Some movies have reputations in terms of music, in other words, original soundtracks (OST) of a film. In this regard, comparing the popularity of a film with the popularity of its OST could be interesting. To narrow down a data range, I decided to focus on the works of Hans Zimmer, who is a renowned film composer and has received a range of honors and awards, and compare the checkouts of OSTs, in which Hans participated or composed, and the checkouts of their original films from the Seattle Public Library dataset.

Due to a lack of a data field for authors in SPL dataset, I searched 'Hans Zimmer' in the SPL website (https://goo.gl/MyxFH6) and filtered to see only music CDs. The list includes not only his solo composition works but also works in which he partial contributed by writing only several songs. Based on the search result, I chose main movies for my visualization and queried as below.

SELECT
    YEAR(cout) AS Year,
    MONTH(cout) AS Month,
    SUM(CASE
        WHEN (title LIKE 'Blade Runner 2049%soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Blade Runner 2049 OST',
    SUM(CASE
        WHEN (title LIKE 'Blade runner 2049%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Blade Runner 2049',
    SUM(CASE
        WHEN (title LIKE 'Dunkirk%soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Dunkirk OST',
    SUM(CASE
        WHEN (title = 'Dunkirk' and bibNumber = '3276118') THEN 1
        ELSE 0
    END) AS 'Dunkirk',
    SUM(CASE
        WHEN (title LIKE 'Hidden figures%original score%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Hidden Figures OST',
    SUM(CASE
        WHEN (title LIKE 'Hidden Figures%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Hidden Figures',
    SUM(CASE
        WHEN (title LIKE 'Kung Fu Panda 3%music%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Kung Fu Panda 3 OST',
    SUM(CASE
        WHEN (title LIKE 'Kung Fu Panda 3%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Kung Fu Panda 3',
    SUM(CASE
        WHEN (title LIKE 'Interstellar%soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Interstellar OST',
    SUM(CASE
        WHEN (title LIKE 'interstellar%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Interstellar',
    SUM(CASE
        WHEN (title LIKE 'Son of God%Motion Picture Soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Son of God OST',
    SUM(CASE
        WHEN (title LIKE 'Son of God%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Son of God',
    SUM(CASE
        WHEN (title LIKE 'Man of Steel%soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Man of Steel OST',
    SUM(CASE
        WHEN (title LIKE 'Man of Steel%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Man of Steel',
    SUM(CASE
        WHEN (title LIKE 'Inception%Music%Motion Picture%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Inception OST',
    SUM(CASE
        WHEN (title LIKE 'Inception%' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Inception',
    SUM(CASE
        WHEN (title LIKE 'Sherlock Holmes%motion picture soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Sherlock Holmes OST',
    SUM(CASE
        WHEN (title = 'Sherlock Holmes' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Sherlock Holmes',
    SUM(CASE
        WHEN (title LIKE 'Madagascar%motion picture soundtrack%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Madagascar OST',
    SUM(CASE
        WHEN (title = 'Madagascar' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Madagascar',
    SUM(CASE
        WHEN (title LIKE 'Madagascar%2%Music%Motion Picture%' and itemType LIKE '%cd')
        THEN 1
        ELSE 0
    END) AS 'Madagascar Escape 2 Africa OST',
    SUM(CASE
        WHEN (title = 'Madagascar Escape 2 Africa' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Madagascar Escape 2 Africa',
    SUM(CASE
        WHEN (title LIKE 'Madagascar 3%Music%Motion Picture%' and itemType LIKE '%cd') THEN 1
        ELSE 0
    END) AS 'Madagascar 3 Europes Most Wanted OST',
    SUM(CASE
        WHEN (title = 'Madagascar 3 Europes most wanted' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS 'Madagascar 3 Europes Most Wanted',
    SUM(CASE
        WHEN (title LIKE '12 Years A Slave%Music%Motion Picture%' and itemType LIKE '%cd')
        THEN 1
        ELSE 0
    END) AS '12 Years A Slave OST',
    SUM(CASE
        WHEN (title = '12 Years A Slave' and itemType LIKE '%dvd') THEN 1
        ELSE 0
    END) AS '12 Years A Slave'
FROM
    spl_2016.outraw
GROUP BY YEAR(cout) , MONTH(cout)
ORDER BY YEAR(cout) , MONTH(cout)

Duration : 463.161 sec

The main visual concept is to combine a circular timeline of an item's release date with a linear timeline of an item's checkout. This is because my main purpose in visualization is to see the trend of checkouts not its amount in detail since an item is issued and to compare the trend of two items, the OST and DVD of a film. For comparing two items, OST's trend is flipped in terms of the y-axis of a linear timeline, in other words, values reflecting checkouts, and placed DVD and OST's data together along with the x-axis of a linear timeline. And a film's checkout timeline starts at a certain point of a circular timeline which means the earliest released date among the film's OST and DVD. The length of an arc from the point of a released date to the end of a circular timeline is the same with the length of each film's linear timeline from the point.

Because of the length relationship between a circular timeline and linear timelines, placing visual objects are somewhat limited in a space. It requires an exact calculation for a radius of the circle and length of a timeline unit segment representing a month. And it has an overlapping issue among linear timelines because of a released date-based visualization approach. Visualizing checkout values as a bar graph, each checkout value is normalized based on the maximum checkout of an item. What I expected in this relative bar graph is to see when an item is the most popular in time and how gradually the popularity changes. Also, by placing a DVD's bar graph and OST's bar graph in a flipped direction, we can compare two items' trend together. A visual theme I intended is a bar-shaped waveform (https://goo.gl/XFHCmS).

In terms of color scheme, I differentiated a bar graph's main color according to an item for comparison. A red color represents OST checkout data and a white color is for DVDs. And its saturation is proportional to a bar's height which means the stronger color is the more popular an item is at the moment. But, I will do more modification and experiment in color for better visual aesthetic.

Although each item has a small bar graph, it is enough to discern the trend of checkout and to compare two items of an OST and DVD. In general, it shows the similar checkout trend between OSTs and DVDs. Usually, an OST is issued earlier than its DVD. This fact also can be confirmed by watching the first bar of a film's OST graph which exists prior to that of a film's DVD. Also, the popularity of the OST has a peak right after the issued date and gradually decreases even after the DVD is released.

All work is developed within Processing 3.3.6
Source Code + Data