Shoutout to Mixxx, it's one of the examples of open source being able to match and in many ways surpass the enjoyability of their closed-source for-profit alternatives. I regularly give it as an example alongside things like Blender to show that a better, more free software world is possible outside of just developer tools.
I love Mixxx, I use it daily as a music player. I know it is a virtual DJ deck, but it is also a fantastic music player.
That being said the following sentence makes me a bit concerned "The current user interface of Mixxx has served us well, but as technology evolves, so must we."
As far as I know QWidget is neither deprecated nor are there any plans for that.
I think Mixxx has one of the best UI's I ever encountered, in design and execution. Please don't botch this.
Yeah, I know both QWidgets and QML quite well and I also don't see a strong reason to port. Genuinely custom painting of highly detailed elements in QML is either pretty annoying (comparable to doing it in OpenGL, and then you still need to do user interaction without much help from the framework) or slow if cobbling it together from existing items or using a canvas. I've helped fix particular issues on a project that needed highly detailed custom painting in QML (think rendering text in an editor component, though it wasn't that), but in the end it failed for performance reasons. I knew of some really good developers on that project, so it wasn't plain incompetence. Edit: I think it can be done, it's just a lot of avoidable work.
Mixxx might not have that problem, not sure. The waveform view is presumably custom painted, but shouldn't be too difficult. There may or may not be others.
Every user who uses a screen reader has been asking for this. Also, though I can't speak to this personally, apparently the theming engine is much easier to maintain in QML which is a big time saver for the devs.
FWIW I've found mixxx to be much more stable than serato or traktor (commercial stuff). I can't say whether they focus on bugs or features like your criticism suggests.
I have no problems with Mixxx in that regard. I run it on Linux and for me it is rock solid. It can't help if your computer crashes though.
I used to use mpd for a while which automatically continues playing where it left off after a reboot.
What I learned from that phase is that the music I enjoy in the evening is rarely the same I like blasted in my ears mid-song at full volume in the morning. Terribly annoying feature, will not use again.
As a Mixxx user of 8 years with many life shows and even longer ones at home I have never had Mixxx crash on me, not even once.
I had some UI annoyances, so I support the effort and trust the devs to make the right choice. Otherwise I can still just run the old version till 3.x turns stable.
I like the way you organize your music in crates and playlists, which are permanent structures and then you arrange everything for the occasion[1] in the AutoDJ, which is ephemeral.
I recently looked and bought into the DJ tech today. I know mixxx, used it many times. And Rekordbox isn't perfect either. But you can't even compare. If the one is Dj Software, mixxx right now, is just a music player with DJ theme
An alternate view on this is Rekordbox is software that you have to pay for, tries to upsell you to a monthly subscription, has arbitrary limitations on what hardware it will work with and by default sends data about every track you play to some 3rd party service.
Mixxx doesn't do any of that and it's absolutely powerful enough to use for DJ sets.
I bought the hardware without knowing that and have a device bound license. It's weird that the software literally degrades when the device isn't attached.
The hardware limitations are crazy and weird, especially when you realize that parts of the limits are purely trough the visualisation. But when you have a supported device the performance is amazing even on older hardware. (I run it on a Wyse 5070)
Like all non free (or any software) it's a double sided sword. However it's one of the few industries where the monopoly software is actually not bad at all and doesn't force you into monthly payments when you buy their devices (which is basically industrie standard anyway, and not actually that much more expensive).
Mixxx is a great piece of software, even better that it's OSS and runs linux. Exciting to see a new major release.
I've been using it lately (2.x) just mixing internally and with cue points set up in my tracks, and using the laptop keyboard for shortcuts. Works really well. For the occasional xfade etc I do have an external MIDI controller mapped to the xfade, levels, and eq but could get away without it for my style of mixing.
Been DJing off and on for 25y and producing electronic music for 30y.
FYI this is not a new major release, this is work starting on a new major release. At least two more versions of 2.x are planned and 3.0 is not yet actually scheduled.
May not have been what you meant, but I wanted to clarify for others reading this comment: 3.0 is quite a ways off.
I got out of DJing and into tech because of my love for dental insurance. For a while I tried to influence the direction of mixxx and was getting great feedback from other DJs but none from the developers.
This was my great introduction to why I will never be a great developer, and why open source will never be on par with paid professional software. I admire their efforts, but mixxx will always be a hobby project in my eyes.
This was around 10 years ago, but at the time the suggestions on the forums were more towards fidelity and 5.1 surround sound. I can't recall exactly, but something of the sort. My guess is that these were people with audio backgrounds who wanted to commit to the project.
I was making suggestions that were more for DJs. My suggestions were (vaguely from memory) keyboard and midi shortcuts for going back 4 - 16- 32 beats. Looping the same number of beats. setting exact cue points (think off beat) and setting synchronized (auto move to the closest beat if bpm was set). Nothing crazy and things that I felt would be necessary for most DJs.
I must have set my alerts far to high because I was bombarded with emails for a week until I completely unsubscribed. At the time I was also getting interested in XBMC and felt their community was much more welcoming and the software was simpler to understand.
Recently got a mix board and have been learning on mixxx. If anyone has any general guides for learning to dj in mixxx I’d love to know. Honestly, I feel a little lacking in the fundamentals of digital audio so I’m probably going to focus on that before going too deep, but so far mixxx has been great! It was pretty hard for me to navigate, but that’s 100% because I have no idea what I’m doing. Curious to seeing if the update will help with discoverability.
80% of DJing is the generalized skill of picking music that fits the vibe, that you can practice with any software. 10% is doing transparent and/or engaging transitions and the rest of the 10% is technical chops, being fluid and being able to recover when you inevitably fuck up. Approach your learning accordingly :)
As a DJ since age 16 who earned a living from it once:
The number one most important skill is to pick the right music in the right moment. Period. Online many people will tell you about beatmatching, transitions, the best gear and all that, and sure, it is good if you can do that or buy that. But better a DJ who picks good music with crude transitions than one who picks boring music with stellar transitions. My digital sets I do with my laptop, Mixxx and an audio interface, no controller, just keyboard shortcuts.
Even if you just crudly made hard cut-transitions and don't match beats you can still give people a great evening if your music selection and the order in which you present it is good. I once had people partying till early morning with my phone (android/poweramp) as a source, as it wasn't a planned gig.
That means you (should) invest significant time in listening and sorting your music and have an clear idea which types of DJ set you like. A lot of this will depend on the style(s) of music as well. Putting on dub is very different from straight techno which is very different from let's say playing an 80s post punk set or something entirely eclectic. And a lot of that depends on the venue, the crowd etc. A good DJ will react to the circumstances around them.
As for Mixxx, most of the metaphors and parts of the UI in DJ software is derived from vinyl DJing. A crate would be a literal crate filled with vinyls. Since vinyl is heavy as shit, bringing all your music was/is not an option and DJs packed literal crates, for a specific gig, etc. The two Decks on Mixxx are refering to the record players (typically two "Decks") you would have with vinyl. In the middle you have a mixer section, just like the mixer between the two record players.
Your task is (a bit simplified): selecting a song, playing it, selecting another song and then transitioning to that song ad infinitum. So you load up a song in the left deck hit play and have it play and you move the crossfader to the left so only the left deck can be heard. You then search for the next track while the song is playing and load it up on the right deck. Mixers typically have a PFL button (pre fader listen) that allow you to "privately" listen to channels even if their volume fader (or the crossfader) is turned down (this is why DJs have headphones). You basically listen with one ear to the the thing the audience hears and with the other to the next song.
Now the song on the left deck is nearing its end and you have another song on the right deck, you try to make the two align in time, tempo, frequency, pitch in such a way that the transition sounds good and then you use the crossfader to blend over. Or you could just wait for song A to finish and then hit play on song B if you don't care too much about the transition. Or you could mix only the bass from song A with the hihats from song B, add effects, shout something into the mic and then play a song from a third deck, while all while listening to song requests.
For the tempo matching Mixxx has a button that does that for you on each deck, not like if you DJ with vinyl where you have to do that by ear. The mixer has an EQ (changing the level of Bass/Mids/Highs at fixed frequencies) and a Filter (cutting away high and low frequencies with variable frequency, allowing you to avoid situations where two basslines/hihats are playing at once during a transition.
You can go arbitrarily complex, but as mentioned at the beginning, music selection and situational awareness is king, everything else is just details as long as you don't make jarring mistakes with jumps in level or prolonged gaps of silence etc.
This generally results in a bulky and oversized UI with excessive padding. The current UI is satisfactory and lightweight. Why fix something that's working well?
Every single dj software is super crammed and has no empty space in it. This is true because djs cannot switch views while performing. There’s a lot of info which must be visible at all times on 1 single page. Regardless of UI framework or target device, it will continue to be like that because the target audience needs the interface to be dense.
1. better readability in the tables: good and probaly themable
2. tiny waveforms in the table: very cool, allows you to quickly judge the structure and sonic qualities of a track without listening! A real time-saver in a live situation
3. Library overhaul: the left pane always felt a little clunky: can't say much from the sceeenshots, but doesn't look bad
4. Tooltip: better than the old ones for sure
5. Enhanced support for lightweight plattforms: cool, not gonna use it, but hey
6. better search: This is a big one. Search always felt a bit unfinished. It worked, but sometimes find a song you knew existed was faster using the browser..
7. Interactive Settings: nice one as well. The settings were probably the least polished part of Mixxx, this increases the speed once more.
Also if there's a tiny network blip before buffering is complete everything goes to hell. I don't get streaming DJing in general, seems like something serious DJ software should avoid (unless they're targeting the casual party crowd, which is fine)
This isn't about videos I guess. I had parties where guests where demanding songs (every DJs dream) that I obviously didn't had on my drive. That gives you three options currently:
1. Tell them to fuck off, which may or may not work in any given context
2. Play the music in the browser and then hop back to Mixxx in time (dangerous)
3. Downloading the song via 'yt-dlp -x --audio-format mp3 --audio-quality 0 <URL>', fixing the questionable metadata (optional) and then loading it into Mixxx (cumbersome)
It would be cool if you could just paste the URL into Mixxx, have it cache a local version, render the waveform, store it in a special temp crate or what and then you can play it. But I guess yt-dlp isn't the most stable dependency...
from what I've gathered, it's a licensing nightmare. Of course it's better than the current alternative which is piracy, but it's a drop in the bucket with regards to the people who will use it in this specific licensing agreement.
Shoutout to Mixxx, it's one of the examples of open source being able to match and in many ways surpass the enjoyability of their closed-source for-profit alternatives. I regularly give it as an example alongside things like Blender to show that a better, more free software world is possible outside of just developer tools.
I love Mixxx, I use it daily as a music player. I know it is a virtual DJ deck, but it is also a fantastic music player.
That being said the following sentence makes me a bit concerned "The current user interface of Mixxx has served us well, but as technology evolves, so must we."
As far as I know QWidget is neither deprecated nor are there any plans for that. I think Mixxx has one of the best UI's I ever encountered, in design and execution. Please don't botch this.
Yeah, I know both QWidgets and QML quite well and I also don't see a strong reason to port. Genuinely custom painting of highly detailed elements in QML is either pretty annoying (comparable to doing it in OpenGL, and then you still need to do user interaction without much help from the framework) or slow if cobbling it together from existing items or using a canvas. I've helped fix particular issues on a project that needed highly detailed custom painting in QML (think rendering text in an editor component, though it wasn't that), but in the end it failed for performance reasons. I knew of some really good developers on that project, so it wasn't plain incompetence. Edit: I think it can be done, it's just a lot of avoidable work.
Mixxx might not have that problem, not sure. The waveform view is presumably custom painted, but shouldn't be too difficult. There may or may not be others.
Every user who uses a screen reader has been asking for this. Also, though I can't speak to this personally, apparently the theming engine is much easier to maintain in QML which is a big time saver for the devs.
The one killer feature any DJ software can have - and it's a tough one - is....
Not crashing during playback.
But hey why refine and fix bugs when you can constantly iterate on new features no one asked for (sadly this applies to most open source software).
FWIW I've found mixxx to be much more stable than serato or traktor (commercial stuff). I can't say whether they focus on bugs or features like your criticism suggests.
I have no problems with Mixxx in that regard. I run it on Linux and for me it is rock solid. It can't help if your computer crashes though.
I used to use mpd for a while which automatically continues playing where it left off after a reboot.
What I learned from that phase is that the music I enjoy in the evening is rarely the same I like blasted in my ears mid-song at full volume in the morning. Terribly annoying feature, will not use again.
As a Mixxx user of 8 years with many life shows and even longer ones at home I have never had Mixxx crash on me, not even once.
I had some UI annoyances, so I support the effort and trust the devs to make the right choice. Otherwise I can still just run the old version till 3.x turns stable.
I am intrigued, what makes it a good music player for non-DJs?
I like the way you organize your music in crates and playlists, which are permanent structures and then you arrange everything for the occasion[1] in the AutoDJ, which is ephemeral.
I also like the BPM and key display.
[1] In my case just my current mood
I use Apple's app Music on my Mac. Unfortunately, still the best music library I know about for macOS. But maybe others have other suggestions.
I recently looked and bought into the DJ tech today. I know mixxx, used it many times. And Rekordbox isn't perfect either. But you can't even compare. If the one is Dj Software, mixxx right now, is just a music player with DJ theme
An alternate view on this is Rekordbox is software that you have to pay for, tries to upsell you to a monthly subscription, has arbitrary limitations on what hardware it will work with and by default sends data about every track you play to some 3rd party service.
Mixxx doesn't do any of that and it's absolutely powerful enough to use for DJ sets.
I bought the hardware without knowing that and have a device bound license. It's weird that the software literally degrades when the device isn't attached.
The hardware limitations are crazy and weird, especially when you realize that parts of the limits are purely trough the visualisation. But when you have a supported device the performance is amazing even on older hardware. (I run it on a Wyse 5070)
Like all non free (or any software) it's a double sided sword. However it's one of the few industries where the monopoly software is actually not bad at all and doesn't force you into monthly payments when you buy their devices (which is basically industrie standard anyway, and not actually that much more expensive).
I don't mean to suggest it's a weird situation - requiring people to pay for the software (directly or indirectly) makes perfect sense commercially.
I’d be interested to know what the main differences are
Mixxx is a great piece of software, even better that it's OSS and runs linux. Exciting to see a new major release.
I've been using it lately (2.x) just mixing internally and with cue points set up in my tracks, and using the laptop keyboard for shortcuts. Works really well. For the occasional xfade etc I do have an external MIDI controller mapped to the xfade, levels, and eq but could get away without it for my style of mixing.
Been DJing off and on for 25y and producing electronic music for 30y.
FYI this is not a new major release, this is work starting on a new major release. At least two more versions of 2.x are planned and 3.0 is not yet actually scheduled.
May not have been what you meant, but I wanted to clarify for others reading this comment: 3.0 is quite a ways off.
The fact that Mixxx works as well as it does on Linux is a testament to human ingenuity and perseverance. It is truly The People's mixing deck.
I got out of DJing and into tech because of my love for dental insurance. For a while I tried to influence the direction of mixxx and was getting great feedback from other DJs but none from the developers.
This was my great introduction to why I will never be a great developer, and why open source will never be on par with paid professional software. I admire their efforts, but mixxx will always be a hobby project in my eyes.
What methods were you using to try and influence the direction of Mixxx?
This was around 10 years ago, but at the time the suggestions on the forums were more towards fidelity and 5.1 surround sound. I can't recall exactly, but something of the sort. My guess is that these were people with audio backgrounds who wanted to commit to the project.
I was making suggestions that were more for DJs. My suggestions were (vaguely from memory) keyboard and midi shortcuts for going back 4 - 16- 32 beats. Looping the same number of beats. setting exact cue points (think off beat) and setting synchronized (auto move to the closest beat if bpm was set). Nothing crazy and things that I felt would be necessary for most DJs.
I must have set my alerts far to high because I was bombarded with emails for a week until I completely unsubscribed. At the time I was also getting interested in XBMC and felt their community was much more welcoming and the software was simpler to understand.
Related. Others?
Mixxx: GPL DJ Software - https://news.ycombinator.com/item?id=42769871 - Jan 2025 (133 comments)
Mixxx, a free and open source DJ software - https://news.ycombinator.com/item?id=31731814 - June 2022 (14 comments)
Open source digital DJ: Integrating Airtime and Mixxx - https://news.ycombinator.com/item?id=2580273 - May 2011 (4 comments)
Recently got a mix board and have been learning on mixxx. If anyone has any general guides for learning to dj in mixxx I’d love to know. Honestly, I feel a little lacking in the fundamentals of digital audio so I’m probably going to focus on that before going too deep, but so far mixxx has been great! It was pretty hard for me to navigate, but that’s 100% because I have no idea what I’m doing. Curious to seeing if the update will help with discoverability.
80% of DJing is the generalized skill of picking music that fits the vibe, that you can practice with any software. 10% is doing transparent and/or engaging transitions and the rest of the 10% is technical chops, being fluid and being able to recover when you inevitably fuck up. Approach your learning accordingly :)
As a DJ since age 16 who earned a living from it once:
The number one most important skill is to pick the right music in the right moment. Period. Online many people will tell you about beatmatching, transitions, the best gear and all that, and sure, it is good if you can do that or buy that. But better a DJ who picks good music with crude transitions than one who picks boring music with stellar transitions. My digital sets I do with my laptop, Mixxx and an audio interface, no controller, just keyboard shortcuts.
Even if you just crudly made hard cut-transitions and don't match beats you can still give people a great evening if your music selection and the order in which you present it is good. I once had people partying till early morning with my phone (android/poweramp) as a source, as it wasn't a planned gig.
That means you (should) invest significant time in listening and sorting your music and have an clear idea which types of DJ set you like. A lot of this will depend on the style(s) of music as well. Putting on dub is very different from straight techno which is very different from let's say playing an 80s post punk set or something entirely eclectic. And a lot of that depends on the venue, the crowd etc. A good DJ will react to the circumstances around them.
As for Mixxx, most of the metaphors and parts of the UI in DJ software is derived from vinyl DJing. A crate would be a literal crate filled with vinyls. Since vinyl is heavy as shit, bringing all your music was/is not an option and DJs packed literal crates, for a specific gig, etc. The two Decks on Mixxx are refering to the record players (typically two "Decks") you would have with vinyl. In the middle you have a mixer section, just like the mixer between the two record players.
Your task is (a bit simplified): selecting a song, playing it, selecting another song and then transitioning to that song ad infinitum. So you load up a song in the left deck hit play and have it play and you move the crossfader to the left so only the left deck can be heard. You then search for the next track while the song is playing and load it up on the right deck. Mixers typically have a PFL button (pre fader listen) that allow you to "privately" listen to channels even if their volume fader (or the crossfader) is turned down (this is why DJs have headphones). You basically listen with one ear to the the thing the audience hears and with the other to the next song.
Now the song on the left deck is nearing its end and you have another song on the right deck, you try to make the two align in time, tempo, frequency, pitch in such a way that the transition sounds good and then you use the crossfader to blend over. Or you could just wait for song A to finish and then hit play on song B if you don't care too much about the transition. Or you could mix only the bass from song A with the hihats from song B, add effects, shout something into the mic and then play a song from a third deck, while all while listening to song requests.
For the tempo matching Mixxx has a button that does that for you on each deck, not like if you DJ with vinyl where you have to do that by ear. The mixer has an EQ (changing the level of Bass/Mids/Highs at fixed frequencies) and a Filter (cutting away high and low frequencies with variable frequency, allowing you to avoid situations where two basslines/hihats are playing at once during a transition.
You can go arbitrarily complex, but as mentioned at the beginning, music selection and situational awareness is king, everything else is just details as long as you don't make jarring mistakes with jumps in level or prolonged gaps of silence etc.
This generally results in a bulky and oversized UI with excessive padding. The current UI is satisfactory and lightweight. Why fix something that's working well?
Every single dj software is super crammed and has no empty space in it. This is true because djs cannot switch views while performing. There’s a lot of info which must be visible at all times on 1 single page. Regardless of UI framework or target device, it will continue to be like that because the target audience needs the interface to be dense.
As a long time Mixxx user What do I see there:
1. better readability in the tables: good and probaly themable
2. tiny waveforms in the table: very cool, allows you to quickly judge the structure and sonic qualities of a track without listening! A real time-saver in a live situation
3. Library overhaul: the left pane always felt a little clunky: can't say much from the sceeenshots, but doesn't look bad
4. Tooltip: better than the old ones for sure
5. Enhanced support for lightweight plattforms: cool, not gonna use it, but hey
6. better search: This is a big one. Search always felt a bit unfinished. It worked, but sometimes find a song you knew existed was faster using the browser..
7. Interactive Settings: nice one as well. The settings were probably the least polished part of Mixxx, this increases the speed once more.
I wish this had youtube video mixing support. This is something i cant find in other software.
this went out of style a long time ago. when you play videos out, everyone just stares at the screens.
Also if there's a tiny network blip before buffering is complete everything goes to hell. I don't get streaming DJing in general, seems like something serious DJ software should avoid (unless they're targeting the casual party crowd, which is fine)
Works well for karaoke though for intermission songs.
The gold standard for that though seems to be VirtualDJ.
This isn't about videos I guess. I had parties where guests where demanding songs (every DJs dream) that I obviously didn't had on my drive. That gives you three options currently:
1. Tell them to fuck off, which may or may not work in any given context
2. Play the music in the browser and then hop back to Mixxx in time (dangerous)
3. Downloading the song via 'yt-dlp -x --audio-format mp3 --audio-quality 0 <URL>', fixing the questionable metadata (optional) and then loading it into Mixxx (cumbersome)
It would be cool if you could just paste the URL into Mixxx, have it cache a local version, render the waveform, store it in a special temp crate or what and then you can play it. But I guess yt-dlp isn't the most stable dependency...
from what I've gathered, it's a licensing nightmare. Of course it's better than the current alternative which is piracy, but it's a drop in the bucket with regards to the people who will use it in this specific licensing agreement.