We have a guest article today, from DJ G-Force (@Scythe on our Discord). This is definitely a more advanced workflow so this information is very welcome. Thanks!

---

Every DJ is different, cares about different things, makes decisions based on different factors and as such, their workflows for preparing their music collection should also be unique. A good workflow gives you a strong framework to process your collection in a reliable, repeatable way. A good workflow for me, may not be a good workflow for you, what was a good workflow for me today, is likely to change and evolve over time into something better.

What I want to document in this article is my current workflow for improving my music collection and some of the decision-making that led me to this, so that you can make your own decisions to start refining your own workflow. I’ll also note that a workflow is just a guideline, you won’t always follow it to the letter, and sometimes you’ll take shortcuts for various reasons and that’s ok. We’ll even cover some tricks to catch things that have fallen through the cracks later in the article.

A quick background on me: I’ve been DJing for 15 years in the underground/hard dance music scene and have collected a very large collection of music during that time. Musically, I play just about everything from Hard Trance and Hard Techno (140-160bpm), through to Happy Hardcore (160-180bpm), through to Gabber (170-200bpm), Frenchore (190-220bpm) and beyond. Like most, my music collection practices were sloppy for the longest time and really it’s only been since the discovery of Lexicon and starting Twitch Streaming (which dramatically increased the total time spent DJing) that I made a decided effort to improve my music organisation practices.

What’s important to you?

A workflow should only focus on what’s important to you relative to your music, with that in mind, let’s take a look at all the possible things we can check, populate and use relating to a single track and work out what’s important to you and what’s not:

• Core Metadata - Absolute minimum metadata for a track (Artist, Title)
• Basic Metadata - Universally supported metadata fields (Album, Genre, Year, Rating, etc)
• Advanced Metadata - Lexicon specific metadata (Energy, Danceability, Popularity, Happiness, Extras)
• Album Art
• Track Colour
• BPM & Beatgrid
• Key
• Cue Points (These can have sub categories like mix markers, sample loops and track component markers)
• Tags (There’s a whole separate rabbit hole of what tag categories and tags are important to you)

Of these, I place the highest value on having correct Artist, Title, Year, Rating, Beatgrid, Key and Genre (Though I use a genre tag category rather than the default genre field).

Given the high energy and relative obscurity of my music I haven’t found much use myself in the Energy/Danceability/Popularity/Happiness metadata, particularly with the values Lexicon is able to populate automatically. For a wedding DJ or someone playing much more commercial music I could see these being immensely useful, I’ll still scan for these values because why not, but I don’t typically make use of them when building playlists.

I also personally have not found a good system for utilising the track colour feature. With so many genres in my collection I haven’t come up with a colour system that’s both memorable and meaningful across multiple genres. I find way more use out of the tags system so I mostly ignore the track colour feature.

Using the best tool for each job 

Given the obscurity of my music beatgrids and key can be hard for some programs to pick up. Given the size of my collection, I’d prefer to reduce the amount of manual intervention required for these values so I  make use of one of the best things Lexicon unlocks, allowing us to use different DJ tools for different tasks. Namely, Traktor is regarded as having the best beatgrid detection algorithm (Rekordbox is not too far behind) and Mixed in Key has the best Key detection. So I wanted to work both of these into the workflow. If your music collection is less obscure there might not be any tangible difference between the results of different DJ app scanners so you might prefer to just do both your grids and Key scans in a single app.

The workflow for new music

I try to batch process things wherever possible as this workflow is a bit heavy-handed for just a few tracks. All newly purchased music gets thrown into my “Incoming” folder on my NAS which is mapped to my Lexicon Watch Folder and gets picked up automatically. Once I decide it’s time to sit down and process them it looks like this.

1. All the incoming tracks get added to a new “%Month% Analysis Playlist ” playlist under the Lexicon Folder (I use the Lexicon playlist folder as a space for miscellaneous and temporary playlists that aren’t meant for DJ use). Then I’ll sync the playlist to Traktor Pro. 
2. In Traktor, I select all tracks in that playlist, right click and “Disable Analysis Lock” as this is sometimes enabled by default. Then right-click “Analyze (Async)” with BPM and Beatgrid selected. Note: I only use the free Traktor demo for analysis and it works great. Even with large batches of analysis if the 30-minute demo timer is reached it will continue to process your analysis before you have to quit. Once the analysis is complete we can close Traktor and use the Sync -> Import inside Lexicon to import that beatgrid data back to Lexicon.
3. Next up we want to get our Key information. To get our playlist into Mixed In Key (MIK) we can use the Sync -> iTunes option in Lexicon to output our playlist to an iTunes Library file. In MIK, we click Add Tracks -> Import iTunes -> Import from File, select our library file Lexicon created, select the playlist and select import.
   Note: In MIK under the Update Tags Settings I have “Update Common Tags” unchecked, and “Update custom ‘Initial Key’ tag” checked, with everything else unchecked. This ensures that MIK only updates the “Key” metadata field that Lexicon can then read.
4. Once MIK has finished analysing the tracks we can switch back to Lexicon, select all our tracks in the playlist and right click -> Reload Tags and Toggle everything off except the “Key” values. Now we have all our tracks with accurate grid and key information. Next up, metadata and tags.
5. The state of metadata for the incoming tracks would determine what gets included here but I’ll often use some Smart Fixes and Recipes to ensure that the Artist/Title fields are clean and correct. Once those two values are clean I’ll run the Find Tags & Album art tool. Usually with the following settings:
   
   - Genre, Energy, Danceability, Popularity, and Happiness all checked for Find & Overwrite
   - Album Art checked for find, but not overwrite
   - If I know some tracks are re-released, I’ll also find and overwrite the Year field as well to try and get the original year of the track. Sometimes I’ll have to do this manually if the scan can’t find the original year.
   
   
6. Now we get into tagging and rating. Because I hoard a lot of music it’s important for me to tag and rate tracks wherever possible to make finding suitable stuff later a lot easier. The only tags I really try to populate for every track as part of this incoming music workflow though are my Genre tags, as long as those exist for a track I can at least categorise things at a high level and tracks start to flow into my various smartlists (that’s a whole different topic). I’ll sometimes do other tagging and star ratings for tracks during this process but I’ll often leave that for later, either in dedicated music tagging/review sessions or while doing Twitch DJ sessions where I can tag/rate on the fly as I try new music in a low pressure live situation.
7. With those tags set our tracks are pretty much done being processed for now. I’ll select all the tracks in the playlist again and right-click -> Use -> Write Tags to File to ensure that everything we’ve populated in Lexicon is also written straight to the files.
8. Lastly, we can select all our tracks from the Incoming tracks page and mark them Done to move the tracks out of our Watch Folder and into their final destination. As for how you prefer to have your final music organised on the hard drive I don’t strongly recommend any particular structure. With things like Lexicon and your existing DJ tools, the actual file paths don’t matter that much anyway, all that matters is that your tracks are organised and don’t move around on the system unexpectedly. My personal folder structure currently looks like %First Tag% / %Artist% / %Original Filename% where First Tag is the first genre tag on the track. 

Improving tagging your tracks over time

Having a rock-solid workflow for new incoming tracks is great, but what about your existing collection, what about when you need to take a shortcut and add some new tracks at a moment's notice for a gig that evening? smartlists to the rescue! You can use smartlists to create intelligent utility playlists to catch songs that require further work from you. How you utilise these is entirely up to you and what information you want to catch but I’ll walk through some stuff that works for me to give some inspiration.

Let’s circle back to what’s important to me, Grids and Key are already done, Artist & Title should be mostly ok for existing tracks (Though occasional use of the Artist Cleanup tool and various Recipes are helpful as you spot things to correct). That leaves us with Year, Rating and Tags. I have a playlist folder for each of the primary genres I play that contains relevant playlists for those genres. There are also smartlists that automatically pick up tracks that get tagged with the Genre tags as well. Within the folder I can make a ToDo folder with 3 smartlists:

• In Playlist -> In Playlist Folder -> “Genre Playlist Folder” & Rating -> Does not Exist
• In Playlist -> In Playlist Folder -> “Genre Playlist Folder” & Tag Category -> No Tags from Categories -> “Genre”
• In Playlist -> In Playlist Folder -> “Genre Playlist Folder” & Tags -> Has all these tags -> Re-Release 

The first two are pretty straightforward, they’ll load up any tracks in the playlist folder that don’t yet have a rating or genre tag respectfully. When I’m in the mood to do some tagging of a particular genre I can load up these lists and get straight to it. The third one dealing with Year is a little more nuanced. Most tracks in my collection have a year date but If I come across something while mixing or tagging that I know is a re-release and doesn’t have the correct original year I’ll just tag it as a re-release. Then I can come back to this list periodically when I’m in the mood to sit down and find original years for those tracks. You can see pretty clearly that setting things up so that you can do similar tasks as bigger batches is my goal here.

I also have some similar smarlists set up that aren’t limited to a specific playlist folder that allow me to sit down and go through various other untagged/unrated music in my collection. As long as I add the Genre tag, those tracks will start finding their way into the various smartlists within each Genre for further action.

Tagging and rating tracks in Rekordbox as a Lexicon user

Being a DJ on Twitch gives me plenty of time in a low-pressure environment with a crowd to test and experiment with new tracks. I have a StreamDeck setup that allows me to easily tag and rate tracks in Rekordbox while I’m mixing/streaming allowing me to feel like I’m continuously making organisation progress on my collection. If you’re just using regular playlists it’s pretty straightforward to just import that playlist back to Lexicon after your session but it’s a little more tricky when you’re using smartlists since RB can’t import a smartlist without doing an import of the whole collection which is time-consuming and not recommended unless you’ve also done a full collection sync to Rekordbox just prior. Instead, In Rekordbox after a session I’ll go to the histories list and turn my tracklist from that session into a playlist, which I throw into a histories folder. Then in Lexicon I can import that history playlist to ensure the tracks I played get updated in Lexicon.

I’ll note that I make use of the “Specific Fields” option on import to only import the values I intended to change. Tags are always updated so I’ll usually just have ratings selected as I don’t usually make other changes in Rekordbox. While most beatshift issues have been resolved, reducing the number of times the beatgrid/cues get imported back and resynced reduces the chance that any very minor beatshift issue gets multiplied over time. Of course, there are times when I’ll have added new cues in Rekordbox or needed to make a grid correction and if that occurred I’ll include them on the import.

---

Thoughts, comments or questions? Come chat about your workflow on our Discord.

Do you have a Lexicon life-hack? We love guest articles so don't hesitate to share.