Using the ECEM demonstrator at http://ecem.wemcouncil.org, as supplied by Lucy Haughey, I have been doing some initial tests today to get a sense of what the data might sound like presented as it is from the Demonstrator. So far I have done 3 tests exploring the variables in the data set to a certain extent.
Originally I tried using the whole browser screen as a sample, but quickly found that the map was quite small in the sampling screen and therefore was not creating much in the way of sound across the range of bells. So I have created a series of screen recordings using Quicktime so I could focus on just the area where there was some change in colour.
The part of the image that is most useful here is in the centre and I have mostly focused on this in the tests.
A preliminary test to see how the movie capture system worked. I did not properly capture the data I was playing as I wanted to check that BellHouse would play it reasonably first, but it was loosely playing information about air temperature on a monthly basis from 1979 to the present.
This is a short excerpt of the sound of BellHouse playing this movie.
As you can see and hear, the variation in this video is pretty wide and across the map of Europe. This means that BellHouse is playing most of the bells on a regular basis. BellHouse has 2 main variables, the first is the “sampling rate” (how often the software will sample the colour change) and the “sampling threshold” Here BellHouse is set up to sample every 2 seconds and with a high threshold of 9.
The end result is that, because of the significant colour changes across the map most if the time, it plays most of the bells most of the time. Also, we can see that because the demonstrator is playing changes every half a second or so, BellHouse is only able to record one in every three or so of the changes.
One thought that strikes me here is that it would be useful to be able to find out if there are specific datasets that the scientists here would like BellHouse to play as there are so many variables to explore. The following 2 tests are to explore the range of variables and see if there are ways to explore how BellHouse might play them.
In this test I have chosen these parameters:
Countries, Historical, Climate, Snow Depth, Monthly, Absolute Values, Jan 1979 – Dec 2016. This is recorded as Quicktime movie.
In this video there is a lot less colour change which makes for a quieter sound but there are some interesting rhythms set up as the Nordic, Alpine and Eastern European countries change seasonally.
Here you can hear that this sets up a nice rhythm on BellHouse which, because of the seasonal and geographical nature of the snowfall, means that only a few of the bells are played in this recording with an occasional new sound coming out. However quite a few bells are outside of the video and so will never register as the format is wrong for the set up of the grid in the sampling software. So as an experiment I flipped the frame to horizontal to see what that would sound like.
In these early tests we don’t yet have the capacity to show both BellHouse and the browser at the same time, but we aim for that once we have practiced and explored the data more fully.
The parameters used in this test were: Clusters, Wind speed, 10m, Monthly, Absolute Values, Ensemble mean, REP 8.5, 1979-2030.
The change in the variables to windspeed has created a much more detailed breakdown of the map with many more areas of changing colour. This is giving us a much more nuanced effect and it is interesting to see how this affects the way that BellHouse plays the changes.
There is still an issue about how quickly the information is played and I feel that it might be useful here to edit this in some video software in order to slow it down so that bellhouse can react to each colour change.
- There are clearly a huge amount of variables that we can play with here. What would be important and useful to explore?
- The speed of the transitions in the animation could be slowed down to see what effect that would have on BellHouse playing the data more comprehensively.
- Making the video screen and the sampler fit is a useful thing to work on, though part of being playful with the data is that even when flipped the pattern is similar enough in how it plays, it just plays different bells.
- Being able to video the whole recording playback in better quality is something to do and should be up an running next week.
- Often when there is less going on, such as in Test 2, there is more to listen to. If the data change is too diverse for too long, it is harder to identify patterns.
Playtest 2 28th October
Following discussions with Lucy Haughey, the Communications Officer at the WEMC, I spoke to Professor Alberto Troccoli, the MD and co-founder of the WEMC in Australia to discuss the use of the new WEMC demonstrator, TEAL. We discussed a variety of ideas about how we should explore some of the variety of information the the TEAL model can demonstrate.
In this new series of playtests, of which there are 5, I have decided to focus initially on Europe exploring temperature, wind speed and precipitation. The last 2 recordings feature a more global perspective, exploring temperatures across Europe, Africa and Asia,
By way of an introduction to the TEAL demonstrator please see the WEMC website (https://www.wemcouncil.org/wp/projects/c3s-climate-and-energy-education-demonstrator/)
This is a publicly accessible tool for educational purposes which is a step forward from the previous demonstrator in it’s simplicity (despite a similar quantity of information available) and user experience.
This first recording is of European average temperatures by country measured seasonally from 1979-2020. In the settings for this I have pushed the sampling threshold up to 20 as the changes of colour would otherwise cause nearly all of the bells to ring all of the time.
The passage of the data in this film is over 5 minutes and so there is no loop in this film. However there is a rhythm in this recording which reflects the more marked changes in average temperatures in the eastern half of Europe and also of Turkey. There is a tonal pairing of higher and lower tones which seems certainly in the later half of the recording to keep reappearing which makes this really interesting to hear and reflects strongly the seasonal changes in the data.
Recording 2 – Windspeed Europe Seasonal 1979 -2020
The colours in this film, green, are of much less contrast and so the sampling threshold is set much lower to pick up the changes.
Recording 3 – Seasonal European Precipitation: 1979 – 2020
In this recording and the accompanying animation, most of the change, and therefore the greater proportion of the sounds, is in the middle columns of the sampler and therefore across the middle of the BellHouse. I think this is denoting a greater variation in precipitation in the countries with a more continental influence where there might be said to be a higher contrast in rain/snowfall on a seasonal basis.
Recording 4 – Annual Global Temperature 1979 – 2020
The sounds of this recording are much quieter as viewed globally, the annual temperature changes are less marked than in the previous recordings. There are some quiet spaces and again the countries in this animation are smaller and so do not fill the same space in the sampler..