Digital Stereo Sampler
Having set the music industry standard in sampling with their S1000, Akai are now taking moving pictures into their stride. Simon Trask gets frame-accurate on the S1100.
The demands being placed on the digital sampler are greater than ever. Akai's newest sampler represents the company's response to those demands.
THE S1100'S ENHANCEMENTS involve not only new software features but also superior circuitry, with 20-bit DACs providing improved s/n ratio and dynamic range. The S1100 also introduces onboard digital effects processing, something which is only otherwise available on Ensoniq's new EPS16 Plus sampler.
Perhaps the most significant new feature on the S1100 is the SMPTE cue list, which can be used in conjunction with the sampler's built-in SMPTE/EBU read/write capability to turn the S1100 into a machine well-suited to audio/visual post-production work. Also significant is the S1100's ability to address up to 32Mb of onboard RAM. The machine comes with 2MB of RAM fitted as standard, but you can upgrade the memory bit by bit - or rather, megabyte by megabyte - using 2Mb and/or 8Mb boards (EXM005 and EXM008). The full complement of memory gives you a maximum sampling time of 12 minutes 40 seconds (mono sampling at 22.05Khz), which means that mono 44.1kHz and stereo 22.05kHz give you six minutes 20 seconds, and stereo 44.1kHz gives you three minutes and ten seconds. Of course, you can use a mixture of stereo and mono, 22.05 and 44.1kHz, sampling - while if you use the S1100's resampling facility to convert samples to even lower sample rates (for special low-fi effects, perhaps) you can give yourself even more sampling time.
Like the S1000, the S1100 includes dedicated software for programming Akai's ME35T trigger-to-MIDI interface (in fact, up to two units can be independently programmed), which is good news for any drummers or percussionists wanting to play sampled kit and percussion sounds on the S1100 from pads via an ME35T.
The S1100's rear panel provides the complement of analogue and digital connections required on today's professional sampler. Thus you'll find Left/Mono and Right stereo audio outputs, eight polyphonic individual audio outs, a mono effect send output, a headphones output, a footswitch input, the standard trio of MIDI In, Out and Thru sockets, SMPTE timecode in and out sockets, a SCSI port (for hooking up hard disks, optical drives and CD ROM units), an XLR digital audio output for transmitting stereo audio data in AES/EBU format, and two expansion slots. The optional IB04 digital interface board provides both coaxial and optical digital links, allowing direct digital sampling at up to 48kHz and hard disk data backup to DAT. The maximum size of hard disk which can be used with the S1100 is an impressive 510Mb.
IN MANY RESPECTS the S1100 is the same machine as the S1000, so much about it will be familiar to S1000 users - not least the front panel, which has the same layout as the S1000. The S1100 can access up to 200 samples and 100 Programs in memory, or up to 400 samples + Programs + Keygroups. An S1100 Program can contain up to 99 Keygroups, each Keygroup being a collection of up to four layered samples assigned to any area of the keyboard, with the option of velocity-switching or velocity-crossfading between layered samples. If you set Keygroups to overlap on the keyboard you can also crossfade between them.
Additionally, the four samples within a Keygroup can be routed through a common filter (18dB/octave low-pass, still with no resonance), LFO and amplitude-envelope configuration, with a second envelope that can be set to modulate pitch and/or filter cutoff. However, some measure of independence is provided by amplitude and filter-cutoff offsets for each sample, while individual samples can be given their own stereo pan setting and output offset (relative to the output assignment of the Program).
The Mix page allows you to define for each Program an overall output level, an individual output assignment, the level at the stereo outs (this allows you to remove a Program from the stereo mix when you want it to go to an individual out only), a stereo pan position and the level at the effects send.
"The S1100 also introduces onboard digital effects processing to Akai's range, something which is only otherwise available on Ensoniq's new EPS16 Plus sampler."
The S1100 is 16-part multitimbral via MIDI, in that it can respond to all 16 MIDI channels. However, the way it goes about defining Program response is interesting. You can assign each Program to a single MIDI channel or to Omni, and give it a note range, an octave transposition amount, a polyphony amount and a response priority (low, normal, high or hold) that allows you to determine which Programs should have priority when polyphony is tight. If you give two Programs the same Program number (a Program's individual identity is established by its name rather than its number) and assign them both to the same MIDI channel, when the equivalent MIDI patch is received both Programs will be called up. Whether or not they are layered or split depends, of course, on their note range settings.
Also included on the MIDI front are a MIDI Receive Filter page (so that you can filter out certain types of MIDI data on each MIDI channel independently), a MIDI note Peak Program Meter page (a convenient way of checking the MIDI channels on which notes are being received), a MIDI Monitor page (so you can check incoming data) and a MIDI Exclusive page. The latter allows you to dump the S1100's Program and Sample data via MIDI SysEx, using MIDI handshaking protocol. Data can be transferred using either standard MIDI Sample Dump or an Akai-specific format. Alternatively, because MIDI transfer of sample data can take an extremely long time, Akai have given you the option to send your SysEx dump via SCSI instead - some generic sample editing software can read and write sample data via SCSI.
What of the sampling process itself? It's straightforward enough: you select stereo or mono sampling, the sample bandwidth (10kHz or 20kHz, referring to the S1100's two sampling rates: 22.05kHz and 44.1Hz), the original-pitch playback note, the sample time and the sample start method (input level, MIDI note or footswitch1) on the Reel page. Then you call up the Rec2 page, which presents you with an onscreen PPM which allows you to set the correct input level in conjunction with the front-panel Rec Level knob. Then you initiate sampling and watch the S1100 trace out the sample across the LCD window as it records. Once sampling has finished, you can trigger the sample from the Ent/Play button or a connected MIDI keyboard.
There are three sample edit pages: Trim, Loop and Join. All you have to do for Trim is set the sample start and end locations while playing the sample to check the right positions, then press the Cut softkey and the dirty deed is done. The Join page allows you to copy a part of a sample to another sample, splice a sample to another sample (all or part of in each case) and mix two samples (or parts of two samples) together. When splicing you can also set a crossfade amount if you need to smooth over the splice point.
The Loop page allows you to program up to eight loops per sample, and define how long each loop will play before the S1100 moves on to the next loop. You can use these multiple loops to do clever things like break up speech samples, repeating certain words or groups of words several times.
For a straightforward single loop you set the loop time to Hold, which means that the sample will continue to loop until you release the key. To find the right loop you have to edit two parameters, At and Length. "At" is the loop end point, Length defines the loop length. As you adjust these parameters, you see two vertical bars moving across the graphic sample display in the left half of the LCD, while the right half is updated to show the sample "curves" on either side of the loop point.
Personally I'd prefer to be able to set both the loop start and end points as absolute positions, as you can on, say, Roland's S770 - it's a simpler and quicker method than Akai's.
Each time you press the Find softkey, the S1100 goes in search of a zero crossing point, the received wisdom being that this will make a good loop point. Sometimes it will, sometimes it won't - but it's still a good feature to have. If all else fails, you can turn to Xfade, which crossfades the sample data on each side of the loop point, over a range defined by you. As this alters the actual sample data, it's a good idea to make a backup copy of the sample first. Finally, you can select either Loop In Release, Loop Until Release, No Looping or Play To Sample End (the latter being useful for one-shot drum and percussion sounds).
"The S1100 allows you to program up to 50 effects patches, drawing on four basic effect types: reverb, chorus/flange, stereo pitch shifter and echo/DDL."
TIME-STRETCHING, WHICH is also available on the S1000 and S950, allows you to lengthen or shorten a sample without affecting its pitch. You could, for instance, lengthen or shorten various drum and percussion sounds or special effects (in the latter case perhaps playing a sample in "slow motion" to complement a slow-motion section of film). But the main value of time-stretching lies in it allowing you to bring sampled performances into sync when the originals are running at different BPMs. I say "performances" because they could be anything: breakbeats, spoken words, a capella vocals, guitar riffs, horn riffs...
The S1100 provides a choice of two stretch modes: Cyclic and Intell. Cyclic is the more straightforward of the two, employing a fixed interpolation rate throughout the sample. For this mode you can set a cycle length, though quite what makes a good length isn't readily apparent. Still, you can get the S1100 to search out what it considers to be good lengths - rather as you can get it to search out good sample loop points for you. However, Cyclic mode is really only suitable for individual instrument samples, and even then it can come out a bit "warbly".
Intell mode is, as you might guess, more intelligent - it varies the interpolation rate according to the sample content. This is the one to go for in the majority of cases. There are two parameters specific to this mode: quality (the time that the S1100 spends working out cycle lengths) and width (the width of stretch crossfading).
The stretch range in both modes goes from a reasonable 25% to a quite extreme 2000% of the original length. Helpfully, as you change the percentage figure the S1100 updates the parameter fields, displaying both the time of the resulting stretched sample and the number of sample words it will take up.
In Intell mode, the stretch processing time rises quite drastically as you set higher quality and width values. For instance, a 200% stretch (doubling the time) on a 2.73-second 44.1kHz mono sample took just under 30 seconds with quality and width values of ten, but two minutes 40 seconds with values of 50 and five minutes 30 seconds with values of 99 (the maximum). It's best to begin with low values because these can often be quite adequate, especially if you're not doing extreme stretches.
The less the stretch percentage is, the quicker the processing. For instance, a 110% stretch on the same 2.73-second sample with quality and width values of ten took nine seconds. Going in the other direction seems to take even less time - a 75% stretch took six seconds. Given that most stretches will probably be within a ±25% range, time-stretching takes up negligible time and the results it produces don't affect the tonal character of the sound notably (something which can happen on much greater stretches).
THE S1100 ALLOWS you to program up to 50 effects patches, drawing on four basic effect types: reverb (large hall, medium hall, large room, small room, plate one and plate two), chorus/flange, stereo pitch shifter and echo/DDL. These patches form a common pool of effects which can be drawn on by the Programs - each Program number can be assigned an effect patch, which will be called up automatically when the Program is selected. Programs using the same number will automatically have the same effect patch, but if you want to use differently-numbered Programs together you'll need to assign them the same patch, because not more than one effect patch can be active at the same time. When any Program having a digital effect assigned to it is selected, individual outs seven and eight automatically assume the role of dedicated effect sends, and so lose their individual output function.
The FX page allows you to select a Program and assign an effect patch to it. However, if the Int parameter in this display is set to Off, the patch won't be activated. The other routing options are Int, Ext and I+E. If you select either of the latter two, the audio signal present at the sample inputs is routed through the selected effect and out via individual outs seven and eight (along with the internal Program in the case of I+E). You could use this feature to record an effected signal to tape without having to sample it first, and then sample the effected signal off tape (unlike Ensoniq's EPS16 Plus, the S1100 can't resample internally with effect). In a performance situation rather than a sampling session, you could plug, say, a microphone into the S1100 and use the sampler's effect processing to add reverb to vocals.
"Akai have responded to the requirements of the A/V and broadcast industries, which represent new markets for the company alongside the musical instrument market."
The S1100's effects are of a reasonably high quality, and quite versatile, but personally I prefer the character, power and flexibility of the effects on the EPS16 Plus. There again, while the 16 Plus has great distortion and overdrive effects, the S1100 has a versatile and very useful stereo pitch shifter - which you can use, for instance, to match up the pitches of different performance samples, recording one or more effected samples to tape and then resampling them back off tape.
PRESSING THE UTILITY button takes you into SMPTE Cue List territory, a new feature for an Akai sampler but a logical progression for a company whose samplers are often used in A/V contexts. Also, the combination of SMPTE Cue List and built-in SMPTE read/write facility makes the S1100 a suitable companion for the company's new DD1000 optical disk-based digital audio recorder.
New it may be, but Akai's implementation of what they call QPLAY mode has been well thought out and well implemented. One cue list can reside in the S1100 at any one time, and can contain up to 250 events. A SMPTE offset time can be entered at the head of the list, allowing the cue list to run from whatever timecode position you want. The S1100 can run as either master or slave with another SMPTE device.
Each event consists of a SMPTE time together with a Program, a MIDI Note On or Note Off, a note number and note velocity. Taken together, the event parameters tell the S1100 when to play, what sample(s) to play and with what velocity to play them. If the sample being played has looping set, in order to tell the S1100 when to stop playing it you need to specify a matching note off for the note on (which of course uses up an event). One event doesn't need to end before another one begins - remember, this is a polyphonic, multitimbral sampler you're dealing with. You can also take advantage of the S1100's individual outs and built-in effects.
If you want to compile a cue list for adding sound effects to picture, the best way can be to punch in the hits "live" while watching the relevant section of film. Fortunately the S1100 allows you to do this - you just press the Grab softkey, start the cue list running in sync with the timecode and press the F4 softkey at the relevant points. Once the event timings have been entered, you can always fine-tune them when you go into cue-list edit mode and assign the relevant samples to the events.
Other cue list functions allow you to insert, delete and slip backwards and forwards in time individual events or blocks of events, and sort events into chronological order. Cue lists can of course be saved to and loaded from disk, either as part of an Entire Volume (all data held in memory) or individually. You can store as many cue lists as you want on disk.
AKAI HAVE BEEN RESPONSIBLE for some pretty abysmal manuals in the past - and received plenty of criticism as a consequence. So it seems only fair that they should be praised for providing the S1100 with an extremely good manual. Written in a clear, friendly, methodical style, it doesn't assume too much knowledge on the part of the user but at the same time avoids being patronising. Also, a combination of thorough contents page and well thought-out index helps you to get straight to whatever you want to read up on. Oh, and all you environmentally-friendly musicians will be glad to know that Akai have printed their manual on environmentally-friendly recycled paper.
THERE'S NO DOUBTING the professional credentials of the S1100 - but then it can't really go far wrong considering it's based on a tried and tested industry standard. Akai have a good thing going with the S1000 - deservedly so - and they know it, but it's never a good idea to rest on your laurels in the hi-tech instrument business. Last year Roland finally fulfilled the promise of their earlier samplers when they brought out the stereo 16-bit S770, a serious piece of kit, as you'll know if you read last year's review. Could it be coincidence that Akai released preliminary details of the S1100 at around the same time, and brought in price reductions on their existing samplers? Surely not. But then the success of their samplers means that they can afford to be very competitive on pricing - indeed, they announced further price reductions at this year's Frankfurt Music Messe. And due in July at a current cost of £1899 is the S1100EX, a 16-voice-polyphonic expansion module for the S1100 which can be connected to the 1100 via SCSI and used in either one of two modes: voice expansion up to 32 voices, or multitimbral use in which a maximum of seven S1100EXs can be chained to an S1100 as SCSI devices.
What Akai have done with the S1100 is respond to the requirements of the A/V and broadcast industries, which represent new markets for the company alongside the well-established musical instrument market. Whether the many musicians who already have S1000s should trade them in for an S1100 depends on whether or not having the highest sound quality, the biggest memory and the possibilities offered by SMPTE cue lists rate as important considerations. And now there's another consideration to bear in mind: Roland are bringing out a cheaper version of the S770, called the S750, which at £2809 including VAT (17.5%) is set to be serious competition for the S1100 (see the Frankfurt Show Report elsewhere in this issue for more details on the S750). If you're in the market for a professional sampler then you need to check out both these machines.
Price £3299 including VAT at 15%. Prices are subject to change due to the recent VAT increase.
More from Akai (UK) Ltd, (Contact Details).
Review by Simon Trask
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!