Double your ESX memory without buying or installing RAM!:
Take the audio file in a DAW (I use Logic Pro) and after you have imported the audio you want, change the session sample rate to 88.2kHz. This should make the samples one octave higher and twice as fast. Then export the audio file as a 44.1kHz file. The samples can either be a one-shot hit, or a stream of hits which you can later slice up in the ESX into individual hits. When you import the sample, make sure to go in and change the “SAMPLE TUNE” setting to put it down an octave, as it should have pitched everything up one octave when you converted them into 88.2kHz. This works best with low-frequency samples like kick, bass, snare, toms, or if you want a “lo-fi” 22kHz sound. If your digital audio program does this different, basically you want to make the sounds play twice as fast, preferably without doing any processing to them. In theory you can pitch things us 2 octaves, making them 4x as fast, and quadruple your memory, but it would make your samples be at 11kHz which is quite lo-fi. Using this technique, I can record a long section of multiple hits which I can slice up later so that I have one “Stack” of a particular type of hits. This way I only use up 1 voice slot on the sampler, but fit a lot in like a bunch of 808 kicks or whatnot. Daniel Kerr shows how to slice them in his tutorial here: YOUTUBE
Stretch mode setting length to 128:
This is a nice little effect for getting some “granular synthesis” type of sounds. Basically, take a very short sound with lots of high-end energy and make it even shorter by adjusting the start and end points in the sample editor section of the ESX. Make sure to truncate the sample so that it is truly only the short part you hear, instead of being a longer sample which you only hear a portion of. Then in the Sample editing section of the ESX, go to “Stretch Step” length and set it to 128. This will now make that tiny, short sample play back over 8 bars time. You can get some interesting sound effects and eerie effects. This is cool to then take and run through a filter with max resonance, which can be tuned to a pitch to create some interesting tonal textures. The part can also be output to the physical output #3 or #4 and run back into the Audio Input in order to add some trance gating by sequencing the “Audio In” track. Filter modulation or effects on top of the pitched texture. COOL.
Shuffling breakbeats with 8 bar stretch part:
This is a great sampler and sequencer, but many people complain that you can’t rearrange the order of “slices” like you might with a REX file. If you use the Drum or Stretch tracks instead of using the Slice tracks, you can play with motion sequences to get roughly the same type of sound. To get the most accurate and easiest to automate loop, you will want to select a loop which is 8 bars long. The reason for this is that the Stretch track need to know how many Stretch points in the track to keep pinned to the beat. If you use an 8 bar loop, this works out to 128 points.
Now all you do is set the motion sequence to “trig hold” and play with the start time knob. When dealing with motion sequences in the “trig hold” mode it offers a resolution of 128 values. Bingo! What is nice is if you have an 8 bar loop with hihats or other sounds playing on each 16th note, each one of the values of motion sequence will automatically align with one of the 16th notes. (You can use shorter loops, but to get everything to play properly you have to go in after the face and use step edit to alter the motion sequence value for start time to divisions of 128. For instance if you have a 1 bar loop in the Stretch part, the values need to be 0, 8, 16, etc. between 0 and 127. This makes each start point align with each 16th note of the audio sample.
Daniel Kerr has some great tutorials on Youtube if you haven’t already checked them out. He is the ESX guru. He create a great tutorial for making Glitch drums here
Side-chain compression using LFO saw:
This one might seem a little obvious, but I didn’t think of it right away. If you want to do the old “side-chain compression” trick with your beloved ESX, but don’t know how, there is a tremendously simple way. Use the Modulation section on the synth/pad you want to be pumping and set the speed to BPM SYNC and 1/4 note. Set the shape to saw wave and the depth to a negative setting and the destination to AMP. Depending on the depth, it should now swell on each 1/4 note so if you have a four on the floor kick pattern it will appear that the synth/pad is being side-chain compressed by the kick. If your kick pattern changes, you can use motion sequencing (automation) to automate the speed to coincide with the hits. (For instance you have a kick on 1, 3, and 4: set the beginning tempo of the Modulation at 1/2, then on beat 3 set it to 1/4).
Using the BPF+ as an EQ:
Another easy one. I discovered how cool the BPF+ is. It basically adds a certain resonant frequency back into the full frequency signal. This is cool on kicks. Turn the resonance up a lot, and the cutoff towards the bottom and you should be getting some cool low-end resonance on the kick. It is like a single band Parametric EQ.
Using sample length to determine pitch:
This is a little obscure, but useful. If you have any sample with a length of 675 ( or multiple of) in the ESX and you load it into Keyboard mode, you will get the note “C”. I discovered this because of the Adventure Kid samples. All of those samples are the sample length, yet when you look at some in a DAW, they are often not perfect sinewave type samples. Some were more noise-like. I find this useful because you can grab any snare, noise or other sound and chop its length to 675 and you won’t need to get a tuner. The other main reason this is useful is that if you are sampling a synth and you sample yourself playing the note “C”, you shouldn’t need to spend too much time in the ESX finding the loop point. As long as the “Loop Start” is 675 samples before the “End”, it should loop properly without glitches. This is super handy for re-sampling sing cycle waves as well. I often layer two bass notes, then resample the tone, chop and truncate it to 675 and I have a nice new bass synth tone to work with. Like layered oscillators, although you can’t detune then and resample a single cycle. Note that half the note length equals one octave higher and double the length equals 1 octave lower.
Other note lengths I have found are:
Set the length of hits to a note length in the Stretch Step:
This can be useful when you are trying to create cool time sync’d hits. For standard snares and stabs it doesn’t make as much of a difference, but if your stab has any bit of pulse to is or swell from beginning to end, putting it into your stretch track will make it’s tempo sync to the pattern.
(An aside with this that is nice is that you can set up only certain hits to be Stretch Samples, which makes only those sounds available to be assigned to the Stretch Part. Meaning, you can create a simple pattern and in real time, choose different time stretched hits to play during playback. I find this cool when doing Glitch Mob style “Space Braf” hits on beat 4. Check out my post here with some free downloadable space barfs. )
I don’t have a Kaoss pad or hands on effects processor, but I like to manipulate the sound of the mix, like a LPF sweep on the whole mix. The ESX effects processor can do this, but every sound needs to be assigned to one effects processor set to LPF. Seems silly when each part has a filter, even the audio in. Well, fear not, there is a way. Take your Audio In track and assign its output to 3-4. Now plug in output 3-4 to your sound system. Now take output 1-2 and use a short cable to plug it back into the ESX Audio Input. Essentially your whole mix goes out 1-2 then back into the ESX and plays “through” the Audio In track. Now on the Audio in track, make sure to set your pattern to play a note at the beginning of each bar with a Gate Time of 16 for the first beat. This will let the audio go through unaffected, but you can use the Roll, Filter, Modulation and Arpeggiator to do some cool live manipulation of the entire mix, and you can use motion sequencing on this Audio In track as well. The alternative to this, which is sometimes easier would be to just re-sample the pattern and assign it to a pad, but then you are locked in to one performance. With this tip, you can still alter the pattern and all tracks in real time.
This one is simple and I may have just missed it in the manual, but instead of transposing a part by holding the transpose button and turning the knob, you can hold the transpose button and select the note on the 16 step keypad. It seems that 0 transposition is equal to the note C, so anything you select up or down on the keyboard transposes it. Up or down 2 octaves!
Figure out how long your loops are with MATH!
You can calculate the length of loops using the BPM calculator HERE halfway down the page you will see a calculator to find the sample length of a whole note at 44.1k. This information along with the fact that the Korg has an approximately 1300 sample delay when resampling has made it possible to truncate the beginning and end of my sample by numbers alone perfectly.
So far that exhausts my knowledge of the ESX and any tips that I have, but I have only owned it a few weeks. More tips might be on their way.
P.S. comment or subscribe if you like my tips.
P.P.S. Got some new ones here, with video
You can play “yo-yo” with your sounds, and use motion sequences to stretch a sample to play as long as you want. The video uses percussion sounds, but it can apply to synth type sustained sounds being played on the “one-shot” pads…
Another video here for you about sample start modulation. REX’ify your loops and re-organizing them.
Another video here about Effects manipulation using Motion Sequences.
Another video here about using Motion Sequences of the Effects processors on the Audio In track.