This question is about the classic 1980s drum machine known as the LinnDrum (LM-2) by Linn Electronics.

Some drums have multiple buttons assigned to them. I'm looking for detailed information about what different buttons do. Apparently, they affect the volume of the drum sounds. But how so? The user manual (available here) contains the following information:

PLAYING THE INDIVIDUAL DRUMS AND PERCUSSION: Drum buttons are situated in the lower left portion of the front panel. The printing below each button describes the instrument you hear when that button is pressed. Some instruments have more than one volume, permitting programmable dynamics: SNARE has three; BASS, HIHAT, CABASA, TAMBOURINE, and RIDE CYMBAL all have two. The higher the number, the greater the programmed volume. [...]

Asked more precisely:

  • What is the difference in dB between the different volumes?
  • In addition to volume, is there an audible difference between the different buttons? (E.g. distortion or aliasing caused by the internal processing of the machine)

I do not have access to the machine itself, only to a number of sample packs. Unfortunately, it is never mentioned which button triggered the recorded samples. Any information would be much appreciated.

By the way, I do not have any commercial interest in this information. I'm just curious about this iconic instrument.


I can't provide a full answer, but from recollections of the time.
I never personally used a Linn2, I only had the Linn1 then jumped to the Drumulator when that was released.

Each button directly accesses a distinct sample, of increasing velocity/loudness, so the snare would go from a light tap to a full hit. Similar for the others. the hat would have 'light', 'heavy', 'open'.

The reason for one button per velocity was, it didn't have velocity-sensitive buttons… that simple.
They existed on some keyboards, but I'm guessing they couldn't yet make them small enough to get in a drum machine.
Either that or they were trying to hold inside that 3 grand price tag & it would have pushed it too far over. There's always a economic component to any decision like this, & in the early 80s a 'good synth' cost three grand, which they would be very aware of.
The first drum machine I'm aware of with actual velocity-sensitive buttons was Emu's Drumulator, in 83. After that, everybody had to follow suit.

I've no idea what the actual level differences are between samples - I never thought to measure them - they were just as natural-sounding as possible, because you weren't really doing much additional processing to them at the time. They had to work right out of the box. A bit of EQ, maybe a little light comp if you had them to spare. Such as the Valley People comps were really new, something fast enough for drums hadn't really existed much before that.

I also couldn't tell you about any additional distortion. We always assumed there was 'none' for a given meaning of 'none', as each sample was levelled inside the machine already, so the loudest wasn't hitting the reds, the softest came in way below. These were I think 8-bit devices, very probably with analogue amps at the end of the chain. They sounded how they sounded & no-one ever thought to measure them. They were musical instruments not scientific devices.

  • Thanks for the insights. Everything you wrote makes perfect sense to me. I wouldn't have imagined it any other way. I'm aware of the scientific nature of the question, as opposed to the musical nature of the instrument. It boils down to this sentence: "They had to work right out of the box." I would simply like to know how to set the volume (and pitches) of my samples so that they sound as "out of the box" as possible. I'm really puzzled why the creators of the sample packs didn't think of this.
    – Streck0
    Jan 2 '21 at 15:52
  • My best (& probably most cynical) guess would be that whoever sampled them tweaked every sample to normalise it & forgot they should be at apparent levels, not absolute for their user-base to get the most immediate use from them, without having to ask what levels they should be at ;)) People used to modern normalisation techniques probably don't think the same way as Roger Linn, who wanted his 3 grand box to just be playable.
    – Tetsujin
    Jan 2 '21 at 16:09
  • True. And I'm not even saying that I'm unhappy with the sample quality. It's just the marketing texts go on and on about the meticulous attention to detail in the sampling process while it's clear that the only thing they cared about was proper recording equpiment. Obvious things like round-robin samples and default settings of the machines are mostly ignored.
    – Streck0
    Jan 2 '21 at 16:30
  • I'd be tempted to ask whoever made the sample set - they would have the original data… possibly even recorded at relative levels so the op amps are at parity; but who knows, maybe they tweaked levels for every recording… which would be 'smart' on the one hand '& dumb as a bag of rocks' on the other ;))
    – Tetsujin
    Jan 2 '21 at 16:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.