Gsoc2018Ideas

Student Project Ideas for Google Summer of Code 2018

This page lists the suggested projects for students working on Mixxx as part of Google Summer of Code 2018. Each of these projects represents something that we think would make a really big difference to our users and that we as a development team are really excited about. For advice on how to get in touch and how to apply, you should read GSoC Advice.

A GSoC application that simply repeats the project description will NOT be accepted. We expect you to think about the feature and how it aligns with Mixxx's goals, outline potential use-cases and propose a plan for implementing a solution.

Also, an application from a student who we are not familiar with will NOT be accepted. You must have some interaction with the Mixxx community before the days leading up to the application deadline for your application to be accepted.

Beatgrid Enhancements

Mixxx uses information about the tempo and location of beats for many features including sync, looping, and quantizing cue points. Currently, Mixxx assumes that tracks have a constant tempo for their entire duration. There is an option to disable that assumption, but it goes to the other extreme and treats the tempo as always changing. Both of these do not work well for lots of music. There should be an intermediate solution that allows for marking sections of a constant tempo and sections where the tempo is changing.

One way to approach this would be to mark sections with changing tempos by indicating which beats are downbeats and how many beats are in each measure. Then Mixxx could place the beat markers by dividing the space between the downbeats by the beats per measure. This would allow the tempo to change between downbeats (gradually or suddenly). For example, this video tutorial demonstrates how Serato DJ handles this. However, Serato and other DJ software assumes that all music has a 4/4 time signature, which is incorrect for lots of music. A proposal for this project should allow for handling tracks that change time signatures.

Adding this information to the beatgrid would allow existing features that rely on the beatgrid to work better for a wider variety of music. Additionally, new features that rely on knowing the location of downbeats and beats per measure could be implemented. You are encouraged to propose some possibilities for new features relying on this new beatgrid information in your application.

The analyzer that Mixxx uses provides the exact locations of each beat. This might be used to automatically guess whether to treat a part of a track as a constant tempo or changing tempo. Ideas for algorithms to detect downbeats and beats per measure automatically are welcome if you already have a strong background in signal processing, but this should come last after completing the rest of the project. If you do not have this experience, we welcome applications that would assume 4/4 by default and allow users to adjust this manually.

A strong application will list some specific tracks with changing tempos that will be used for testing the new features. These should include both tracks played by live musicians and tracks produced on a grid in a computer. Also, the application should propose how the new information available in the beat grid could be edited by users in a fast, intuitive way. Students with backgrounds in music theory and/or playing percussion instruments are encouraged to apply for this project, but these are not requirements and we welcome your application if you do not have that experience.

Effects Customizations

Mixxx's effects system is really flexible, but there is room for improvement to allow users to customize their setup more. Currently, users can load effects and set their parameters and Mixxx will remember this state when they start Mixxx again. This functionality could be expanded upon by allowing users to save and reload custom effect chain presets. That would allow users to have quick access to more effect chain setups than they can load at a time (most controllers only have controls for 2 effect units although Mixxx provides 4). These presets could be exported and shared by users online.

Additionally, it would be convenient for users to be able to arbitrarily rearrange and hide effects parameters. Most controllers only have knobs for 3 effect parameters but some effects have more than 3 parameters. Adding this would allow users to decide to have access to different parameters than the developers decide are the best ones to access on a controller.

Another feature which would be useful would be implementing a way for users to set custom per-effect defaults so each time they load a particular effect the parameters would be set how they want (and together with the above feature, the parameter order would be set too).

Cue Point Enhancements

Currently, Mixxx's hotcues are limited. They cannot store any information other than a position in a track. It would be helpful to expand the capabilities of this in a number of ways. For example, letting users label hotcues with custom text and set their own color coding for hotcues. Setting specially marked mix in and mix out markers would be helpful both for live performance and for telling AutoDJ when to start automatic crossfading. Storing multiple loops per track that could be activated with a hotcue would be helpful too.

A collection of ideas for improving cue points can be found in the Launchpad blueprint.

A strong application will include mockups for how these new features could be accessed by users in an intuitive user interface that does not clutter the screen with an excess of information. We suggest studying how other DJ applications handle cue points for inspiration, but do not copy how another program works exactly. If you do not have access to proprietary applications, you can search YouTube for tutorial videos and read the manuals of other DJ software.

MusicBrainz Integration

Mixxx uses AcoustID to identify tracks by fingerprinting their audio data. Subsequently the results are used to query the MusicBrainz database for metdata about the identified track. Currently we don't utilize the full potential that the MusicBrainz database is providing. We are reading just a few track properties to complement missing metadata.

MusicBrainz IDs

All entities in the MusicBrainz database are identified by UUIDs. These UUIDs could be used for various purposes:

• Detecting exact duplicates or variants of a track independent of the actual audio encoding. Use Case: When migrating your files from lower quality MP3 to higher quality FLAC encoding, obsolete MP3 files could be identified based on their IDs and proposed for removal.
• Relocating a track after it has been moved. Identifying tracks by their IDs will be much more reliable than by a combination of some properties. Use Case: You reorganize your files using a tool like Beets and don't want to lose all your carefully crafted crate/playlist/history contents when Mixxx is not able to asscociate your tracks in the library with the new file locations.

We recently implemented the import/export of MusicBrainz IDs according to the Picard Tag Mapping proposal. The next step is extending the MusicBrainz client for retrieving and the Mixxx database and storing those IDs. Afterwards your proposed features can be added based on these IDs.

Custom Tags

One essential feature that Mixxx is missing is the ability to assign and manage custom tags to tracks. We have already collected some ideas what and how to store this information in Launchpad #1743702. MusicBrainz records user-provided tags and ratings in their libary which could be used as a starting point and for synchronization. In the Mixxx database simple textual tags could be stored in an inverse index Tag String -> Track ID. It should also be possible to attach the custom tags of a track to the file by exporting/importing them as file tags.

Live Metadata Output

Many users would like to be able to show what music Mixxx is playing from other applications. There are lots of formats this information could be output to, for example MPRIS, HTTP APIs of various web services like ListenBrainz, or simply writing to a plain text log file. This project should make an extensible foundation that will make it easy to add new output formats. A proposal for this project should identify specifically which output formats will be implemented and explain some use cases for them.

Transition Effects

Currently effects are basically applied to one channel only. If you like to use them for transitions, you have to controls more than one knob at a time. It would be nice to make effect transitions as easy as cross fading.

During this project you need to define and implement an extensible way to control transition effects with a single knob.

A simple example is a "EQ transition" The EQ knobs are used to avoid to much bass during long cross fades. Currently the DJ has to turn various EQ knobs play/pause and the cross fader simultaneously.

Once this project is implemented the DJ should only select "EQ transition" and move the cross fader.

Something Else!

As always with Summer of Code, you aren't limited to the suggestions we've made here. If you've got a great idea for a project involving Mixxx then we're looking forward to hearing about it. We recommend spending more than a few days using Mixxx and participating in the community to develop a better understanding nof areas where Mixxx could use improvement. Our bug tracker is full of wishlist bugs and other ideas scattered throughout, so if you browse through it, you may find many more ideas for GSoC projects.

IMPORTANT: You should contact us first to get feedback if you're going to submit a proposal for your own project idea!