Spektral Delay
German Native Instruments have been known for many years for their powerful and of good quality virtual instruments, as well as plug-ins or as stand alone versions (i.e. they don?t need software as a sequencer), especially their emulation of famous real instruments which had a great impact (Hammond organ with the B4, Prophet 5 with the Pro 52, DX 7 with the FM 7?). The quality of the filters and modulations created by NI could have been especially heard with Absynth and Kontakt. But until now, no effect software has emerged from the imagination and mastership of the engineers and programmers of the brand.
That?s what has been done with this first release of a virtual effects series named NI- Spektral Series: the Delay.
You could say : a delay unit again! Yes, but it is in fact a hybrid software, with such sonic possibilities that it can be uses like an instrument. And all this power is due to a particular and innovating functioning mode.
Explanations follow !
Classical effect units, whatever they are hardware or software, work on what are called data blocks. That is to say they cut the sound they receive in many little sound pieces they separately treat (applying the effect you chose). Then they ?stick? them to give, at the end, the affected sound. So they only work on time units.
That?s on this point that the NI-Spektral series is different. Indeed, once the sound you want to treat is cut, the software analyses, in real time, the spectral content of this sound (thanks to a proceed called Fast Fourier Transform (see the article on the ECVST web site here)), and cut it into many frequency bands on which the effect will be apply separately. The MI-Spektral Delay works on a frequential basis. In other words, you can apply an effect (here it is a delay) only on the bass or the treble, depending on the result you want. Basically, if you record a guitar and if you want to apply a delay on the bass only, you need Spektral Delay.
To have the same result with hardware, you would need as many pass-band filters as frequencies to extract? and as many delay units as frequencies to treat!
Here, you have all this in a single unit, and in real time. But that?s not all, because SD (as we will name Spektral Delay until now) offers other possibilities to modify the sound, in addition to the simple delay.
Spektral Delay is bundled in a box which contains the software installation CD (for Mac and PC, and for all the versions: VST Plug-in, DirectX Plug-in, and stand alone version), the manual and the documents concerning the registration (those who have the Internet will do it online, which takes less than five minutes).
Il faut bien l?admettre, le manuel, même s?il est bien traduit, a très peu d?intérêt. Il se limite à un catalogage des fonctions du logiciel. Aucun tutoriel de mise en route. Les explications pour l?installation sont réduites à leur strict minimum. Par exemple, on y parle d?un répertoire NI-SPEKTRAL DELAY DEMO, où sont supposés se trouver des programmes et des samples de « démarrage rapide » : ce répertoire n?existe pas ! De même, aucune explication n?est donnée quant à l?utilisation du soft en mode plug-in (insert ou send ?). Il semblerait que NI (Native Instruments, que nous nommerons ainsi à partir de maintenant) aient jugé que leur programme était réservé aux initiés en raison de sa complexité. Je ne sais pas? mais quand-même. Les manuels sont souvent rédigés par des ingénieurs et pas par des musiciens, il serait temps de penser un peu à eux. Un petit tutoriel en ligne aurait été sympathique pour illustrer les fonctions énumérées dans le manuel. En revanche, pour ceux qui sont habitués à ce type de soft, il y aura moins besoin du manuel.
Once this (little) bad surprise is gone, we introduce the CD in the CD-Rom driver, double click on the Setup.exe file, and the software will install without any problem. By choosing ?Custom Installation? you can, if you want it, install only the Plug-In version (VST or DX), or only the Stan Alone version, or both. Once the installation is complete, when you first launch the program, an activation key is asked. Here, no need to call or to go on the Internet to get the number, you just have to give the one which is on the box and in the manual: this is a good point for NI. You will be asked, from time to time, to insert the original CD and to recompose the activation code. You just have to get these elements near you. You cannot copy the CD (don?t ask me if the system is sure, I did not try). On the naked face of the CD are two little holes that forbid the copy.
For this test, I installed every version of the software, knowing that the sequencer is Cubase VST 5.1 under XP. We will see further that the stand alone version has its advantages, especially for those who create samples.
During the first uses, my computer crashed many times! In the plug-in mode, the fact of clicking on a program in SD simply closed Cubase! So I visited the NI site on the Internet. After having entered my personal information and my serial number, I could download the 1.5 update, with an re-written audio engine, much more steady. You will also get access to a special SD forum on the NI site.
To be noted: after the update, it is impossible to uninstall the software in a classical way. The installation logs are overwritten by the update. So you will have to uninstall it in the old way, cleaning all by yourself (system and registry? thanks for newbies!)
To configure the software (choice of the audio soundcard driver, MIDI options?), you will have to open the Stand Alone version. This part is relatively well documented, so no trouble.
Concerning the CPU consumption, this is very reasonable. In plug-in mode, about 8% on the CPU gauge in Cubase. In stand alone mode, Windows XP gives me 4% of the CPU used. More than correct, given the power of the stuff ;o) The computer used for this test is a Celeron II Tualatin over clocked to 1450 MHz, with 512 Mo RAM and two hard drives 7200 rpm (one for the system and one audio).
Here we are, now SD offers two options : plug-in or stand alone.
SD can function inside every sequencer that supports VST 2.0 or Direct X, as Cubase, Logic or Sonar. And it can also function alone (to simply treat sounds) or beside a sequencer (which will pilote it through MIDI, for example, to automate and record the changes you make). The choice will depend on what you will do the most. If you are interested in the modification of samples you re-use in a tracker (such as Acid), you will choose stand alone. If you use it to mix and process audio tracks in real time, you will choose the plug-in mode.
In the plug-in mode, as every other software of this type, you will just have to choose an audio track, then to select insert or send effects (SD functions in both modes). The software can also be used on VST-I tracks, bu, be careful! During this test, I realized that clicking on a program number in SD selected the same program in the open VST-is! Strange, isn?t it? You will have to play with the ?thru? options of you sequencers. Would this be a bug ?
A little clever originality: when you choose to use SD as a send effect, the Dry/Wet button of the software disappeared, to not be redundant with the ?send/feedback? button of the sequencer. As with their virtual instruments, NI chose polyvalence. The software will adapt to every context.
The software is so rich that we will have to go straight on! The better is, at the beginning, to play with every parameters to understand all the subtleties, then to learn to master them, which will happen faster than we could expect.
The interface is very professional and smart, and is pleasant. Once the few minutes of astonishment because of the functioning of the software and of the many parameters, all is easily accessible, understandable and clear, as always with NI.
To be noted : when you point on a button or a window, some explanations will appear in a dedicated corner of the interface. Very well thought, despite the fact that all is in English. Even if your servant perfectly speak English, one should have to translate this software for those who are not Shakespeare?s friends ;o))
The functioning principle (or the reading principle) of a sound graph is easy simple. It is the mathematic calculation to get it which is difficult ;o) On the horizontal axis, you will find the time, represented by vertical lines (in fact, the clocks of audio data, or frames). On the vertical axis, you will find the frequencies which are present in each frame. The more high-placed on the graph is a pixel, the more high is the frequency. Next, the more clear is a pixel, the more important is the amplitude (i.e. the volume) of the frequency. If you enter a nice pure wave at 50 Hz, you will simply see a clear line at the bottom of the graph, and conversely if it is at 20,000 Hz ? it will be at the top. The vertical cutting in frequency bands in the other windows are the same that in the sound graphs. So you exactly know which frequency is proceed. The left graph correspond to the in signal, and the right one to the sound graph after proceed (rebuilt). The whole on the left and right channel, of course!
In this window, you can modulate (i.e. transform) the original sound, before sending it in the other treatments. It is a insertion point in the signal routing. You have 11 modulation algorithms (13 in the 1.5 version), from inversion effects to signal deterioration ones, and striking rhythmic effects (a drum loops in random order in real time? wow!). Depending on the algorithm chosen in the rolling menu, buttons appear to set many parameters, such as speed, feedback, affected frequencies? you can imagine that all is allowed before applying the delay itself !
It is a big equalizer which will allow you to change the sound tone by increasing or diminishing the chosen frequency bands. The vertical axis shows the frequencies and the horizontal axis the attenuation degree in dB. Very simple.
For every frequency band, we can choose a different time delay. Vertical axis: still the frequency bands; horizontal axis: delay time in milliseconds. Nothing complicated.
As for a classical delay unit, this matrix allows to choose the amount of signal which goes back to the delay, that is to say in fact the number of repetition with decreasing amplitude. Once again, the vertical axis shows the frequency bands, and the horizontal axis the amount of feedback. So you can set this amount separately for each frequency band.
All is done graphically, thanks to several tools located in a menu above the edition matrixes: with the pen, you just have to click or to drag-and-drop to affect the chosen parameter. Combinations of keys allow to draw a line, or to copy forms.
You also can select areas you can modify together in one action, or to move the whole matrix along the two axis (as if you would equally modify each band). Of course, every parameters can be automated through MIDI. You just have to create a MIDI track for SD, and to record the changes (the same way you record the mixing automation in a sequencer).
For each matrix, you have a ?link? button (L) which obviously allows to link the left and right channels (changes in one matrix are at the same time done in the other), a ?bypass? button (to cut the effect), two buttons ?copy top? and ?copy bottom? (to copy the settings from one channel to the other one) and a ?monitor? which allows to know which frequency is affected. To conclude, we have in every window a screen which gives the delay time, the attenuation level (dB) and the feedback degree (in percentage).
All these elements can be synchronised with a chosen tempo or with the tempo imposed by the sequencer. We can even choose the rhythmic values of the synchronisation to the tempo: crotchet, quaver, semi quaver? many possibilities!
Well, no! In fact, we could even say it is just a small part of the software. What we have just seen is what is always on screen. But when you look in the software menus, you will see all the power of the Spektral Delay, especially its LFOs.
In the interface top, we can find five buttons which allow to get five different menus (which appear at the right of the buttons), and some of them are really powerful.
The ?file? menu simply allows to choose, as in the Windows (or MacOS) Explorer, the effects programs that were already saved and that we want to change. We find the usual commands (change program, save program...) and a banks navigator. Very simple and well-built. The software is bundled with a lot of factory programs created by famous people in the world of the C.A.M (among them is Craig Anderton, the boss of Music Player and Keyboard). Some are there especially to show the possibilities of the software, but their artistic interest is questionable.
The ?editor? menu allows to transform the curve drawn in the edition matrixes. After having selected the concerned frequencies bands, we can do absolutely all. It is quite complex and difficult to explain, but here it is: we have six buttons to choose one of the six edition matrixes (three matrixes on two channels). Once chosen, we have tools that allow us to invert a curve (horizontal or vertical mirror effect), to smooth it... and to give it particular shapes, thanks to generators (square, sinus, triangular waves...). Basically, that?s as if you had Photoshop! This will allow us to very precisely edit our graphs.
The ?Settings? menu allows settings and many options: number of frequencies bands (from 64 to 1024!), graphs scale, maximum time of the delay, MIDI options...
The ?About? menu, of course, gives information about the software. Nothing very interesting, but it is good to know which software version you have (especially for the updates).
And I kept the best for the end! The ?LFO? menu (Low Frequency Oscillator) allows to create effects that evolve in time! That is extremely simple but extremely powerful and inventive too. We have three LFO. For each one, we select the parameter which will be modulated (among twenty-four! For example, the modulation parameter number 3 of the left channel, or the X axis of the left channel, delay section...), then the wave form (sinus, square, triangle, saw, random) and last the amplitude! Once the LFO starts, we will see our parameter evolve in real time depending on the chosen elements (slowly and regularly for a sinusoid, abruptly for a square, randomly...). At this moment, we can see one of the powers of Native Instruments. It is possible to create pads from a drum loop !
Ton conclude, we must know that the quasi totality of the parameters can be automated through MIDI. Impressive, isn?t it ?
In the top left-hand corner can be found the master part, which control the totality of the software. Apart the programs loading/saving buttons, we can find a ?Dry/Wet? balance, a ?Gain? button, a general bypass (which allows to standby the effect), a mute (which turns off the effect) and a freeze (which blocks the signal).
To be noticed, the 1.5 update adds, just below the master window two bar-graphs which allow to monitor the in level and the processed one.
As I said it earlier, the stand-alone version has, at the bottom of the interface, a sample player; which can be used as any WinAmp (with a navigator and an explorer, play, stop and loop buttons...). It also allows to record or save in Wave he result of the original sound processed. You just have to click on ?record? and then ?stop? and the result is recorded in a previously chosen folder.
Apart the fact that this use is interesting for sound designers who create sounds or loops, it can also be interesting if your computer is a bit low to use SD in plug-in mode. You process your audio file in the software in the stand alone mode, you record the result, the you import that in your sequencer.
system would have been much more simple and precise. Here, you have to launch the recording, in real time (as in your sequencer), the launch the sample reading, then stop the recording. It is too problematical and you have to re-loop your samples. Not easy !
Apart some mistakes of youth (especially quite poor manual, and some little bugs), we are impressed by the possibilities of this software. Of course, those who never manipulate this kind of machine will be completely lost, but the graphic edition allows to understand rapidly, and... using his ears, the newbie will get it maybe better than the engineer.
The quality of the effects, the richness of the filters and the modulations, the extended editing possibilities, make Native Instruments coming near to the perfection. For a first attempt (it has to be reminded that it is their first effect unit) it is a master stroke. I bet that the future NI-Spektral software will be up to the Delay.
But it remains the question of the artistic interest of such a powerful software, able to transfigure (disfigure?... as you wish!) a sound. Filters and LFOs means electronic music? No. Spektral Delay knows how to adapt to every style of music. And if you like strange and weird sounds, you can enjoy during long hours. Even I as a guitarist rather turned towards big rock, I have been impressed. Maybe I will use it... sure I will.
In fact, It is a software with which we play, even if its quality and its (only apparent) complexity make of it a work tool which every professional will have to get.
Between the beginning of this test and the writing of this article, I received Cubase SX. So I did some tests again. No problem to be signalled, I even find the software sounds better, but it is probably psychosomatic.
Note : 19/20
Pros : smart and well-built interface, extended possibilities, not so difficult start, quality of the filters and modulations, the adaptability (not easy but present), a demo version is available. l
Cons : Cons: the manual, the uninstall problem, the software which is not in French, and that?s all!