Skip to end of metadata
Go to start of metadata

Pie Chart 1

The following sonification sketches are based on the attached sketch to explore different ways of sonifying a Pie chart. Unlike the Early Sound Sketches, these ones are made based on a real pie chart and the timing of each value is very close to an accurate measure. A legend has been added to the piece. In the first sketch, the timing and values have 1:1 proportion, however, the timing has been cut into half for the rest of the sketches. Before listening please start with low volume and then adjust based on your preference. The goal is to spot the dominant Operating Systems market shares by listening to the following sketches.

Sonification-PieChart-Sketch#1-Legend.mp3

Sonification-PieChart-Sketch#2-Legend.mp3

Sonification-PieChart-Sketch#3-Legend.mp3

Sonification-PieChart-Sketch#4-Legend.mp3

Sonification-PieChart-Sketch#5-Legend.mp3

Received Feedback:

  • #1 and #2 -both of these make the relative proportions really clear. Prefer not to hear a pause between the segments, since you’ve got the bell indicating the start of a segment. Remove the Bell sound all together and let the segments flow to each other. It would be easier for user to listen to segments in order.
  • #3 - Difficult to differentiate between the individual sounds.
  • #4 - For the shorter/smaller segments, it may be difficult to hear it over top of all the other sounds. But it is really interesting to see how it describes the relative proportions between the segments as well as each segment relative to the whole. It would be easier for user to listen to segments in order.
  • #5 - This one really confusing - Couldn't tell why some segments overlap and others don’t, and in general couldn’t figure out what was happening.

Pie Chart 2

Pie Chart 2 sketches have been edited based on the feedback received from the team. The following changes have been applied:

  • Sketch#1:
    • The segments are ordered from large to small to make it easier for listener to spot the more dominant segments.
    • The bell sound indicating start of a segment has been removed. The moment of silence between segments has been removed.
  • Sketch#2:
    • The segments are ordered from large to small.
    • The bell sound indicating  start of a segment has been removed.
    • The double bell sound that indicates looping is replaced by single bell sound.
  • Sketch#4:
    • The Highlighted pieces are ordered from large to small.
  • Sketch#5:
    • sounds are split into smaller pieces and are more distributed throughout the piece.

Sonification-PieChart-Sketch#1-Legend-ordered-no bell.mp3

Sonification-PieChart-Sketch#2-Legend-ordered-no bell.mp3

Sonification-PieChart-Sketch#4-Legend-ordered.mp3

Sonification-PieChart-Sketch#5-Legend-Edited.mp3

Pie Chart 3

Round 3 sketches were developed based on the previous rounds’ sketches, but other features were also added. Different sounds and legend styles were used in order to reproduce the following pie chart:

Sketch #1-R3 – Audio legend is placed in the beginning; different sounds were used to represent each section, with the duration being a metaphor for size; nonstop play back.

Sonification-Sketch#1-R3.wav

Sketch #2-R3 – Audio legend is placed in the beginning of each section; different sounds were used to represent each section, with the duration being a metaphor for size; sectioned play back.

Sonification-Sketch#2-R3.wav

Sketch #3-R3 – Audio legend is placed in the beginning; only one sound is used to represent the market share, and the sections are separated by a bell sound; nonstop play back

Sonification-Sketch#3-R3.wav

Sketch #4-R3 – Audio legend is placed in the beginning; only one sound is used to represent the market share, and the sections are separated by a bell sound; sounds to indicate ten units and units were added as a metaphor for numbers; nonstop play back.

Sonification-Sketch#4-R3.wav

Sketch #5-R3 – Audio legend is placed in the beginning of each section; different sounds were used to represent each section, with the duration being a metaphor for size; sections are presented in pairs to help compare them; sectioned play back

Sonification-Sketch#5-R3.wav

Sketch #6-R3 – Audio legend is placed in the beginning of each section, but is considered part of the sonic chart; only one sound is used to represent the market share, and the sections are represented by echoed legend; a sound was added to represent ten units; the market share sound gets faster as the sections gets smaller; nonstop play back

Sonification-Sketch#6-R3.wav

Sonification Sound Tracks for Pie Chart Authoring Tool

The following sound tracks can be used to create the sonic pie chart. They are kept under 4 seconds long and made ready for looping.

2.wav

3.wav

4.wav

5.wav

6.wav

7.wav

8.wav

9.wav

10.wav

11.wav

12.wav

13.wav

Sonification Stacking Experiments

Some early and rough experiments with stacking sounds.

Strings stacked left

Strings stacked right

Strings stacked center

Experiments with Percussive Sounds and Legend Placement

Stacked percussive sounds (right justified), no legend

Stacked percussive sounds (right justified), with legend

Sequential percussive sounds with integrated legend

Sequential percussive sounds with integrated legend and "backdrop" sound

Sequential percussive sounds with legend up front

Sequential percussive sounds with single "backdrop", no legend

Some thoughts:

  • how can some of these be combined in different ways to offer the user a fuller experience of the chart/data?
  • are the pauses between percussive sounds too long? what level of accuracy do we need in terms of correlating duration with values?
  • the "backdrop" sound which indicates the full duration of the chart is quite long when it runs for the total length (sum of all parts) - in this example I have shortened it arbitrarily but made it consistent in each part such that it provides an indication of relative length of each section to total
  • if the legend is integrated, is it better to use the same sound for all parts? need to consider consistency if different presentations are combined (i.e. maintain unique sound for each section)

Single sound

Single sound, integrated legend (ordered largest to smallest)

Single sound, integrated legend (duration to value).mp3

Single sound, integrated legend (frequency to value).mp3

Single sound-density to value-integrated legend.mp3

Density

Density-to-value-mapping zoomed out

  • this sketch maps a somewhat arbitrary density value to the segment value
  • density could be mapped according to: number of "beeps" (directly proportional to) value/10 - or perhaps the denominator could be user-configurable and adjusted with a "preview"
  • minimum number of beeps = 1 (for multiple segments with small values, this could be solved by e.g. lumping them together as "all others", then perhaps allowing the user to zoom in on sections of the data to get more detail)

Density-to-value-mapping zoom in 1X

  • this sketch takes the previous sketch and "zooms in" on the sound
  • here, the total duration of each segment remains the same (while the density decreases / total number of beeps decreases)
  • this could also be achieved by "stretching out" the data such that the total duration increases - this would probably be a more intuitive approach for the user

Density-to-value-mapping zoom in 2X

  • this sketch takes the previous sketch and "zooms in" on the sound again
  • here, the total duration of each segment remains the same (while the density decreases / total number of beeps decreases) 
  • this could also be achieved by "stretching out" the data such that the total duration increases - this would probably be a more intuitive approach for the user

Density-to-value-mapping zoom to units

  • this sketch doesn't actually map density, but could provide a way of "zooming" in further on the density sample to allow the user to accurately count the units
  • maps one beep-like sound to units of 10 (G) and same lower pitch sound (F) to units of 1
    • each 10-unit sound is approximately 1/8sec long, each beep separated by approximately 1/4sec
    • each 1-unit sound is approximately 1/24sec long, each beep separated by approximately 1/8sec
  • unit sounds are representational, and do not relate in a durational way (i.e. sound for 1-unit is not 1/10th the length of the sound for 10-units)
  • total duration of each segment is consistent (includes pauses at end of any section where necessary)
    • (note: because, for example, 6 1-units have a longer total duration than one 10-unit, I decided to keep the total duration of each section the same, since total duration of the sounds does not map directly to value)

Density-to-value-mapping zoom to units octave change

  • this sketch is exactly the same as the previous sketch with an octave shift
  • provides and example of a possible user-configurable pitch level

Stacked Right with Units - Pitch and Timbre

  • duration of segment is mapped to value (duration of each segment calculated as X% / 10 sec)
  • segment sounds are "stacked right" such that they come in a different points but end at the same time
  • a click track is also added here to allow unit counting (clicks every 1/2 sec or 5%)

Stacked right, units, pitch + timbre

  • different instruments playing notes of different pitches represent each segment

Stacked right, units, pitch only - strings

  • same instrument playing notes of different pitches represents each segment (strings)

Stacked right, units, pitch only - clarinet

  • same instrument playing notes of different pitches represents each segment (clarinet)

Stacked right, units, timbre only

  • different instruments playing the same note represent each segment

Combinations - Stacked + Sequential

to do!

 

  • No labels

10 Comments

  1. Murilo Crivellari Camargo

    The Round#3 sketches are very interesting. Great job:)

    Sketch#1: I like how easily different sounds are being recognized in this piece. However, it still has the problem of forgetting the legend when you get closer to the end of the chart.

    Sketch#2: This is a very interesting approach. I think you could just use the same sound, because here only the duration matters and when you are separating different sections you want the user to be able to compare the length of each section and hearing different sounds may distract the user.

    Sketch#3: Here first we need to map a number to a section and then remember what that number represented. If user takes notes, this is a very practical approach, if doesn't, probably user has to replay the legend a few times to map the numbers to sections.

    Sketch#4: This is my favourite!!! It's so easy to recognize how large each section is and it also conveys a second layer of information which enables user to approximately calculate a number related to that piece. So far, I think this sketch represents a pie chart better than all the other sketches that we have developed.

  2. Murilo Crivellari Camargo

    Hey great job coming up with all these different variations!

    1: definitely found it difficult to remember the legend near the end, but it's easy to get a feel of how small/large the different pieces are in relation

    2: I like this one the best, I think it easily differentiates between the sections but still lets you compare by feel and I think the different sounds helps in distinguishing the parts from one another

    3: I find the addition of numbers corresponding to the legend is still hard to remember, and simply adds more cognitive load for the user. I agree with sepi, they'd have to take notes to remember

    4: The use of units is nice but I find it too rhythmic and it's too easy to start bopping away to the beat as opposed to listening to the sonic chart for information

    5: Pairs could be nice for comparison but then it feels more like multiple bar graphs being compared and it may start getting too long

    6: The echo is distraction imo but I like the combination of sounds here better as it didn't get too beatsy, also the units sound was great. Wondering though how you would represent a more specific unit? not sure if introducing another sound would make it too messy

  3. Hey, I liked Sketch 4 and 3.1 the best.  They were simple and easy to follow.  Although I liked 3.1 for its simplicity and not being too busy, I'm still leaning towards 4 even though it's a little busier because it gives you the actual percentage for each category.

  4. Hi guys, I just checked out Dana's 3 samples and I was able to understand the "strings stacked right" sample the best.  I don't know why the stacked-left sample was harder for me to decipher, but for me it was easier to distinguish the strings coming in on the stacked-right sample.  Hope this helps, and thanks.

  5. Hi guys, sorry I took so long to check this out.  My preference for the percussive experiment is the second one, the stacked right-justified one with legend.  I think this one will also give the user an experience that's pretty close to what you guys experience when you see a visual pie chart.  The one with the integrated legend and backdrop sound was good too, as long as the backdrop sound is a consistent length for each section (so the user gets a better idea of how much larger one section is compared to others).  Thanks, and take care.

  6. Hey Dana, I just checked out your new file.  I don't think the single sound is as effective, at least not when it's a continuous sound like a flute.  If it was a percussive sound with differing numbers of beats per section then it would be a lot easier to figure out the differences in duration.  With the flute I found it harder to figure out the differences in section duration, especially for the first 2 sections.  Thanks, and take care.

  7. Thanks for your feedback Maria! Regarding your second comment, we were actually just discussing something like that - breaking up the sound to represent units instead of having one continuous sound. I think Sepideh is going to try this and then we'll get it posted to the wiki.

  8. Hey Dana, I checked out the 2 single sound, integrated legend files you added.  The Frequency one didn't do much for me, but the duration to value one was a little better than the first single-sound file.  The difference between section duration is still a little hard to figure out, but if you separate those little notes just a little more I think it'll work fine.  Thanks, and take care.

  9. Hey Dana, thanks for the new round of sketches.  Out of all of them I think the last 2 are the best ("Density-to-value-mapping zoom to units" and "Density-to-value-mapping zoom to units keychange").  Personally I don't have a preference between the 2, but those 2 are easy to figure out.  Thanks again, and take care.

  10. Hey guys, I've checked out the pitch & timbre files, and for me they're not as informative as the "Density-to-value-mapping zoom to units" or "Density-to-value-mapping zoom to units keychange".  But that's just my take on it, other people might find it useful.  Thanks for all your hard work.