In the last article on the blog, we covered how Cyanite’s Similarity Search can be used in music catalogs. In this article, we explore another way to search for songs using Dynamic Keyword Search and how to leverage it to create mood- and contextual-based playlists. 

Rather than relying on a reference track, Dynamic Keyword Search allows you to select and combine from a list of 1,500 keywords and adjust the impact of these keywords on the search. This is especially helpful to create playlists where songs match in mood, activity, or other characteristics. 

But before we explain how this feature works, let’s explore how playlists are created. What makes a perfect playlist? Why are playlists so essential when utilizing a music catalog? And how can the Dynamic Keyword Search help with that?

There are three techniques for playlist creation:

1. Manual creation (individually picking songs) 
2. Automatic generation and recommendation 
3. Assisted playlist creation. 

Historically, manual creation has been the most basic and old approach. It involves picking songs individually for playlists. It might be the simplest technique but the amount of time and effort that goes into it can be overwhelming. Imagine you are working 100,000 audios in a catalog and have to create an “Energetic Workout” and “Beach Party” playlist. 

Automatic generation uses various algorithms to create playlists with no human intervention. One of the most famous ones is, for example, “Discover Weekly” by Spotify. 

Assisted playlist creation uses music technology to guide and support manual playlist creation. 

In the research by Dias, Goncalves, and Fonseca, manual playlist creation was found to be most effective in terms of control, engagement, and trustiness. This means that people trust handmade playlists. Also, manual creation provides the most amount of control over the outcome and it engages editors in the creation process. 

Automatic creation was found to be the most effective in adapting to the listeners’ needs. There is no manual control involved, so automatic tools can adapt and change playlists in no time. 

Assisted techniques were found to be most effective in terms of engagement and trustiness whilst being quick to create. They also performed well on the song selection criteria. Song selection has been defined as the most critical factor in the playlist creation process according to this study. However, while song selection is considered very important, the question of what makes a song right for the particular playlist is still open. Apart from that, assisted techniques proved to be optimal in control, and serendipity and they also can adapt to listening preferences rather easily. 

To anticipate things already: The Dynamic Keyword Search is exactly such an assisted technique in playlist creation.

Playlists have been known to be the ultimate tool for promoting music. We already covered the ways artists can get on Spotify and other people’s playlists in other articles on the blog. But creating playlists can also be beneficial for catalog owners and catalog users, be it professional musicians or labels. Here is why: 

• You can realize new and passive modes to exploit and monetize your catalog. If you make it easier for your users and/or customers to explore your catalog, you directly increase its value.
• Playlists are used as a promotional tool to showcase the works of an artist or the inspirations behind the artist. This article recommends creating two playlists: a vibe playlist and a catalog playlist for brand engagement and streams. 
• Playlists help organize music by theme or context. 
• With playlist creation features, users save time on finding the right fitting songs
• Playlists can be indexed separately in search results. This helps music get discovered. 

So playlist creation tools in a catalog are pretty important. Similarity Search is one of these tools. Another one, which we focus on in this article is Dynamic Keyword Search.

Cyanite’s Dynamic Keyword Search allows for searching tracks based on multiple keywords simultaneously where each keyword can be weighted for its impact on the search. This feature leads to more relevant search results with less time-effort spent on search.

Usually, the keywords you choose represent your idea of what you’re searching for. But you don’t have full control over the search. With Dynamic Keyword Search, you can increase the precision of the search results by adjusting the impact of the keywords on the search. So you can express exactly what you’re looking for. There are 1,500 keywords to choose from representing such characteristics of the song as mood, genre, situation, brand values, and style. These keywords’ impact on search can then be adjusted on the scale from -1 to 1 from no impact at all to “heavy impact”.

Cyanite Dynamic Keyword Search interface

Not all playlists are created equal. Some are better than others. This study outlines 5 characteristics of playlists that can indicate a good or bad playlist. The authors of the study assumed that user-generated playlists could be an indicator for the algorithms to create good playlists. Here are the 5 playlist characteristics they outlined: 

• Popularity – most user-generated playlists feature popular tracks first. This, however, is not too obvious though but grabbing the attention spans of the listeners from the start is important. 
• Freshness – playlists should contain recently released tracks. Most playlists in the study contain tracks released on average in the last 5 years.
• Homogeneity and diversity –  playlists on average cover a very limited number of genres so playlists should be rather homogenous. However, diversity plays a significant part in listeners’ satisfaction so it should be incorporated into the playlist as well.
• Musical Features – in terms of energy, playlists with a narrow energy spectrum with a low average energy level are preferred, but there can be some high-energy tracks in the list. 
• Transition and Coherence – the similarity between the tracks defines the smoothness in transition and coherence of the playlist. Usually, user-generated playlists have a better similarity in the first half and a lesser similarity in the second half. 

As the study deals with a variety of user-generated playlists, it can’t be said that all of them were equally good playlists. But the criteria outlined above can help improve playlists by understanding the character of the playlist. With Dynamic Keyword Search, you can control such criteria as homogeneity and diversity, musical features such as energy level, and similarity between the tracks to ensure transition and coherence. 

PRO TIP: To improve a playlist’s transition and coherence you can combine the Dynamic Keyword Search with our Similarity Search to further filter music on Camelot Wheel. The Camelot Wheel indicates which songs transition harmonically well giving you an extremely powerful tool to perfect the song order. You can find a deeper explanation of that in this article.

Here is how to access Dynamic Keyword Search in the Cyanite app. This feature is also available through our API. 

1. Go to Search in the menu and select the Keyword Search tab. Choose whether to display results from the Library or Spotify. 
2. Select keywords from the Augmented Keywords set. For example, these are some of the keywords in the list: joy, travel, summer, motivating, pleasant, happy, energetic, electro, bliss, gladness, auspicious, pleasure, forceful, determined, confident, positive, optimistic, agile, animated, journey, party, driving, kicking, impelling, upbeat. We recommend selecting up to 7 keywords out of 1,500. 
3. Adjust the weights for each keyword from 1 to -1 to define their impact on search. For example, let’s set  the search input as sparkling: 0.5, sad: -1, rock: 1, dreamy: 1 
4. Scroll down for search results. The search results will return tracks from the library that are dreamy, slightly sparkling, and not at all sad. They will also all be rock songs.

Dynamic Keyword Search can be requested from our support team.

There are various ways to create playlists from manual creation to automatic and assisted techniques. An assisted approach that combines automatic and manual creation has proved to be the most effective in playlist creation. It meets almost all the editors’ needs such as providing control over the process, maintaining a high level of engagement and trustworthiness, and offering a good selection of songs. However, the automatic approach is fast developing and algorithms might substitute human work completely in the future. 

Our Dynamic Keyword Search feature can help you create playlists as one of the assisted techniques. It can provide search results that take into account the search intent  in terms of keywords and the impact of those keywords on search. This doesn’t mean that Dynamic Keyword Search replaces the manual work completely, but it can help artists, labels, and catalog owners do the creative work and engage fans and listeners with the support of the right tools to save time, money, and effort. This is what we’re striving to achieve here at Cyanite – to help you fully unlock your catalog’s potential.

Let us know if this article has been helpful and stay tuned for more on the Cyanite blog! 

I want to try Dynamic Keyword Search – how can I get started?

If you want to get the first grip on Cyanite’s technology, you can also register for our free web app to analyze music and try similarity searches without any coding needed.

Sources

From manual to assisted playlist creation: a survey by Ricardo Dias, Daniel Gonc¸alves, Manuel J. Fonseca

Effects of recommendations on the playlist creation behavior of users by Iman Kamehkhosh, G. Bonnin, D. Jannach

Other articles in the series include: 

How to Turn Music Data into Actionable Insights: This is an overview of the Data Pyramid and how it can be used in the music industry. 

An Overview of Data in The Music Industry: This article gives a list of all types of metadata in the music industry.

Making Sense of Music Data – Data Visualizations: This article explores data visualizations as the second step of the pyramid and gives examples of visualizations in the music industry. 

Benchmarking in the Music Industry – Knowledge Layer of the Data Pyramid: This article deals with Knowledge and how it is used to benchmark performance and set expectations.

Intelligence can be described as the ability to choose one decision out of a million other decisions based on knowledge of how these decisions might affect the outcome. 

Intelligence can be generated by the machine, for example, a self-driving car is a form of intelligence that scans the environment and can predict the course of action for the next section of the road. In the music industry, intelligent decisions are still, for the most part, made by humans by examining information, reading graphs and charts, memorizing past outcomes, and monitoring real-time data. In this article, we’ll explore some of the emerging intelligence technology in the music field so keep reading to find out more.

1. Recommendation systems that learn and adapt effectively to individual users’ preferences

Recommendation systems already use some sort of prescriptive analytics when they make a selection of songs based on past user behavior. Recommendation systems can also take into account the sequence of songs and context that affect the enjoyment level of the playlist as a whole. As previously played songs influence the perception of the next song, the playlist can be adjusted accordingly. The ability to prescribe a listening experience by recommendation systems is, perhaps, the most common and well-developed example of intelligence in the music industry.

Additionally, recommendation systems can prescribe music that directly affects user behavior. This project, for example, uses data from running exercises, predicts the future running performance, and recommends songs that maximize running results. It does so continuously, as the system stores and learns from each updated running exercise record.

To learn more about different types of recommendation systems, check out the article How Do AI Music Recommendation Systems Work. 

2. Automatic playlist generation based on context

Generating music or suggesting existing music based on the context is an analog of a self-driving car in the music industry. The music adapts to the listening situation to amplify the current experience. For instance in video games, where music adjusts to the plot as the user progresses through various levels of the game. More on that in our article on Omniphony engine that explores adaptive soundtracks and music context in game development.

Such systems are also used as location-aware music recommendations for travel destinations (when music is chosen based on the sightseeing place you visit), or computer vision systems for museum experiences (when the artwork dictates the audio choice). In these cases, the constantly changing environment serves as the basis for recommendations. 

Another example of intelligence in this field is generating music in the metaverse which is a virtual environment, that includes augmented reality. The metaverse can be accessed through Oculus headsets and a smartphone. Currently, virtual streams and concerts are already conducted in the metaverse, so it is only a matter of time till the curated immersive experiences that can adjust to the audience’s needs will be delivered using intelligence.

3. Prescriptive curatorship – What’s going to be hot next? 

Prescriptive curatorship entails an understanding of how up-and-coming artists and tracks will perform and who is more likely to break in the near future. In the past, platforms like Hype Machine indexed music sites and helped find the best new music. 

Nowadays, there are systems that can predict future hits and breaking artists automatically. For example, Spotify is developing algorithms that can predict future-breaking artists. The algorithm takes into account the preferences of the early adopters and then determines whether the artist can be considered breaking. This data can then be used to sign deals with the artist at a very early stage.

4. Tracking changes in music preference distribution  – making music that hits the current preferences or even future preferences

Unlike prescriptive curatorship that relies on a group of experts, music preference distribution numbers serve artists to show how their chosen genre and formats fit audience demographics and how music can be changed for current or future preferences. The general consensus in the music industry is that music preference algorithms come after the music is produced, otherwise all music will end up sounding the same to mimic popular artists. 

There is not yet a system that would automatically recommend changing the content of the song based on what users prefer. Nevertheless, attempts to use the numbers to create songs people will like are still being made.

5. Royalty Advances

Royalty advances are a complex task that requires comprehensive tracking of music consumption across all platforms. Distributors such as Amuse and iGroove offer a royalty advance service that is able to predict upcoming payout amounts so that artists can invest in their music long before the actual royalties kick in. These systems analyze streaming data to calculate upcoming earnings. 

Recently the topic got even more attention through the hype of NFTs. Crypto-investors want to predict future royalty payouts and the value of their asset. 

Future platforms most likely will be able to prescribe a course of action regarding which distribution platform to focus on based on the predicted royalty amounts. 

The main concern in the industry is how far can one go with technological intelligence considering that music is a creative activity and the human element is still largely prevalent. An intelligence machine that can tell which music to produce based on a prediction of future user preferences generally prompts an adverse reaction in the industry. 

Nevertheless, intelligent decisions to adjust the content of songs or to sign future-breaking artists identified by the AI can already be made by the artists and labels based on available data. 

At Cyanite, we provide our API for access to data and the development of any kind of intelligence engines. As always, at each level of the pyramid, the quality of data plays a vital role. Cyanite generates data about each music track such as bpm, dominant key, predominant voice gender, voice presence profile, genre, mood, energy level, emotional profile, energy dynamics, emotional dynamics, instruments, and more.

If you want to get the first grip on Cyanite’s technology, you can also register for our free web app to analyze music and try similarity searches without any coding needed.

This analysis example gives you a comprehensive overview of Cyanite’s auto-tagging scope. Whether you are thinking about integrating Cyanite into to your platform or just want to get a general idea of AI tagging in action, this data sheet is for you!

To show you how it works, we analyzed the New Music Friday playlist by Spotify (Germany). To get the data sheet, leave your email here.

How to Turn Music Data into Actionable Insights: This article is the first one in the series. It reviews all layers of the Data Pyramid and shows how to turn raw data into actionable insights. 

An Overview of Data in The Music Industry: This article presents all types of metadata in the industry from factual (objective) metadata to performance metadata. It is an essential guide for music professionals to understand all the various data sources available for analysis. 

Making Sense of Music Data – Data Visualizations: This article discusses the second step of the Data Pyramid – Information and how information can be presented in the form of data visualizations, making it easier to comprehend large data sets. 

At this step, attempts to look into the future and predict outcomes are made. The more specific the problem or context you’re observing is, the more precise your findings will be at this step. The Knowledge Layer produces analytics that help benchmark performance. As Liv Buli from Berklee Online University puts it, at the Knowledge layer you can tell that the artist of a certain size sells well after performing on the TV show and use this information to guide strategy for other artists of the same size. As a result, knowledge makes it possible to look at data in the industry-specific context and understand how you compare in relation to past successes and to competition. In that regard, benchmarking and setting expectations is the final outcome of the knowledge step. 

Benchmarking can take different forms within the music industry: 

Types of benchmarking 

• Process benchmarking

This type of benchmarking deals with processes and aims to optimize internal and external processes in the company. You can improve the process by looking at what competitors are doing or setting one process against another. Processes relate to how things are done in the company, for example, the process of uploading songs to the catalog. 

• Strategic benchmarking

Strategic benchmarking focuses on the business’s best practices which is often more complex than the other two types of benchmarking and includes: competitive priorities, organizational strengths, and managerial priorities. For example, an assessment of how fans responded to the brand sound in the past can help devise a long-term sound branding strategy.

• Performance benchmarking

Performance benchmarking compares product lines, marketing, and sales figures to determine how to increase revenues. For example, as a marketing campaign for a music release develops over time, it can reveal the most vital money channels for exposure. Sales figures can indicate how artists compare to one another in terms of profitability. 

1. Analyzing hit songs and popular artists to discover new talent

You can utilize Cyanite’s technology to analyze what’s currently working in the music industry in different markets. In particular, you can analyze popular songs and understand what makes them successful in terms of audio-based features such as genre, mood, energy level, etc. Further, you can use the Similarity Search to find tracks with a similar vibe and feel. It then helps you discover and identify new talent which may go along the same lines as current successes. Of course, that is not the whole story of making a hit but it gives you a pretty solid foundation of hitting the current zeitgeist.  

2. Analyzing popular playlists to predict matches

We specifically described this use case in The 3 Best Ways to Improve your Playlist Pitching with CYANITE article. You can analyze existing popular playlists such as Spotify New Music Friday or Spotify Peaceful Piano and see what songs are usually featured. This information can then help to understand the profile of the playlist which allows you to find the perfect fit and increase the chances of getting into the playlist. It also supports you in describing the songs to the editors in the way that song is accepted.  

3. Analyzing marketing and sales numbers to allocate marketing budgets and support event planning

For example, record labels can analyze the performance of their artists and adjust marketing campaigns accordingly. Insight can help change the direction of the marketing campaign, choose appropriate channels, and allocate marketing budgets more effectively. In events planning, you can analyze event venues and identify the most relevant cities and venues based on past events. 

4. Analyzing the artists’ styles to identify opportunities for song-plugging

If you’re a music publisher looking to get a placement on a new release of a successful artist, analyzing their previous style and matching it to your catalog of demos would be the way to go. 

In this case, some audio qualities of the song are not important, as it will be eventually re-recorded. To analyze songs, a regular Cyanite functionality can be used including the Keyword Search by Weights where you can search your demo-catalog by the analysis results of the successful artist on weight-specific keywords to get the most relevant results. 

5. Analyzing fan engagement to identify audience segments

You can also analyze artists’ performance by looking into fan engagement on social media and music platforms. Through understanding the fan’s demographics, interests, and lives, you can create custom audiences for new artists or deepen fan engagement for the same artist based on past campaigns. This use case has been thoroughly described in the article How to Create Custom Audiences for Pre-Release Music Campaigns in Facebook, Instagram, and Google.

6. Analyzing trending music in advertising to find the most syncable tracks in the own catalog.

Sync licensing which includes finding the sync opportunities and pitching specific songs can benefit from data analysis and benchmarking. Trending music in brand advertising can be analyzed to reveal the brand’s sound. This sound will then be matched to specific songs in your catalog making a strong case in the pitch to the brand in terms of sound-brand-fit. If you are interested in this use case and how data can be used in sound branding and sync licensing check out the interview we did with Vincent Raciti from TRO – About AI in sound branding.

– Knowledge is about the past, not the future. At the knowledge layer, you only have information about what happened before. Usually, information about the present (though some tools provide access to real-time data) or the future is not taken into account. It is important to remember this limitation as past performance is no guarantee to future results.

At the stage of Information, data is structured and organized so it can be interpreted. Information requires less time to find relevant data but there is still a lot of effort involved. 

At the Knowledge level, information is put into context and can be used for benchmarking and setting expectations. This context can be historical and involve past successes or it can relate to the position of others in the market. What kind of knowledge is derived from information depends on the initial data set and on the ability to store the memories of past successes. This process of turning data into knowledge takes a whole new form when machine learning and deep learning techniques are used as they significantly speed up the process of data collection and can memorize tons of data. However, a lot of knowledge in the music industry is still derived manually by looking at the past outcomes and trying to apply them somehow in the present.  

If you want to get the first grip on Cyanite’s technology, you can also register for our free web app to analyze music and try similarity searches without any coding needed.

1This article is a continuation of the series “How to Turn Music Data into Actionable Insights”. We will explore the second step of the Data Pyramid – Information and Visualizations. For details on the first step – Data – you can see the article here.

The Information stage of the Data Pyramid involves organizing and visualizing data. Organizing is often done with the use of business intelligence, while visualizations are presented in the form of pie charts, line graphs, tables, etc. This stage of the pyramid answers the question “What happened?” (or What is happening?) in the business. 

Data becomes information when it is structured and organized. Some think that organizing and structuring data is not enough and only when data is useful and meaningful then it can be considered information. To combine these two views, we will consider that data becomes information when it is structured and meaningful. The final goal of the information stage is to come up with data visualizations that combine data and information so that conclusions can be made.

Data visualization is the process of translating large data sets and metrics into charts, graphs, and other visuals. Visualizations combine both design and computing skills. There is also psychology involved in how people read and interpret visuals.  

So data visualizations pursue several goals: to communicate information effectively and to present the information in a visually appealing way. Some people even say that data should look sexy and an ideal visualization should stimulate viewer engagement. A great example of this is the work by the London-based Italian information designer Tiziana Alocci who uses visualizations for many different use cases: album covers, corporate visualizations, and editorial infographics.

Data visualizations that are used for business decisions are usually expected to be visually appealing and clear and communicate information. So design, data science, and computing knowledge should really work together to create effective data visualizations. 

Before data can be analyzed and visualized it is important to decide which data to use as well as ensure the quality of data. For different types of metadata used in the music industry, see the article here. Once the dataset is chosen and analyzed, it can then be visualized either with the help of a designer or using one of the three types of Music Data Visualization Tools:

1. Music Analysis and Discovery Tools

These tools show major characteristics of music such as genre, mood, emotional profile, energy level, etc., as well as show relations between artists and tracks. The idea is that similar tracks can be analyzed in detail and combined into playlists or recommended to customers. Cyanite usually falls into this category of music data visualization tools. 

2. Music Visualization Tools for Researchers

These tools are used for research purposes to prove a thesis or provide an overview of the field. For example, Ishkur’s Guide to Electronic Music was originally created as a genealogy of electronic music over the course of 80 years. It consists of 153 subgenres and 818 sound files

3. Marketing Tools

These tools present data visualizations that can be used for sales, advertising, and marketing purposes. They visualize data about consumer preferences, artists’ popularity, track consumption, industry trends, etc. which is not audio sound data. 

An example of marketing visualization tools could be web applications such as Pandora AMP and Soundcharts that provide data and visualizations to derive information. Tools like Tableau and Plot.ly allow you to upload raw data and get industry reports. 

The general visualization techniques to help represent data in an effective way are trend charts, comparison charts, pie charts, connections, maps, etc. These can be constructed manually or by the computer.

For example, spectrograms are used by Cyanite to train the computer to identify patterns in the audio sounds.

Cyanite’s moods are presented in a comparison chart where overarching mood and the least present mood can be easily identified.

Genre is represented in a trend chart that shows how track’s mean value of genre changes throughout the duration of the track. 

In this project, representing the history of rock, a connection chart shows how different artists relate to each other in a linear or hierarchical way. 

To learn more about data visualization techniques see the article here. 

We asked companies specializing in data analysis and visualization how data visualizations are used in their work. 

AMP Sound Branding works with data visualization experts and depending on where the company plans on using the data, they visualize it in different ways. 

In the research on automotive industry sound, for example,  AMP used a combination of Polar area charts and line charts to visualize brand moods and compare.

At TAMBR sonic branding, a big chunk of work is to create a shared understanding of musical parameters that surround a brand. 

TAMBR visualizations take some of the subjectivity away when choosing the right music for a brand. However, these visualizations are merely guidelines, not strict pointers. According to the company, magic happens where data and creativity meet.

Data visualizations are a powerful way to present the results of data analysis and gain additional insights. Visualizations can really improve the process of decision-making. They can also be used on their own in the sales process to impress customers. 

However, this stage of the Data Pyramid is also connected to a range of problems. For example, misinterpretations often occur when the output is interpreted as the only truth, disregarding the input dataset and its limitations. Sometimes people rely on visuals so much that they don’t go into exploring the deeper layers of data, missing out on the important information. The human element in algorithms is also a problem. In some algorithms, a human marks the data as important to consider for a machine, so this affects how the algorithm learns and develops. Nevertheless, data visualizations are widely used to present some version of the truth in a clean and digestible format. 

You really can’t ignore the simplicity of data visualizations and their ability to navigate the viewer’s attention to the key information. Yet, to get a simple graph, tons of data mining and data analysis work is usually required. 

If you want to get the first grip on Cyanite’s technology, you can also register for our free web app to analyze music and try similarity searches without any coding needed.