I'm mainly talking about hard trance, electronic, rave, house, progressive, etc. from the 90s till the 2000. So no real recordings of instruments and stuff like a rock song

I wanted to know a little better how things were used in this period. I kinda know the gear and the setup; there's even a video on the internet called "How to make techno music in 1994" that shows Westbam and some other guys in their studios. The thing is I'd like to know a bit more.

For example. I know there were some basic, rudimental sequencers back then. How were they used? I'm guessing they were linked via MIDI to a synth and you could program a line on the computer the same way we do today with a DAW? Then you would record it or what? Or I find a good melody on my synth: ok, what now? Do I put the notes in the MIDI software? Or I record myself playing it and then put it in the song? What if I want some automations in it? I need to record the automations as well I guess.

Let's say I have my mixing desk with some elements. I have my bassdrum on 1, bass on 2, some cheesy vox choir synth on 3, a saw wave on 4, percussion and drums on 5 and effects on 6. I run my drum machine and the sound start coming out. Then the synth starts playing, and all the other elements. How does this all work together? Maybe I'm too used to thinking in the DAW way, but I'd really like to know.

A better way to put it is this. Today if I want to make something I make a pattern for the kick, a pattern for a bass, etc and then I lay down "the blocks" in the playlist. Back then there was no playlist, obviously. Or maybe there was, like a sequencer, or a workstation synth? In this case, I'd need to route every one of my instrument to it, right?

I'm really confused and the more I think about it the more I find ways things could have worked. So being a noob in this department I'm humbly asking you if you can tell me more about the real stage of production. Of course there is no general answer, I'm pretty sure everybody had their own way, but I'm looking for a generic answer. If you'll ask in the future, people will say "yeah we used daws, there were virtual instruments and patterns, bla bla bla" and you know.

But back then?

- cya

PS: I wish I started earlier to interest in music making...

 

Well, early '90s you could use a computer to sequence, a standalone sequencer, or an onboard sequencer. A computer sequencer worked pretty much like a modern DAW does, except just the piano roll and not the digital audio. An onboard sequencer would be one built into either a synth/sampler/workstation so you would arm a track and record your notes/patterns using the local keyboard/pads and then assign MIDI out channels to any other equipment you wanted to automate. A standalone sequencer is the same idea as onboard except it's more generic and not a fixed part of another instrument.

Once you started getting into the late '90s then you had all-in-one DAWs that were capable enough to handle digital audio and MIDI together in a modern sense. Once 2000 hit then it was the big bang and pretty much everything started working in the way people are accustomed to now, with VSTs and lots more in-the-box.

Yes, sequencers had pattern sequencing. You'd sequence a pattern and then you'd order the patterns into a song. Each pattern has 16 MIDI tracks and you would typically route one to each specific instrument.

The biggest difference is that things were less visual for the most part. Less painting things into a piano roll with a mouse, and more arming tracks and playing back things in realtime then (possibly) quantizing. Although, look at Cubase for Atari ST and you'll see that things could be done relatively close to today when editing with a computer based piano roll -- just lots more MIDI hardware.

 

Last edited: Aug 20, 2016

And the mixing desk was linked to the sequencer/MIDI software, right? Thanks from the answer mate
Well it isn't that different from what I imagined.
If anyone wants to expand on the subject feel free to!

 

The mixer wouldn't typically be connected to MIDI, no. Fader automation was usually done by hand. I mean, you could control the switching of FX patches and reverb levels and stuff via MIDI depending on your rack gear, but a lot of things had to be more hands on. You weren't going to adjust the gain reduction of your LA-2A compressor by painting a curve into the sequencer or anything. Mixers with motorized faders and that kind of thing were very expensive. It was a lot more hands on. Lots and lots of cables. Lots of setting things up and trying to minimize MIDI latency, etc.

You'd have a sequencer which branched out (either with a patch bay or via MIDI Thru) to your hardware synths/samplers and then the analog outs of the synths/samplers would go into a mixer. Then the sends of the mixer would go into your reverb and you'd use mixer inserts to slap some chorus onto your synths, etc.

It was a bitch. Everyone should be realize that 2016 is a piece of cake. Production is easy as shit, now.

 

So the mixing desk was one of the final parts, after the sequencing and the software? It was like all the outputs together and then from there, after adding the sends and all the other stuff, you would start finishing the song. And by finishing I mean recording or whatever was the method back then.

Yeah I bet. I grew up with this kind of music so I thought understanding better its production during that time would help me a little bit.
I guess saving a project was a harsh thing to do, lol.

 

Yes, the mixing desk is the final element that ties all the audio together and sums (mixes) it. From there it would have gone (typically) to ADAT, DA-88, 4-track cassette, or (if you were very lucky) multi-track reel-to-reel.

"Saving" a project could only be done so far as saving the MIDI patterns in your sequencer and (sometimes, depending on the equipment) sysex data or onboard patch data. Everything else was written down as notes to manually dial in later, or just recorded to tape and you were stuck with what you had recorded in that moment. A lot of times you'd come back to the project and nothing really sounded exactly the same again.

 

Last edited: Aug 20, 2016

What's even worse is that if you had live elements recorded over the top of your sequencing then you still couldn't really go back and change anything sequenced because the live elements would no longer line up properly if you tried to replace the sequenced parts once they were commit to tape. I mean, you could punch-in and replace stuff that way, or you could conceivably sync MIDI to your recorder through FSK, but that comes with compromises of its own.

 

Last edited: Aug 20, 2016

Thanks a lot for the exhaustive answers

 

If you want to experience the pain:

This is one way we used to get it done back in the day.

 

All signals were recorded track by track (First take Kick, second take Snare, third take HH ...). Most Keyboards/Sampler had no single outputs.
We had to sync the Roland MC 500 Micro Composer, later the Atari St with the Roland SBX 80 Syncbox to SMPTE timecode, recorded on Track 24 (or 48) on multitrack machine(s). To get the EMUII or Prophet V Samples - 4 example Kick and Snare - in time, we messured the (SMPTE) offset with a graphited-grease pencil on a 2 track recorder, the referencee signal was recorded (L/R).
For recall FX devices like Reverb, Delay ... we wrote parameter values on a sheet or made photos.
In 1990 I worked the first time on SSL E-Series console (a little later G-Series) with a POOR (estimated today) automation, it was sooo great - in the 90s.

 

From a performance POV, you could try by disabling as much multithreaded stuff in your DAW, and using ONLY FX which have no PDC, and every FX should us 4% of a core or less.

 

Atari ST, Cubase, AKAI S1000 some multi timbral synths!

 

FSK Frequency Shift Keying. I haven't thought about that in decades.

 

Think also of low-spec'ed Mac daisy-chained together to share system resources and/or dedicated to different processes. Also, looping and time-stretching of samples usually had to be done with regard to how long (in beat-increments, per tempo) the loop or sample had to be, and recording it at a speed which would double or halve accordingly in speed when slowed or sped-up X2, within the sampler unit. And this, for some historical perspective, from Wikipedia, regarding Pro Tools:

The first version of Pro Tools was launched in 1991 offering four tracks and selling for US$6,000. Digidesign continued to improve Pro Tools, adding a sequencer and more tracks with the system offering recording at 16-bit and 44.1 kHz. In 1997 Pro Tools had eventually reached 24-bit and 48 track versions. At this point, the migration from more conventional studio technology to the Pro Tools platform took place within the industry

 

In my opinion MIDI had the main part in those years. Programming MIDI was an art, more than now, and there were some applications ahead of their time, now sadly discontinued. One of them was Upbeat (Dr'ts / Intelligent Music, now Cycling74), programming a convincing drum solo was a matter of minutes.

 

Trackers!
Years ago I was obsessed with them and the videogame soundtracks (for example Unreal Tournament '99 used tracker music). This kind of software was the same as the MIDI software we were talking about earlier, right? You just linked everything to it, etc.? I thought tracker worked a bit differently

Thanks all for answers

 

They do work differently.

Trackers are basically a MIDI sequencer and a sampler combined together, so you're using a computer for sampling instead of an outboard hardware sampler. That's more conceptual vagueness than technological truth, but that's the gist. Some trackers would let you trigger external MIDI hardware, but they weren't really MIDI sequencers, per se.

The were a good way to operate if you only had a computer and no other hardware. Plus, it was really easy to share your songs because you didn't have to render down to WAVE/AIFF or record to tape, etc. They were self-contained modules that you could upload to a BBS (pre-Internet) where people could download them or send you their own songs. Kind of like sending a MIDI file and all the patches in your Ensoniq Mirage or Yamaha TX16W except the listener only had to have a computer with a soundcard, and not a MIDI interface and the exact same hardware sampler that you have.

 

A lot of techno/acid type music was actually done by placing some synth output on every channel of the mixer, like a 303's own sequencer, a 909, a 808, some juno or other synths sequenced through cubase or another sequencer. All synced from a device like this: https://www.flickr.com/photos/matrixsynth/sets/72157605344445584/ or cubase could be the master.
A lot of times people would remember at what place in cubase what pattern should be selected by hand on one of the other devices, or cubase was put in a loop and was never changed and the only thing that would change were all the other internal sequencers.
You could then mute all the devices by throwing faders open and closing them. You could send to fx sends by turning the send knobs. A really cool thing of the time was that people would make endless loops of FX sends creating all sorts of morphing feedback sounds. Sweep the EQs, cut the bass, sweep the midrange on saw sounds. Stuff like that.

In those days you would "play" the track live and record it. Play it 10 times and pick out the best recording for the release.

*edit: actually I know that there are hardware geeks that still work like this.. Mostly for acid like styles still..

 

Last edited: Aug 21, 2016

"Playing" the mixer and FX like that is also a large component in dub music.

 

I hope the auto subs work!

(I see that half the time or more the subs are utter bullshit, but at least it gives you an inside of the vibe of the early scene in Amsterdam)

@28:00 they work a little bit with just a sampler..

@8:21 is a typical setup

 

Last edited: Aug 21, 2016