Even if I am fond of “real” acoustic instruments,I have always been intrigued by “synthetic” wind instruments,be them sampled or synthesized. For example,I really like some flute patches that come with the Patchman Turbo VL upgrade chip for the Yamaha VL70-m,and I find them quite expressive and playable with my Akai EWI 4000s. Alas,I always failed at finding a decent implementation of a virtual sax.
Using a sampler rather than a synth – including synths that use physical modeling to replicate the sound characteristics of a real wind instrument,as is the case of the VL70-m – surely delivers a result closer to a natural sax sound,yet I have always been disappointed by the lack of playability and the inability of achieving the many timbre variations that only a real sax can offer.
When I decided to record music for sax quartet,the choice was between purchasing a real tenor and baritone (to be added to my alto and soprano) or trying again with a sampled instrument. This is how I got to Samplemodeling and its range of products,which includes brass (Trumpet,Trombone,and French Horn &Tuba) and woodwind instruments (Saxophones,Clarinets,Flutes,and Double Reeds).
Their list price ranges from 159 euros (about 200$) for products with two instruments (e.g. Trumpet,Trombone and Clarinets) up to 259 euros (330$) for packages with four instruments,such as Saxophones (which includes soprano,alto,tenor,and baritone) and Double Reeds (oboe,English horn,bassoon,and contrabassoon). These price tags make Samplemodeling products more expensive than libraries that offer hundreds instruments for a fraction of a dollar each,but this comparison is misleading,because none of the instruments in these libraries can really compete with Samplemodeling.
The company has been on the market for a while and gained a very high reputation among musicians,especially EWI players. The first line of their products is based on the Native Instruments’ Kontakt platform,and in fact they include the free Kontakt Player for users who don’t own the complete product. This holds true for brass instruments,but newer woodwind instruments – including The Saxophones – are based on a proprietary engine named SWAM (Synchronous Wavelength Acoustic Modeling). For more tech information about SWAM check this page,but for now it’s enough to say that this platform is a further step towards more realistic virtual woodwind instruments. SWAM instruments don’t require you to install the Kontakt player.
Installation,licensing and required hardware
The good news is Samplemodeling software needs no dongle key (e.g. the iLok dongle),which means you don’t have to give up to a USB port on your computer AND you don’t feel you are treated as a potential pirate. Instead,you get two license keys,which allow you to install their software on two different computers. If you buy a third computer you can have one key deactivated and reactivated on the new hardware. As is common with most vendors,a valid license gives you the ability to download future minor releases of the same product,without having to pay for a yearly subscription. My review is based on version 2.4,but I learned that version 2.5 is about to be made available (for free) to all registered users.
Samplemodeling instruments work on Windows XP and later and on Mac (OSX 10.6 and later). They are provided only as plug-ins in the VST2,AAX,and AudioUnit formats,therefore they require a plug-in host such as Cubase,Live,Logic,or ProTools. If you don’t have a host program,you can use a free one,such as Cantabile Lite. In all my tests I used the AudioUnit version loaded inside Apple MainStage 3.
Interestingly,Samplemodeling virtual instruments are built with the wind player in mind:unlike other software,you must use either a wind controller (such as the Akai EWI or Yamaha WX5),a MIDI keyboard with either a breath controller (such as the Yamaha BC3) or – at the very least – an expression pedal that emits CC11 messages. Without any of these MIDI controllers,SWAM instruments don’t sound at all!
Installing The Saxophones on my MacBook Pro was a breeze. I haven’t used by second serial key yet,therefore I don’t know how it works on a Windows system,but I don’t expect any trouble. I scanned the Samplemodeling forum before getting the software,and the satisfaction level among users is impressively high. This surely means something.
Quite opportunely,Samplemodeling provides three configurations to start with:for keyboard players with an expression pedal,for keyboard players with breath controller,and for wind controller players. Each configuration correctly initializes the program’s parameters to work properly with the corresponding MIDI controller,and in fact all the four instruments of The Saxophones worked “well enough” as soon as I loaded them in MainStage.
But probably “well enough” isn’t enough for you,thus you need to read the manual to get the best out of these instruments;failing to do so means that you are leveraging a fraction of their potential. The documentation is short,clear and concise,and just takes a few minutes. Once I grasped the fundamental concepts,it took less than one hour to tweak the many available parameters until I got the sound and the responsiveness I expected. The parameters are grouped in three distinct sets:
- basic sound parameters,in the main window
- advanced sound parameters,in the Options window
- MIDI CC mappings,also in the Options window
Let’s have a look at basic parameters first,even though they aren’t really “basic”,as you’ll see in a moment. In the top half of the window you can set the intonation (in Hertz or cents),the transposition (in semitones),the pitchbend up and down ranges,the volume,and the pan. You can also control reverb,even though I prefer MainStage’s own reverb plug-ins because they provide more options.
A rectangular area in the center of the main window shows a subset of the messages arriving from the MIDI controller,i.e. pitch bend,velocity,expression (CC11) or breach (CC2),and vibrato. It also provides a graph of the dynamic envelope and shows how the software is interpreting the way you articulate the phase (e.g. legato or staccato). These pieces of information are crucial,because The Saxophones – unlike traditional samplers – uses sophisticated algorithms to understand the musician’s intentions and generates slightly different sounds depending on the result of this analysis. If what you hear isn’t what you had in mind,you should keep an eye on these parameters to understand how the software is analyzing your playing style.
The quest for the perfect parameter value
In the bottom half of the main window you see more advanced controls that directly affect the resulting sound. Your first option is which instrument/microphone combination you want to use:each sax model (soprano,alto,tenor,and bari) provides between 8 and 10 variations,labeled with names such as “Sax 1 Dry” or “Sax 2 Bright”. Each variations results in a slightly different timbre,even though the differences aren’t always noticeable,at least to my ears. After some tests,I found myself focusing on two or three different variations for each sax model.
Except for a few sliders whose meaning is obvious – for example,the amount of growl,the compressor rate,and the release time – the controls in this area require a deep knowledge of acoustic principles and honestly I can’t really claim I could grasp the meaning of each of them. In some cases I managed to have interesting sound variations by moving a slider,in others I found no audible difference,probably because some controls must be adjusted in groups rather than individually.
For example,I found out that lowering the default value of the Dyn.Pitch parameter – which controls random pitch fluctuations in response to different breath intensity – allows me to articulate phases in a way that is closer to my playing style on the real instrument. Likewise,increasing the SubHarm value – which simulates the sub-harmonics that acoustic instruments create one octave below the played note – results in a “richer” sound.
Some parameters are meant to be controlled dynamically by sending CC messages from your MIDI controls. For example,the Growl and Flutter Tongue values should be set to zero because you don’t want these effects in all your phrases:by mapping them to specific CC messages you can enable them on-the-fly and add expressiveness to your phrases.
A couple of sliders affect the noise produced by breath and keys;when playing at low volume,these two parameters add realism to the sound. You can control them using CC messages,too,but in practice their default value is OK for most circumstances.
As if all these options weren’t enough,the Options button (near the bottom-left corner) brings up a window with additional advanced parameters (see right portion of following figure). Here you can set the way portamento and vibrato work,the expression curve,under which circumstances the virtual sax “overblows”,etc. I haven’t played much with these values,because I was already satisfied with their default value,but it’s good to know that you have so many options to tailor the virtual instrument to your taste and style.
The SWAM engine also supports microtones and non-tempered tuning. In the Options window you can select the detuning value (in cents) for each of the twelve semitones,but this operation alone doesn’t change the intonation. Instead,you have to activate the microtuning feature either in the main window or by sending a specific CC message. In either case,you can decide which notes use the alternate tuning. Notice that you can detune any note by the desired amount of cents,but you still have “only” 12 notes available:this means,for example,that you can play non-tempered Indian ragas and other Eastern scales but you still can’t perform any sort of quarter-tone compositions,which would require up to 24 different keys.
Optimal MIDI and EWI settings
The fields in the left portion of the Options window allow you to customize how MIDI CC messages map to most of the parameters just described. The default mappings are reasonable,yet they assume that you have a large number of CC controllers available,such as a MIDI keyboard with a lot of sliders and knobs or – if you are an EWI player – a foot MIDI controller such as Behringer FBC1010 or McMillen SoftStep. Even if I have both these foot controllers,my goal was to achieve good results with the EWI alone,thus I spent some time to configure the software and the EWI for the highest expressiveness. Here’s the configuration I came up with.
In the SWAM Engine:
Trasposition: +3 semitones for alto and baritone,-2 semitones for soprano and tenor – this means that a given fingering delivers the same note that would produce on the real instrument,so for example the low Bb generates more or less the same characteristic tone you get from the saxophone.
PitchBend Up: 0.0 semitones – this is necessary so that you can use the EWI pitchbend plate to send a specific CC message without affecting the pitch.
PitchBend Down: 1.0 semitones – with an acoustic sax you typically don’t “bend” a note for more than a semitone;besides,using a small value allows you to achieve a “pitch vibrato” by rapidly moving the thumb on the EWI pitchbend plate.
Expression: mapped to CC2 (breath controller),the standard setting if you select the wind instrument configuration.
Vibrato Rate: mapped to CC1 (default),to allow you to apply vibrato using the modulation wheel (on a keyboard) or the pitchbend up plate on the EWI (by using the configuration described below).
Portamento Time: mapped to CC5 (default)
Growl: mapped to CC4,so that you can add growl effect by pressing an auxiliary key on the EWI (see below).
Overblow: mapped to CC64,so that you can achieve this effect by pressing a key on the EWI (see below).
On the EWI:
You should configure the EWI to send note velocity and breath information (CC2) with each note,and not send volume (CC7) messages. Sending velocity doesn’t really change the note volume or timbre (which depends on CC2),but can affect other behaviors,for example the transition time from note to note during glissandos.
To match the SWAM settings described above,you should map pitchbend up values to the CC1 (modulation) message,which lets you easily have a variable degree of vibrato by simply sliding your right thumb up the pitchbend plate. The Glide plate should map to CC5,so that it controls the portamento in legato phrases.
The Hold key – i.e. the auxiliary key closer to the mouthpiece – should be mapped to CC64 and send the value 65 (or any value greater than 64) when pressed the first time,so that you can activate overblow mode by just pressing this EWI key. Pressing it again disables overblow.
The Octave auxiliary key should be mapped to CC4 so that you can achieve the growl effect by pressing a single key;in the setup you should specify a value between 30 and 127,depending on the amount of growl you desire. Unfortunately,the EWI can only send a single value,thus you can’t apply a varying degree of growl unless you use a MIDI foot controller. Pressing the key again returns to normal (non-growl) sound. Alternatively,you can give up to either overblow or growl,and instead use one of the EWI auxiliary keys to activate microtuning.
Finally,I strongly recommend that you reduce the delay that the EWI introduces between playing a fingered note;the default value for this parameter is 7 and you probably never edited it,but Samplemodeling tech support recommends adopting a lower value,say 3 or 4,to reduce the number of “ghost” MIDI notes that the instrument sometimes emits and that is sometimes responsible for spurious notes,clicks,etc.
It’s time to answer the questions that always come up when speaking about virtual instruments:do these instruments sound as good as the “real” ones? Can “The Saxophones” replace real,acoustic saxes?
Unfortunately,the only answer I can provide is:it depends. More precisely,it depends on why you want to use a virtual instrument in the first place and what are your expectations.
Let’s start saying that the quality of sound is fully satisfactory:if you play single notes or slow musical phrases,the saxes in “The Saxophones” can be hardly distinguished from acoustic instruments. Same consideration applies if you want to record a big band section:to get the idea,listen to the demos in this page. In the hands of a good EWI player,a virtual sax can be also used to play ballads and medium-tempo songs,as this video demonstrates.
Samplemodeling instruments were accurately recorded in an anechoic room with expensive mics and using first-class instruments played by professional musicians,thus a recording session based on these sampled instruments can often deliver better results that those you typically obtain in your home studio. If you don’t have four saxophonists at hand,“The Saxophones” is your next better option.
The sound of all the instruments provided in the package is equally good,yet I have my preferences. In my opinion,the baritone and the soprano delivers slightly better results,then comes the tenor and finally the alto. But again,this is my personal opinion and it probably depends on me being an alto sax player,which means that have higher expectations for that specific model.
Playing these virtual instruments in live gigs can be challenging. The detail that makes the biggest difference between a virtual or real sax isn’t the sound itself,rather it’s the playability of the virtual instrument:a sampled instrument just cannot render the countless timbre and articulation variations that you can achieve with an acoustic instrument,such as alternate fingerings and different tonguing techniques;you can sing in the instrument,use multiphonics,produce vibrato by using either the breath or the jaw,and so forth. You can’t reasonably expect that a virtual instrument can implement all these techniques,even though Samplemodeling offer some interesting possibilities,for example in the way they implement growl and flutter tonguing.
Just remember that every single detail of how these instruments sound can be controlled via CC messages:if you really want to get the best results you should read the manual,learn how to tweak the most important parameters,and maybe use an EWI with a MIDI foot controller (or a master MIDI keyboard with a breath controller and a lot of sliders and knobs).