, since the ceiling in Blackrock Castle is quite low. It is the Epson EB-410We.

The lens ratio is between 0.48-0.65, which is quite remarkable. This means, that at a 2 meter distance, the projected image could still be about 4 meters large. It is not surprising that this particular projector is the most expensive one on the list. Here's the product brochure.

The major focus now has to be, to get a hold on one of those projectors. In total Crawford College has two of them available. If both could be gained for the project, the best possible technical setup for the projection would be achieved. With only one of them, the problem of overlaying two projections with different lens ratios will come up. Since the second image of the visualized data projected onto Stellarium would be the smaller one in that case, ways of eliminating the black background of this projection will have to be found, to avoid having a light black square on the large projection.

http://www.epson.co.uk/Projectors/Epson-EB-410We/Overview

New PC & projector

I managed to get a Dell PC and a small Dell projector over the easter break (see also: https://twitter.com/#!/clemens_anzmann/status/185779172096999425/photo/1). Thanks a lot to Paul Green for this!
The PC will probably be the same machine, the final project will run on in Blackrock Castle, so possible problems that may occur on that particular machine can be found early.

Also, with the projector over the easter break I can finally start doing user studies with a projection on the floor and the Microsoft Kinect. This is very important to improve the interaction procedures of the final piece.

Working with projectors

During a workshop in the Crawford College of Design, different ways of projections where tested for the installation, such as projecting on tissue and overlaying several projected images.
The major goal was, to see the result of a processing sketch (witch a black background) being projected onto a projection of Stellarium. Here are the results:

As for now, a mainly black projection does not seem to affect a projection of Stellarium much. So projecting different ways of visuals for hydrogen onto the projection of Stellarium does not appear to become a problem, as long as the background of the visuals is black.