Journey of a drum & bass track: part 4
In this series Ben Kama takes us on the journey of a drum and bass track from start to finish. Throughout the process he will explain everything he does, the instruments and VSTs he uses, and share with us the various highs and lows of putting a piece of music together.
Follow the entire series here:
By the end of the last entry the tune was starting to sound like... well... a tune. In this entry Ben experiments with some vocals to try and add some layers to the ensemble.
Entry #4: I am not a machine
I must admit it - I cheated a little. I worked on the track just a little yesterday and didn't write about it right after. I checked out the project, just meaning to listen to it again, but I fell into a trap and did some minor changes. The original bassline was good, but this track was taking another direction so I toned down the gritty bits on the bass even more and re-introduced the vocal twist to end of the current clip. Also made a few very slight changes to the drums to avoid repetitiveness and some bigger changes to how the speech sample was processed. Here's a clip that also serves as a reminder of where I left off last time (After uploading I noticed the vocal was very loud in the track, toned it down for the final clip):
The track sounded promising, but to keep it interesting for 4-5 minutes it needed something more to take it beyond the grey mass of mediocrity, and that gets us to today's session. I had an epiphany one night about making my own vocals to the track. It was late and I had to get up at 6 for work, but inspiration usually strikes at the most inappropriate moments so I just scribbled the idea down to a piece of paper. The 2nd and 3rd entry in the notes were added the next night. The whole idea was to deepen the "I am not a machine, I am a man" theme and to give the track some meaning. In this case I wanted to give listeners some idea about self-aware machines. Something very futuristic and sci-fi as is appropriate with drum and bass.
So now when I had time I dusted off my mic and started recording. It took about 5 or 6 takes to get the first half of the written "vocal" down nicely. That's quite much considering there's nothing there in terms of tuning, but like most people, I haven't gotten used to hearing my own voice from outside my head...
It took me a long time to glitch and cut the vocals like how I had "heard them in my head" at the moment of the late night inspiration. I cut the stuff up to pieces, edited some clips in an audio editor (distorted, flanged, fade ins etc) and made a lot of audio editing to the recorded samples. A lot of the inspiration and ideas came from the game System Shock 2, especially the crazed artificial intelligence Shodan. The vocals are very short, but one of FL Studio's shortcomings is it's audio editing capabilites. You can get the job done, but it takes a lot more time than in for example Reaper, which is a DAW I use a lot too. In fact, I do most of the audio editing on my tunes in Reaper in the mixdown phase - but this time I wanted to do it in FL Studio as I want to be able to change the sounds and processing if I want.
If you listened to the first clip in this journal, you might have noticed the first speech sample sounds different from what I originally used. Here's a comparison of the two:
The original sound:
The new sound:
At this point I tried desperately to use my shitty Nokia to record some video to show you how the vocal snippets were processed but it kept on stuttering. Somehow I'm not very amazed at the state of that company at present. So you'll have to do with a written description. (If anyone knows a good free screen video capture software please leave a comment to this post or get in touch on benjamin.kama@gmail.com -ed)
The original vocal processing relied on multiplying the source and using random effecting on it, like bitcrushers, filters and pitchshifting. This new one however takes use of self-modulation. The source is practically the same as before, the same sample three times, but this time a pitched down sample is used as a vocoder carrier signal and 2 slightly less pitched sounds mixed together form the modulator signal. In Image-Line Vocodex you can do a L-R encoding of the source, meaning anything panned right will be the carrier signal and anything panned left will be the modulator. Or the other way around, I can't remember. In any case it's very useful with mono sources as you don't have to fool around with sends at all.
One thing I did on one of the 3 sources was timestretch it to twice as long with a very simple algorithm (sounds like those oldschool jungle timestretches), save the result as a new sample and then shorten it back to the original length. So I was actively seeking any artifacts to the sound. Other than the vocoder, the channel still has a bitcrusher to bring up some edge, an EQ to damp extreme sibilance, a slight chorus and a reverb. It's also layered with the unprocessed snippet to give it more clarity.
Here's the whole result of the processing:
Unfortunately that is all I had time for today. Listening back to the vocal snips I'm torn two ways. On the other hand the voice sounds really nasty and good, but on the other hand it doesn't really suit this track - it would be more at home in some minimalistic bass science tune. I have much doubt about continuing with this direction and I'm pretty sure I'll scrap the idea in the next session and just go for some old fashion sampling. I'm still going to hold on to this voice and maybe I will use it in another track that suits the theme better.
But that's music for you, you can't hit the gold vein everytime.
More about Ben Kama
Follow the entire series here:
- Part 1: The pain of creation
- Part 2: The beat generation
- Part 3: Makeover
- Part 4: I am not a machine
- Part 5: Hook and line
- Part 6: Organized chaos
By the end of the last entry the tune was starting to sound like... well... a tune. In this entry Ben experiments with some vocals to try and add some layers to the ensemble.
Entry #4: I am not a machine
I must admit it - I cheated a little. I worked on the track just a little yesterday and didn't write about it right after. I checked out the project, just meaning to listen to it again, but I fell into a trap and did some minor changes. The original bassline was good, but this track was taking another direction so I toned down the gritty bits on the bass even more and re-introduced the vocal twist to end of the current clip. Also made a few very slight changes to the drums to avoid repetitiveness and some bigger changes to how the speech sample was processed. Here's a clip that also serves as a reminder of where I left off last time (After uploading I noticed the vocal was very loud in the track, toned it down for the final clip):
The track sounded promising, but to keep it interesting for 4-5 minutes it needed something more to take it beyond the grey mass of mediocrity, and that gets us to today's session. I had an epiphany one night about making my own vocals to the track. It was late and I had to get up at 6 for work, but inspiration usually strikes at the most inappropriate moments so I just scribbled the idea down to a piece of paper. The 2nd and 3rd entry in the notes were added the next night. The whole idea was to deepen the "I am not a machine, I am a man" theme and to give the track some meaning. In this case I wanted to give listeners some idea about self-aware machines. Something very futuristic and sci-fi as is appropriate with drum and bass.
So now when I had time I dusted off my mic and started recording. It took about 5 or 6 takes to get the first half of the written "vocal" down nicely. That's quite much considering there's nothing there in terms of tuning, but like most people, I haven't gotten used to hearing my own voice from outside my head...
It took me a long time to glitch and cut the vocals like how I had "heard them in my head" at the moment of the late night inspiration. I cut the stuff up to pieces, edited some clips in an audio editor (distorted, flanged, fade ins etc) and made a lot of audio editing to the recorded samples. A lot of the inspiration and ideas came from the game System Shock 2, especially the crazed artificial intelligence Shodan. The vocals are very short, but one of FL Studio's shortcomings is it's audio editing capabilites. You can get the job done, but it takes a lot more time than in for example Reaper, which is a DAW I use a lot too. In fact, I do most of the audio editing on my tunes in Reaper in the mixdown phase - but this time I wanted to do it in FL Studio as I want to be able to change the sounds and processing if I want.
If you listened to the first clip in this journal, you might have noticed the first speech sample sounds different from what I originally used. Here's a comparison of the two:
The original sound:
The new sound:
At this point I tried desperately to use my shitty Nokia to record some video to show you how the vocal snippets were processed but it kept on stuttering. Somehow I'm not very amazed at the state of that company at present. So you'll have to do with a written description. (If anyone knows a good free screen video capture software please leave a comment to this post or get in touch on benjamin.kama@gmail.com -ed)
The original vocal processing relied on multiplying the source and using random effecting on it, like bitcrushers, filters and pitchshifting. This new one however takes use of self-modulation. The source is practically the same as before, the same sample three times, but this time a pitched down sample is used as a vocoder carrier signal and 2 slightly less pitched sounds mixed together form the modulator signal. In Image-Line Vocodex you can do a L-R encoding of the source, meaning anything panned right will be the carrier signal and anything panned left will be the modulator. Or the other way around, I can't remember. In any case it's very useful with mono sources as you don't have to fool around with sends at all.
One thing I did on one of the 3 sources was timestretch it to twice as long with a very simple algorithm (sounds like those oldschool jungle timestretches), save the result as a new sample and then shorten it back to the original length. So I was actively seeking any artifacts to the sound. Other than the vocoder, the channel still has a bitcrusher to bring up some edge, an EQ to damp extreme sibilance, a slight chorus and a reverb. It's also layered with the unprocessed snippet to give it more clarity.
Here's the whole result of the processing:
Unfortunately that is all I had time for today. Listening back to the vocal snips I'm torn two ways. On the other hand the voice sounds really nasty and good, but on the other hand it doesn't really suit this track - it would be more at home in some minimalistic bass science tune. I have much doubt about continuing with this direction and I'm pretty sure I'll scrap the idea in the next session and just go for some old fashion sampling. I'm still going to hold on to this voice and maybe I will use it in another track that suits the theme better.
But that's music for you, you can't hit the gold vein everytime.
Next: Part 5: Hook and line
More about Ben Kama
Post a comment
Please login to post a comment.About the author
|
