Akai S1100 hard disk recording

Akai's S1100 hard-disk recording option caused quite a buzz on its release, and even now makes this well-specified sampler even more versatile. CHRIS CARTER passes on some hard-won tips for making the best of S1100 HD recording.

Until quite recently, the question we were asked most often by visitors to our studio was "Where's the tape machine?" Like many musicians who record at home, we started with various Tascam and Fostex tape machines and have tried them all, from cassette multitrackers to reel-to-reel 2-, 4-, 8- and 16-track machines. As most hard-up musicians know, the running costs on a 16-track tape machine can be a real burden and you're tempted to start 'economising' by re-using tape and skimping on servicing -- both big mistakes.

About five years ago our 16-track Fostex tape recorder and Akai S950 sampler both desperately needed replacing, and we took the decision to sell them and change over to a MIDI-based hard disk recording system. After months of research, we concluded that the limited systems then available were far too expensive and we almost decided to stay with tape. Fortunately, we were offered a new Akai S1100 and S1000 at a price we couldn't refuse. The S1100 was launched in 1991 as a replacement for the highly respected but ageing 16-bit, 16-voice S1000. The principal differences between the two machines are an improved audio system on the S1100, with 24-bit processing, the inclusion, as standard, of a SCSI interface, a digital effects board, an AES/EBU digital output, a SMPTE interface, and a hard disk recording function (introduced with the version 2.0 software). We decided to take the plunge and began a fretful transition from tape to disk, courtesy of the S1100's HD option. After a couple of early catastrophes while recording vocals onto a Syquest removable disk (quickly replaced by a more reliable optical removable disk drive) we eventually developed an efficient, fast and very reliable system under the control of a MIDI sequencer and mixing down to DAT. While this method of recording may not be to everyone's taste, we found it a pretty painless transition. The S1000/S1100 series of samplers are well known for their ease of use and this, in my opinion, accounts for why they have become an industry standard. Even so, there are hidden depths to the S1100 and a lot of parameters to wade through, and many people just use the basic sampling, editing and playback functions -- so if you're considering HD recording, and have access to an S1100, the following tips could help you achieve some rewarding results, though they should also be applicable to users of later Akai samplers with HD recording facilities. For the purposes of this article I will assume that most readers have a basic working knowledge of the S1100 version 4.3 operating system.

RECORDING VOCALS

When it comes to vocals, a decent microphone will make more of a difference to the recorded sound than any amount of later editing or tweaking. A good microphone technique is also encouraged -- try to get as high a level as you can sung into the mic, and if your vocalist has a soft or quiet voice try to use an active or phantom-powered microphone, as these usually produce a higher output level. This will reduce the need to turn up the input gain too far, because while the input preamps on the S1100 are very good they aren't totally transparent, and increasing the input levels will eventually introduce unwanted noise. If possible, don't put the mic through a mixer first, as the mixer noise level, even if low, will also add more noise to the S1100 preamps. I use an old Sony ECM-56F electret condenser microphone that can be powered by a mixer or a battery but sounds particularly good when running on new batteries and plugged directly into the S1100's XLR input -- no hum and no noise. Although the S1100 jack inputs have more gain available than the balanced XLR inputs they are best avoided for use with microphones, as they pick up mic hum more easily and (because of impedance mismatching) the tonal characteristics are not as well suited to vocals. Other options for microphone recording could be to use a high-quality stand-alone preamp or vocal processor to add EQ, valve warmth, compression, limiting or gating.

It's worth spending some time setting up the input levels correctly on the DD RECORD page (EDIT SAMPLE>DD>DREC>TAKE) before you begin recording. This is because the 'linear' level display doesn't show the correct levels while the S1100 is in DD record mode. For some reason, it always shows the input levels about 30% lower than they really are, and if the input level control is set too high, serious digital clipping will occur. Make sure you know what your loudest peak is likely to be and mark the level control with a wax pencil as your absolute limit for that session.

Any S1100 HD recording works best when running in parallel with a MIDI sequencer. The S1100 DD record mode defaults to note C3/60 on MIDI channel 16 for its HD/DD record/playback trigger. The note should be placed a couple of seconds in front of where the S1100 will drop into record or playback. This forward delay is in addition to the default S1100 pre-delay of 500ms and allows the hard disk plenty of time to jump into action before the recording begins. The MIDI note, or the track it is on, should be locked or isolated so that any successive playbacks always remain in sync. If you intend to play back multiple takes, different MIDI note numbers can be assigned to each take on the DD PLAY page (EDIT SAMPLE>DD>SONG>S.ED), but remember that only one take (either mono or stereo) can play at once. An alternative way of playing back multiple takes simultaneously is outlined below.

TRANSFERRING DD TAKES TO RAM

If you are playing back a lot of different takes that need to be butted very close together or overlap each other, this becomes impossible using the HD/DD recording feature alone. Over the years I've developed a technique for transferring S1100 DD takes into the sampler RAM for better manipulation and control. To achieve this at the highest quality with no loss and no noise, an Akai IB104 digital audio interface must be fitted to your S1100. Using this interface, it's possible to save DD takes to DAT and then digitally load them back into the sampler RAM for editing and keygrouping. A lot of memory is essential for this procedure, with 8Mb RAM the minimum, 16Mb recommended and 32Mb an ideal configuration.

• The first thing to do (particularly if you only have 8Mb of RAM) is choose only the takes that are essential and save these to DAT, either one at a time or in bulk. This is accomplished using the ONE or ALL buttons on the TAKE BACKUP page (EDIT SAMPLE>DD>PLAY>BU.S). Connect a DAT machine to the IB104 interface, using the co-ax or optical connectors, and begin the backup (see the 'Stick to 44.1kHz' box). As the backup proceeds, mark your DAT tape with ID numbers to help guide you through it when you play the tape back later.

• Next go to the DIGITAL INTERFACE page (EDIT SAMPLE>REC1>DIGI) and set the source to DIGITAL and the input to either ELECTRICAL or OPTICAL, depending on the type of connections your DAT machine has; leave the receive rate at AUTO.

• Now go to the sample RECORD page (EDIT SAMPLE>REC2) and set the parameters as you would normally for: NEW NAME, STEREO or MONO, SAMPLE LENGTH and PITCH, and so on. The input level doesn't need to be adjusted, as you're recording from a digital source. The S1100 display should indicate that it's receiving at 44.1kHz or 48kHz, depending on the output of the DAT machine.

• Play back the DAT tape and record the previous DD takes into the sampler RAM a section at a time. At the beginning of each take you will hear a very short burst of digital data -- this is irrelevant information that can be edited out on the sample TRIM page (EDIT SAMPLE>ED1). The takes can also be edited into shorter blocks, deleting any silences or pauses in the process to save RAM space.

Once the vocal takes have become samples, a wealth of editing facilities is available to you -- retuning, stretching, squeezing, reversing, combining, looping, and so on -- and, if you have access to a Mac or PC sample editor, other functions, such as EQ and level re-scaling and special effects, could be used.

RAM BACK TO DISK

When you've assembled your re-sampled takes, they can be put into keygroups, and if you have sufficient RAM it's possible to have 20 or 30 vocal phrases spread across a keyboard. Using a MIDI sequencer you can now arrange and rearrange the vocal takes with ease on a keyboard. Vocal takes can be cut and pasted on screen if you're using a computer-based sequencer, timing can be adjusted and shifted, and takes can be overlapped and double-tracked.

If the S1100 is a major part of your setup, it could be pretty inconvenient having the sampler RAM full of vocals. One of the most useful features of the S1100 is the ability to play back samples and stereo HD takes simultaneously. An often-overlooked part of this duality is the ability to mix down the contents of the RAM -- samples, keygroups, and so on -- onto the hard disk. To achieve this:

• Prepare the MIDI sequencer to play back the rearranged and edited vocal takes, and include a separate note (C3/60) on MIDI channel 16 to trigger the S1100 DD record/playback function, as mentioned earlier.

• Put the S1100 into DD record mode, play the sequencer and record your samples onto the hard disk.

The sampler RAM is now free for loading in instruments, percussion, or whatever. With this arrangement it is possible to have a stereo mixdown of vocals (or instruments) playing back from the hard disk on outputs 7 and 8, with individual outputs 1-6 and the stereo L/R outputs available for more instrument samples, without any loss of polyphony.

AND FINALLY

Many of my comments on vocals can also be applied to other recordable sources, acoustic and electronic. If carefully recorded there is little difference between a DAT recording and an S1100 HD/DD recording, and with its real-time stereo digital output the S1100 is ideally suited to digitally recording your work into any S/PDIF-equipped DAT machine, digital mixer or digital soundcard (PCI, NuBus, etc).

Second-hand S1100s seem to hold their prices quite well and are not seen in classified ads as often as the S1000, as satisfied users appear to hang on to them for longer. Currently you can get hold of an S1100 for around £1000 to £1300 depending on its condition, the amount of RAM and whether it has a IB104 card. Points to look out for when buying an S1100 second-hand are what operating system it has (currently v4.3 and preferably in ROM for much faster booting), whether the display is nice and bright, and whether it comes with a copy of the revised instruction manual, which is pretty important if you want to get the best from the machine. [If you own, or find, an S1100 which doesn't have the software with the HD option, an approved Akai service centre can supply a new EPROM with the latest software. Panic Music (01954 231348) charge £45 for this service. Thanks to Adam at Panic for his help.]

When you consider that the Akai S1100 is a superbly specified industry-standard sampler with probably the largest sample base in the world and a stereo hard disk recorder, you begin to appreciate what a bargain this machine really is.

WELCOME TO THE RSDO

The wonderfully named Real-time Stereo Digital Output is more than it appears to be. Although it uses the professional XLR balanced AES/EBU configuration, don't let this put you off: it is compatible with the more popular consumer S/PDIF standard. I have two leads for use with the RSDO: a long one with a phono plug for connecting to a DAT machine, and a very short one with a quarter-inch jack plug which connects to the IB104 digital co-ax socket, for performing digital mixdowns as mentioned elsewhere in this article. It's very easy to make your own S/PDIF digital lead if you just follow these directions. Looking at the rear of both the XLR and phono or jack plug, connect the pins as follows:

• XLR pin 1 to phono/jack centre tag or tip.

• XLR pin 3 to phono/jack case or shield tag.

Take a look at the accompanying diagram for a more graphic idea of how to make the lead.

TAKE IT FOR A SPIN

To get the most from the S1100 a hard disk is essential, especially if you want to try HD/DD recording. For some reason Akai have a formatting limit of 500Mb, no matter what size the hard disk, so Jaz drives are out! But even a relatively small hard disk can store hundreds of samples, keygroups and programs, and loading and saving times are reduced to seconds. Most types of SCSI drives can be used, but beware of old, slow models: these can cause a stuttering or chopping effect when stereo takes are played back.

An option endorsed by Akai is the use of 3.5-inch optical media. The most popular and cost effective is the 230Mb format, as the original 128Mb format is being phased out and the newer 640Mb type are still expensive. Optical disks are very cost-effective (about £70 for a pack of 10), reliable, and immune to stray magnetic fields. The best buy at the moment must be the dinky Olympus MO230 drive (without a fan, so ideal for audio work, and available from Mac & More on 01442 870300) at £351 including VAT. The price also includes 10 230Mb disks. An even cheaper, if slightly smaller, option is an Iomega Zip drive, although the disks work out more expensive than optical in the long run.

Formatting a hard disk for use with the S1100 can be very slow -- in excess of 20 minutes for a 230Mb optical, though it all depends on how the disk is partitioned. When formatting for HD/DD recording, the disk can be partioned entirely for DD takes, in which case a 230Mb optical would give about 22 minutes of stereo recording. Alternatively the disk can be divided into two or more partitions, with one partition for DD takes and one or more for samples, keygroups and programs. A disk partitioned for DD takes and samples takes the longest to format. But here's another exclusive tip. If you are about to start a session recording to HD and you suddenly realise you haven't got a formatted disk and can't wait 20 minutes, try this.

Go to the S1100 FORMAT page (DISK>FORM) and set the parameters as follows:

Partition size: 1Mb, max: 1.

Now press ARR, followed by YES. Twenty seconds later you'll have a fully formatted disk, ready to roll. The only downside is that the disk can only be used for HD/DD recording, but it's probably best to keep separate disks for samples and HD/DD recording anyway.

STICK TO 44.1kHZ

The sampling frequency of the S1100's real-time stereo digital output and analogue HD/DD recording are fixed at 44.1kHz. Unless your DAT machine will only record at 48kHz you should always transfer or back up at 44.1kHz, because although the S1100 can convert from 44.1kHz to 48kHz when making backups, the pitch of the take will be affected when you load them back, so your vocals will be out of tune. Sticking with 44.1kHz also makes things simpler when performing a digital mixdown.

S1100 BRIEF SPEC

• 16-bit linear sampling at 32kHz, 44.1kHz, 48kHz.

• 16-voice polyphony.

• 24-bit internal processing.

• Stereo hard disk recording.

• Real-time DSP effects.

• SCSI interface.

• SMPTE read/write.

• 32Mb RAM, maximum.

• 200 samples, 100 programs.

• Stereo output.

• 8 assignable outputs.

• Digital output.

• 320-character display.

• Options included the IB-104 S/PDIF optical/co-ax digital interface and various sizes of internal hard disk.

DIGITAL MIXDOWN EXCLUSIVE

Here's a never-before-revealed, exclusive Chris Carter tip for users of S1100s with the IB104 interface fitted.

The S1100 manual tells you that to perform a mixdown from RAM to disk you must connect the two rear analogue L/R outputs to the front L/R inputs. A better way, which keeps everything in the digital domain, is to connect the real-time stereo digital output to the IB104 S/PDIF co-ax input and configure the S1100 as follows (for XLR pin connections see the 'Welcome to the RSDO' box).

• Go to the DIGITAL INTERFACE page (EDIT SAMPLE>REC1>DIGI) and set source to DIGITAL and input to ELECTRICAL.

• Next go to the DD RECORD page (EDIT SAMPLE>DD>DREC) and set source to DIGITAL, d.rate to AUTO, start to M.NOTE+DEL, stereo mix to OFF and mode to either STEREO or MONO.

• Now go to the TAKE page (EDIT SAMPLE>DD>DREC>TAKE), name your mixdown and set the length to a suitable figure.

• At this stage you may find the mixdown level is too soft or loud, and although it seems as if there is no way to adjust the input level, because you're in the digital domain, there is a workaround. Go to the DIGITAL STEREO OUTPUT page (MASTER TUNE>Dout), which sets the level for the real-time stereo digital output socket. Here you have a choice of 0dB (default), -6 dB or +12dB. Make your adjustments and return to the TAKE page.

• Set your MIDI sequencer as described earlier and record a digital mixdown. One of the benefits of this digital mixdown method is that, in theory, an endless number of DD mixdown sessions can be performed without any loss of quality.

TRIGGERED SYNC

Unusually, the S1100 doesn't use a timecode to stay in sync while in DD mode, unless you use the internal (or an external) SMPTE generator to play takes from the QLIST. Instead, the S1100 relies on the timing accuracy of the connected hard disk and MIDI sequencer. I've used various sequencers, both software and hardware, from different manufacturers, and on the whole they stay in sync to the same degree as do hard disks. While writing this I tried an experiment with a drum pattern recorded from a MIDI sequencer onto the S1100 HD with the DD record function triggered by a MIDI note at the beginning of the pattern. I played back the sequencer and the S1100, triggering the DD take from the same note, and apart from some slight phasiness the rhythm patterns were still in sync after eight minutes. Pretty impressive!