Making the Most of your Akai S1000 Sampler (Part 2)
Sample Editing & Manipulation
PART 2: Akai product specialist Steve Howell passes on more useful tips to help you make the most of your new S1000 sampler.
Last month we learned how to get some samples into the S1000 other than by loading them from disk. This month, we'll explore what we can do with those raw samples. The first thing you'll probably want to do with any sound that you wish to sustain for any length of time is to loop that sound. Ordinarily, the mere mention of looping sends even the most die-hard samplist reaching for the aspirin - but I am happy to report that the S1000 makes life as easy as possible thanks to its on-screen visual editing. OK, I'll admit that the on-screen editing with the S1000 is not ultimately as powerful as using an external editing program on a computer, but it sure beats the hell out of waiting to transfer samples back and forth between the computer so that you can hear the fruits of your labours. Personally (and I am sure there will be some who disagree), I find the on-screen visuals of the S1000 more than adequate. In fact, all the library disks have been created using the S1000 alone, partly because there is no computer editor for the S1000 (yet!) but also because I feel that it's only fair to give sounds to our customers that can be realistically edited on the S1000 itself without having to resort to external devices. It really galls me to use library disks on a sampler that were beautifully knocked into shape on a computer editor but remain totally uneditable within the confines of the instrument itself - it smacks of a bit of a con to me. But that's just my opinion. Let's move on...
So, assuming we have a sample already in our S1000, let's start editing. At the time of writing, the basic editing facilities available within the EDIT SAMPLE pages are as follows:
In ED1 (F4):
TRIM (F2) - to cut the start and end points of a sample.
LOOP (F3) - to set up to eight loops within a sample.
JOIN (F4) - to splice two samples together with variable crossfade times.
In ED2 (F5): Tuning and re assigning base pitch, plus loop mode.
TIME (F3) - Timestretch; the ability to stretch a sample by up to 1000% or to shrink it down to 25% of the original sample length without a pitch change.
New sample editing facilities currently under development (but not yet implemented) include:
• Resampling at a variable bandwidth — sampling can still only happen with bandwidths of 10 or 20kHz, but you will be able to resample at a lower bandwidth to save on memory.
• Sample merge - the ability to mix two (or more) samples to create one new composite sample.
• Gain normalisation - sets sampling level to maximum after a sample has been taken to optimise signal-to-noise ratios and match sample levels.
For now, let us concentrate on the editing functions in ED1. To access any of these editing facilities simply press EDIT SAMPLE and then select F4 (ED1) - this will come up on TRIM. Here, as stated above, we can edit the start and end points of a sample. The S1000 automatically looks for the start of a signal, so start point editing is not always necessary, especially for keyboard sounds. But with drum samples it is often worthwhile to get in there and edit start points, otherwise you may find your drum parts sound 'lazy'. When you are in the TRIM page, you are presented with a visual display of the sample's amplitude. The default display is quite 'zoomed in', which is as it should be for detailed editing, and this can be changed as you wish using the ZIN (F5) or ZOUT (F6) buttons. For the time being, let's stay with the default.
Having selected the 'start' field with the cursor knob, you can now use the data knob to change the start point. Notice how the vertical bar in the display moves as you change values. The intention is to line up this bar dead on the attack transient.
Fortunately, you can hear the effect of this editing process by playing the keyboard, so you can combine your ears with your eyes to achieve the best result.
When editing sharp attack transients (especially from drums and percussion) play the sample several octaves lower on the keyboard, as you will hear any delays more easily when sample playback is slowed down so much. By doing this and checking the screen, you can ensure that your samples start dead on time and so don't affect the 'feel' of a track.
Having successfully edited the attack portion, we can now move to the end of the sample. To view the end you can either zoom out to maximum (ie. press F6 several times) or you can select F7, which will toggle the display to show the end of the sample. Next, move the cursor to the 'end' field and adjust the end point using the data knob. Once again, the vertical line will show you exactly where you are in the sound - you ideally want to edit out any rubbish (if any) after the sound.
Having done that, press F8 (CUT). This will discard anything before or after the start and end points you have set thereby freeing up more sample memory for further sampling.
Of course, if your sound is a sustain type of noise, you may not want to edit the end point right now. Alternatively, you may wish to head straight for the LOOP page and do all your editing afterwards - so let's do that.
Pressing F3 accesses the LOOP page. The S1000 display shows the whole sound on the left-hand side, with the actual loop point shown on the right. When creating a loop, you should be watching the left-hand side of the display most of the time looking for good loop points. What is a good loop point? Let me explain...
When setting a loop, the start and end point of that loop must ideally match in amplitude (see Figure 1) so that the transition between the start and end of the loop is relatively smooth. Figure 2 shows what happens if the start and end points don't match up too well - you get a glitch. On the S1000, by referring to the left-hand part of the display, you can see where the best place should be for a good sort of loop.
Having said all that, it's not quite so easy. You see, all we can do with our sampler is match up the overall amplitude changes, but any sound will also have tonal changes within those amplitude envelopes. Even if we can match up amplitude changes perfectly, there may be tonal discrepancies between the start and end of a loop which will also cause a glitch. If a sample won't loop, it is usually the fault of the sample and not the sampler (either that, or you're not doing it right!!). To help, the S1000 has two features, namely autoloop and crossfade looping, which make looping relatively painless.
To set a loop you must first turn the loop on, and this is done by moving the cursor to the 'time' field and setting a value of 9999 on the numeric keypad. This will display HOLD, ie. the loop will last as long as you hold the note down (multiple loops will be explained another time). Now move the cursor to the 'at' field. This sets the position at which you want the loop to occur (this is represented by the right-hand moving vertical line; there are two). You ideally want to move this to a suitable point towards the end of the sound where the amplitudes match (although it can be anywhere you like within the sample). Next, move the cursor to the 'loop length' field and set the loop length so that the left one of these two vertical lines is at a point of amplitude level that is similar to that set by the 'at' field (see Figure 3). You can see whether or not the loop transition is smooth by referring to the display on the right of the screen. By playing the keyboard, you will also hear your loop.
Now in all fairness, this simple procedure is unlikely to produce a perfect loop straight off (although it has happened to me on more than one occasion!), so you can now call upon the autoloop function to assist you. Press F7 (FIND) and keep playing the keyboard. In an ideal world this should produce a smooth loop, but if it doesn't, then press FIND repeatedly until a good loop is found. (By the way, you don't have to repeatedly play the keyboard each time you press FIND or move loop points manually because loop setting on the S1000 is in real time, and you will hear the new loop points as you hold the note down).
If you manage to find a good loop point like this, all well and good; but if nothing worthwhile is happening, you can use crossfade looping. Move the cursor to the 'Xfd' field in the display and set a suitable crossfade time, then press F8 (XFD) to initiate the crossfade loop.
As with last month's episode, the operations outlined above are more difficult to explain than they are to do, so what follows is a potted set of instructions to create a loop.
1. Access the LOOP page by pressing F3 in ED1.
2. Set the loop to HOLD.
3. Move the 'at' point to a suitable position.
4. Set the loop length to a suitable position.
5. Press FIND until a decent loop is found.
6. Set a suitable crossfade time and press F7 (XFD).
Done! But there's an even easier routine. You see, the S1000 automatically searches for the best 'area' of a sample in which a loop is going to be effective. Nine times out of ten, simply going into the LOOP page, setting the loop to HOLD and pressing FIND a few times will give you a good loop almost instantly. If you dispute my claim, bear in mind that it stems from nearly a year's experience of sitting in front of the S1000 almost every day, creating loops! Admittedly, not all are perfect - but most are; and the S1000 is certainly the easiest sampler on which to set loops I have ever used (and I have used a few, believe me!).
Assuming that you have set a decent loop, you might like to go back to the TRIM page to edit start and end points of the sample. Very often, the loop may well occur somewhere in the middle of the sound and so you can cut and discard the portion of the sample after the loop point (see Figure 4). This will free up more memory for further sampling.
What about JOIN? This allows you to merge two samples in a crossfade. For example, you can have a bell sound crossfading with a string sample over a defined period of time. This is easily done and is well explained in the manual, so I won't dwell on it too much. Suffice to say that in the JOIN page, 'A' selects the first sample, 'B' selects the sample you wish to crossfade sample 'A' with, and 'J' is the composite result (which you name using the usual naming functions). Then you set a suitable crossfade time in the 'X-fade over' field and press F8 (GO). After a short while, you will be able to play the result from your keyboard.
If you want to crossfade three samples, there's no problem. Simply use your new composite sample as the basis of another crossfade and join that with a third sample. This process can be repeated endlessly until you have some quite spectacular 'evolving' sounds that would put many a D50 or M1 preset to shame!! The only limitation here is that each new composite sample you create uses up sample memory, which you may run out of. This is not a big problem because, once you've used the source samples, these can be deleted to make room for your new creations. Another word of warning: any loops set in the original samples will be lost when you crossfade them, so your new mega sample will need to be re-looped.
Such samples lend themselves well to multiple looping, and this is something we will discuss next month. Until then, happy editing.
Gear in this article:
Feature by Steve Howell