ROM machine STRT parameter exponential = aaargh

Hey everyone, I’ve encountered a (hopefully) interesting issue… I tried to be cheeky and load “sample chains” into my MD, with the hope that I could essentially plock the STRT parameter to get several sounds on the same channel, thinking that as long as they were evenly spaced, and grouped as square numbers, then they’d correspond directly to the rather coarse parameter settings…

However, it didn’t quite work, and I couldn’t even figure out why, until I looked at the manual just now:

STRT - Controls the start point of the sample. The parameter
is exponential, giving a very fine control over the beginning
of a sample.

This certainly threw a spanner in the works! My evenly spaced samples clearly don’t work with this method. Is anyone emulating sample chains with MD? If so, how are you doing it? Or is it all too difficult and I should just get an octatrack/digitakt instead?

It’d be nice to see a graphical representation of what the STRT parameter corresponds to, or even a table where each parameter setting corresponds to a percentage…

I think STRT is exponential for some machines, and linear for others. I forget the ordering exactly but check the manual, its like the first half of those machines = exponential STRT, the second half = linear


oh my god you are totally right, I kept reading in the manual and it says it right there… dang! Will try that tonight, thanks!


So, would it be possible to make a sample chain grid with that very annoying exponential START? :slightly_frowning_face:

1 Like

Hi,the first 24 slots on rom samples are for short sounds,and the next 24 for llop sounds


Exactly. The first 24 rom slots are exponential and the last 24 rom slots are linear.


So doing Breakcore with a sampled amen break would be easier to know start times for samples by loading the samples in the last 24 slots because they are linear, correct?


…in theory
In reality, the amen break is not PERFECTLY on the grid, so you won’t be super precise like that. You could throw it into Ableton or something and quantise it to prep it for the machinedrum and then it would work perfectly.

But yes the idea is that the last 24 rom slots are able to cut up by changing the start point in a linear way.

Yeah, using a trained ear I suppose would be the best way to make sure it’s aligned correctly. Thanks!

1 Like

That’s what I did for OT with Samplitude. :wink:
Time stretch for each beat.
Amen Break seems to work for MD with 24 last slots, thanks for that great information.

Btw I couldn’t use a 8 secs 128 sounds sample chain with equal length. Not accurate. Maybe there is a max length? Special export settings?

I used C6 and a “stereo” 44.1kz 16 bit wav file made from OT. The sample chain works in OT with the 128 start points.

The ROM machines ROM-25 to ROM-48 operates slightly differently since they are optimized for loops. How the parameters differ from the other ROM machines is described below.

HOLD Set to a value of 127 the amplitude envelope is open for 2 bars. A setting of 1 will keep the envelope open for ¼ of a step (1/16 note)
STRT This parameter is now linear which makes it easy to find a position within a loop. In a 2 bar loop you will find each ¼ note located every 128/8 = 16 parameter steps. So for every 16 steps of the parameter you will find each beat in the loop.
END Also linear and functions basically the same way as the STRT parameter.


Seems to work with Hold. I was using End.
Thanks again for good news!

So could someone please clarify this for me. To make sample chains for the MD, a file should be prepared that has a number of evenly spaced samples and the number of samples should be divisible by 128? To select a different sample you just then p-lock the start parameter? Should the file be uploaded as a loop?

1 Like

IIRC there are 128 possible divisions for the sample but you can create a sample chain as you like, following a 128 evenly spaced grid, and set START / END as you like.
You can load a 4 sample chain for example.
Start 0, 32, 64, 96

IIRC again, loop doesn’t work with certain settings, but loading it with loop setting is not an obligation at all.


Thanks. So there is no limit to the sample length (memory permitting)?

Great expectation! :smile:
I didn’t try a sample with max RAM length, but it should work if it’s the only one.