One of the techniques I learnt for equalisation, mixing, compressing and mastering is to attenuate the levels and frequencies you don’t want, then level the rest of the mix upwards. If you keep boosting what you do want, you very quickly end up with a noisy mix that clips with no room to breathe. Rework becomes impossible and the whole project is compromised.

If the frequencies you do want are very quiet to begin with, the final mix can still be of poor quality. It can be used for artistic effect but usually includes a lot of noise that requires removal and it’s hard to sample the artificially generated noise to create a quality noise reduction filter. It’s a careful balancing act.

I think the same applies to playlisting, specifically the approach I’m taking at an individual level. When I start a playlist I normally put lots of tracks in that I like but I’m undecided whether they really meet the specification. Repeated listening over a number of days helps me to make a decision. In some instances, especially in lockdown, I have to imagine myself at the event or venue. But rather than keep adding tracks that I think will work, I start to remove tracks from the playlist that I think don’t work. This can take a couple of weeks. I’ll eventually whittle down the playlist to the essence of what I think fits the requirements, and only then will I re-explore what could be added.

I use a couple of tools to sort my saved playlists which can help me bring a playlist together. My favourites happen to be Organize Your Music and Playlist Sorter for Spotify. I can create staging playlists very quickly from my already saved library of 4,000 tracks, which can be imported into my work-in-progress playlists. Once again, I start to whittle down what really works in the playlist and save the other tracks for another day. They tend to move to other playlists I save privately as they might have a different use.

Only when I’m happy with tracks that I know well and also believe work well in the playlist, do I start to explore suggestions. At the end of each playlist, Spotify has the “if you listened to that you might like this” approach. Very occasionally there’s a track in there that I think might work but more often than not I discard most of them. It’s not that I don’t like them, I just don’t think they always fit very well.

If I’m lucky enough to find a single track that’s ideal and sums up the essence of the playlist, I can start a radio based on that track but I know it’s swayed by my listening preferences. I sometimes find it ends up on a narrow loop of artists, which is no bad thing for expanding my knowledge of an artist, but with each playlist I try to keep the number of tracks by the same artist to a minimum.

If it isn’t automated, how do you do it?

So how do you go about making a playlist without automated suggestions? It’s time consuming and not easy.

I do use the tools mentioned above to find new tracks. I also listen to playlists I find, or search for artists on Amazon and look to see what people frequently bought together.

I’ve listened to SomaFM for many years and in some ways is partly, inadvertently, the inspiration for HumanPlaylistr. Their USP is, “All music hand-picked by SomaFM’s award-winning DJs and music directors.” It’s great for spending an hour listening to music I may not have heard before, looking up artists’ discographies and exploring related music.

I listen to the music being played around me everywhere I go. At the moment, that isn’t very many places! I’m always intrigued by what’s being played and how it’s been selected. Just before Christmas I was in a local large convenience store and the CD of Christmas anthems started skipping. It became obvious very quickly that the staff chose the music as they rushed from the aisles to sort it out. Any playlist in a commercial environment needs to be considered by the staff who work there. After all, they’re the ones who have to listen all day, every day, and might appreciate rotating playlists of some considerable length. I’ve been to some venues with impressive playlists where I’ve ended up using Shazam as it’s music I don’t recognise.

I actually think Spotify missed a trick by not buying Shazam; the technology is fascinating – How does Shazam work? Music Recognition Algorithms, Fingerprinting, and Processing and An Industrial-Strength Audio Search Algorithm are good starting points (but are quite technical, be warned) – and instead persevered with Echo Nest. They’re two different products as I understand it, undeniably both very smart. Spotify must have sound commercial reasons for it, but once again Apple own the technology that users actively see, hear, and interact with.

I have musician friends who are getting airplay on radio shows that I haven’t heard. They tweet about it and I listen to the show online, listening for artists and music that are new to me. (See HumanPlaylistr Independence Assurance.)

I also listen to the radio to find out what’s being played locally and nationally. I try to follow events and festivals.

As a trained musician (tenor voice, violin, piano, bass guitar) over the last 20 years I’ve performed a wide variety of works as part of orchestras and choirs in venues around Europe. The repertoire covers a variety of works, highlights include: Mozart’s Requiem, Vaughan Williams’ A Sea Symphony, Bruckner’s Mass No. 2 in E minor, Elgar’s Dream of Gerontius, Bernstein’s Chichester Psalms, Rossini’s Petit Messe Solennelle and Berlioz’ Grande Messe Des Morts. I enjoy a wide variety of classical music including contemporary and film music. I named my company mascot after the contemporary composer Max Richter. In this instance I use my experience to create classical playlists. Events such as Weddings are ideal settings for classical playlists.

I believe HumanPlaylistr is responsive to requirements and my USP is me, my saved library of tracks I know incredibly well, and my willingness to craft and curate playlists for anyone who would like my help in specific settings. I bet you already have some ideas what you want for your event or venue so maybe, just maybe, HumanPlaylistr is the service you’ve been looking for.

Of course, once you’ve got the ideal playlist, that’s where you start expanding your ideas and adding what you really want. Those automated suggestions might start being far more relevant.

Then you can amplify what you do want.

1 Comment

  1. Music Discovery – HumanPlaylistr

    February 28, 2021 at 1:07 am

    […] a previous post Attenuate what you don’t want, I gave a bit of insight into how I start a playlist by inputting far too many tracks from my own […]


Join the conversation