Program For Analyzing Music

[steveryan]

I've been thinking a lot about how to choose the order of the songs in a set. In the past I had always just mixed songs that I think sound good together or will compliment each other in some way. But I'm starting to change my thinking by focusing more on controlling the shape of my sets in terms of how the intensity goes up and down throughout.

It's a fairly obvious concept in DJ'ing, to start your set out mellow and gradually increase the energy. As to how high you want to take it is simply up to you. Regardless, your set will always have some kind of shape to it if you are aware or not.

For example, you could play in waves, where the energy goes up and down consistently with no real intense spikes at any given point. Or, you can play on an ascending and descending line where the energy eventually gets high, then goes down a bit, then goes even higher than where it was before, then down again, and then rises to the most intense part of the set, and then gradually works it's way down while going through spikes in energy along the way. If you think about it, these concepts all have their own shape and that's what I'm working on. I want to build sets that have shapes, or more specifically, I want to first choose the shape and then put the records in place to make it happen.

I've recently made a couple of nice sets on an ascending line just by stacking the records according to intensity. While I intend to continue doing this, I've also come across another dynamic that I'd like to incorporate.

Some songs are brighter than others. What I mean is that the high-hat, higher frequency sounds and various supporting elements to the rhythm can be fairly bright and crisp compared to other tracks where those elements are much more dim. An extreme example would be comparing Electro to Deep House. And while I wont be playing Electro, the brightness can vary from song to song with the genre that I play (House, Deep, Tech, etc...)

With that, I've figured that it might be better to mix brighter tracks into darker ones. The reason is that it's easier to fade out and hide a dark record underneath a bright one. If your playing a song that has a wide, low bass kick with dim elements, it will be easy to gradually drop the EQ bands and ultimately fade it away when a brighter track is mixed on top of it.

Conversely, if you were to mix a dark record on top of a lighter one, it might be more difficult to hide that big, bright record when fading into a song that just doesn't have those crisp, bright elements.

So, my idea is to organize my sets according to brightness and energy from low to high. Start the set out with dim, mild tracks and gradually mix your way into brighter and higher energy one track at a time.

It's not always easy though, to tell if one track is truly brighter than another.

I was thinking that it might not be a bad idea to analyze my tracks with an audio spectrum analyzer so that I can really see which songs are the bright ones and which are more dim. I found a website with a free spectrum analyzer, but it's so basic I can't do much with it. I'm kinda looking for something that has numbers I can look at that will give me an idea of the brightness strength.

Does anyone have a suggestion as to what program I can get that will help me with this? Free is best, but I'd pay money for a good one.

[efinque]

I used to have a program that did key analysis of tracks and other information like length, ID3 tags etc but I can't remember it's name, not that it was of any use to me anyway... then there's Mixed in Key which some people have used with good results.

The brightness&dim which you mentioned roughly correlates as energy levels; something that psy-trance event organizers quite effectively make use of.

[David Bowman]

I still use my ears, but here you go. Keep in mind this software is mono only.

http://www.klingbeil.com/spear/


SPEAR is an application for audio analysis, editing and synthesis. The analysis procedure (which is based on the traditional McAulay-Quatieri technique) attempts to represent a sound with many individual sinusoidal tracks (partials), each corresponding to a single sinusoidal wave with time varying frequency and amplitude.

Something which closely resembles the original input sound (a resynthesis) can be generated by computing and adding all of the individual time varying sinusoidal waves together. In almost all cases the resynthesis will not be exactly identical to the original sound (although it is possible to get very close).

Aside from offering a very detailed analysis of the time varying frequency content of a sound, a sinusoidal model offers a great deal of flexibility for editing and manipulation. SPEAR supports flexible selection and immediate manipulation of analysis data, cut and paste, and unlimited undo/redo. Hundreds of simultaneous partials can be synthesized in real-time and documents may contain thousands of individual partials dispersed in time. SPEAR also supports a variety of standard file formats for the import and export of analysis data.

[Not Applicable]

I like Mixed in Key and Platinum Notes.

[steveryan]

I still use my ears, but here you go. Keep in mind this software is mono only.

http://www.klingbeil.com/spear/

Sweet. I'll def check that out. Do I need a DAW? I downloaded Audacity.

[steveryan]

I like Mixed in Key and Platinum Notes.

I'll probably pick up MIK soon just for the energy analysis. I know scales and how they relate to each other, so I won't be using it for harmonic mixing. I'll check Platinum Notes also. Sounds good.

[pete]

There are two routes to follow here:

1:/ Spend time and energy finding the free software to analyse key. Which may or not be accurate. Get the software to edit the Mp3 tags of your collection with the key/genre/energy/colour-of-artist's-underwear/whatever. Set up a system where it almost automatically writes over the tags of your collection. Don't spend time actually mixing the tracks and hope it goes well on the night. End up with Mp3s with messed up tags that you frantically try to remember which key mixes well with which other key at which pitch level.

2:/ Learn your tracks inside out the old fashioned way. Practice actually mixing the tracks in sets that flow well. As a bonus you also learn the song structure as well. Play them out live, learning how the crowd reacts. Spend more energy hunting out the rare gems that the crowd regards as "your tracks" and the blends that the crowd regards as "your mixes", that they appreciate at your gigs.


Needless to say, I have tried both methods. You can probably guess what I ended up using.

[KLH]

Learn your tracks inside out the old fashioned way. Practice actually mixing the tracks in sets that flow well. As a bonus you also learn the song structure as well. Play them out live, learning how the crowd reacts. Spend more energy hunting out the rare gems that the crowd regards as "your tracks" and the blends that the crowd regards as "your mixes", that they appreciate at your gigs.
This is what most DJs will say is what 80% of being a DJ is - always picking the most appropriate song to put on next. The other 20% is style and promotion. Without that first 80%, most would say not to bother with anything else. I agree.

[steveryan]

This is what most DJs will say is what 80% of being a DJ is - always picking the most appropriate song to put on next. The other 20% is style and promotion. Without that first 80%, most would say not to bother with anything else. I agree.

That's pretty much the key to the kingdom. The better you are at that, the more people will love what you play.

But let's talk about the methods we use for choosing those tracks. It's always been the ear. We listen to a song, and sometimes it reminds us of another song; so we try to mix them together. Sometimes we can make it work: sometimes not. Other times we just flip through records until we find something to try. Or we just grab a bunch of tracks were gonna play at the party and then put em in some kinda order.

What I'm trying to do is choose 20-30 "Similar" tracks for my set, organize them by energy and play them in order from dark to light. If I have a few tracks in the middle of my set that are similar in energy, use the brightness to help determine what order I might play them in.

My current method is as follows:

1. Pick one track that I really, really like.
2. Choose 20 or more other tracks that have a similar sound or presence.
3. On paper, write down the names of all the tracks.
4. Give the song that I really like an energy rating of 10.
5. Listen to each of the other tracks I selected and give each one an energy rating according to how much higher or lower the energy is that the main track at 10.
6. Order all of the tracks according to energy from low to high.
7. If multiple tracks have the same energy, figure out what order to play them in and whether to exclude any of them from the set.
8. Pick out an intro track and a conclusion track or two.

So now, what I'm trying to do if figure out how to include brightness in all of this to help me better organize the set.

I downloaded Audacity and the Voxengo SPAN spectrum analyzer today. SPAN is way better than the free one I was using last night. I can isolate the spectrum to show only the frequencies between 2,000 and 10,000 hz, and then have it display only the average in that range.

[efinque]

This is what most DJs will say is what 80% of being a DJ is - always picking the most appropriate song to put on next. The other 20% is style and promotion. Without that first 80%, most would say not to bother with anything else. I agree.

This kind of 80/20 division is very accurate.. however, the 80% is somewhat technical and can be mastered in a few years; the rest you can't learn by reading books or watching tutorials.

I think promotion takes a lot of time and commitment.. style is something that develops over years, and to master that one could start by reading biographies of famous/successful people (preferably ones that are objectively written)