Okay! I am finally certified to make this comparison! Or at least I can say something about two things that I did try.
A somewhat simple monophonic analog synthesizer with two oscillators, one filter and a simple arpeggiator these days can cost as much as a 3 oscillator, 25 voice, 16 part multitimbral, 9 filter, 7 effect virtual analog synthesizer or as much as the entire Korg Legacy Collection.
Building your own analog synth from modules seems to be a popular choice these days and I think this is a very nice direction and tendency in which anyone can contribute with his own module/distortion pedal/effect box etc.
Meanwhile, the legendary beasts of the past still soar at pocket-cringing, stomach-churning prices over at our good friend eBay while improved reincarnations are brought out by surviving companies.
So…what gives?
Any young producer starting out knows that software synths and effects have gone a long way since everything started. There are some good free ones and more than good cheap ones. The ones at higher prices are simply for the tweakers or for the preset users (sorry, folks!) simply because they are so advanced it takes you more time to be creative and shape one sound than it takes you to shape an entire album!
I, for one, have relied solely on freeware since 2005. My only regret was the fact of having to map everything to my MIDI controller and somehow pretend I was not controlling the computer. I wish I had something to sit at without the need of a computer and just tweak it and play it in my headphones on until I get really creative and decide to record. I still don’t have such an interface, but I’m pretty much used to that by now.
Other than the lack of a “hands-on” experience, I’ve been really happy and I have to thank all the people who decided to make their work free. Right now, the market is even bigger with such devs and there’s basically nothing you can’t do having only a computer and a decent audio interface.
Before I go into describing my experience, there are a few things I think it’s worth pointing out:
- People in this field of activity tend to use some words to describe sound that really don’t have anything to do with the actual characteristics of the sound; words like “fat”, “thin”, “weak”, “boomy”, “soft”, “aggressive” etc. Several characteristics of any sound are frequency, duration, amplitude, direction etc. Different values for these parameters can influence one’s perception on a specific sound. The same sound can be perceived differently by two or more people. Even though I criticize the use of the aforementioned words, I still may use them occasionally because I am not able to find a more accurate way of describing it since I cannot find an actual property of the sound to which I can refer to at a given time.
- I do not claim to have a lot of experience. I only hope my ears are set up correctly so that I may be able to distinguish what I am hearing. Normally, I tend to use just my ears but this time I did some measurements with a spectrum analyzer.
- This shouldn’t influence your decision on how your set up should look like. You may regret it if you change your mind after reading some nutjobs opinion on what’s best. Normally, the producers I know get the best results whenever I give them some very basic tips rather than telling them exactly how much eq, reverb send and so on. You should always judge with your own ears. This post serves as a simple reminder to reconsider some alternatives when you’re on a low budget and give them a try.
- If you think my view is subjective, you could be right or wrong. I can’t tell since these are my ears and my brain. My brain produces dopamine, endorphine and other pleasure-inducing chemicals when it perceives a cello, a piano, an Orbital or Prodigy or Juno Reactor track, an electric guitar solo and many others. But it’s not just any cello, piano etc. It’s a very redundant thing to say, really. Of course I have a subjective view! I just wanted to say that I’m a very skeptical person when it comes to people making things better through pure faith that it actually is better. So keep that in mind.
The analog synth I’m talking about is the Waldorf Pulse. Purists will probably say that it’s not a real analog synth, such as the legendary Oberheims, the Jupiters, Moogs and so on. Well, they all sound differently so you’re gonna have to decide on which of them is the real analog synth.
A first somewhat striking difference is that the simple unfiltered saw wave sounds differently than any other saw wave I’ve ever heard. Of course, it was still going through the lowpass filter at its maximum. This still has an impact on the raw oscillator waveform but sadly, you can’t make the wave go around the filter. Almost every digitally produced saw wave has that specific tractor-like buzzing saw with no significant amplitude at lower pitches. There’s also a difference in dB levels at different frequencies, which is probably influenced by the filter: most software synths have an almost equal spread of frequencies. It’s important to note that the Pulse’s oscillators don’t have a lot of high frequency content, reaching only 5000Hz which has a huge impact on the raw sound. The Juno60/106 probably has the type of oscillators that most soft synths are trying to emulate (no, I don’t really mean that. What I meant was the soft synth raw oscs rather sound more like the Junos than the Pulse.)
The problem is: there is no ideal distribution of dB on the frequency spectrum. If everything looks/is flat, you can bring out almost anything you want with a filter or eq. If some frequencies are favored over others, then the synth will be good for sounds in that range of frequencies since those will come out better in a mix.
One thing I noticed, and ultimately believe that this is how our ears work, is that the less the sound is spread over the spectrum, the more we perceive the region of frequencies that is highlighted.
To give you an example:
Image1 – Sound1
Image2 – Sound2
You would probably say that the second example has more bass content but if you look at the screenshot, it actually has the same amount of dB at around 50Hz. This example will become relevant in what I’m about to tell you onward.
Drifting is something common to all analog oscillators. I’ve witnessed it by using a precise tuner. It can be characterised as a slow “wobbling” or pitch change of less than a quarter semitone (plus-minus). It is perceivable, but only if you listen carefully for it and if you don’t program complex patches. I managed to emulate this behavior with decent results.
Filters play the most important role in any synth. That’s why there are so many synths. There are always the 3 or 4 basic waveforms oscillators can have which should more or less sound the same. Filters, on the other hand, don’t. The way the filter shapes the sound makes synths stand out from each other. For anyone who’s studied DSP deep enough, it’s purely exponential curves, so to speak: if you change one value by a tiny fracture, everything is transformed perceivably.
The particular Low Pass filter on the Pulse behaves quite differently from any soft filters I’ve heard: when you increase the resonance, the sound actually looses a lot of frequencies, making the synth go haywire at resonance values higher than 0.5 (to give you a basic idea, at 0.5 it sounds like any other filter at 0.7 or so) This tells me that this particular filter has a very high Q, which explains the somewhat major loss of any other frequency when twiddling with the resonance. Another interesting thing about this filter: it probably has a very long and steep curve – (in some cases) when you increase the frequency, the amplitude of lower frequencies also increases! So if you want to get the most bass possible, you actually have to max out the frequency on the filter (as opposed on how you have to do it in Synth1, for example – lower the filter to bring out the lower frequencies)
I’m approaching the final part of this comparison, which I hope someday will end and people will stop spending so much time on forums and such just to check if everyone has the same subjective view as they do. During the time I’ve spent with my Pulse, I realized why people tend to choose analog gear. It’s one particular VA that caught my attention and made me *snap* the answer out: The Clavia Nord Lead. Oooh, yes. You all enjoy your Leads, right? Almost as much as your analog gear. I wonder why? I mean, seriously: it’s around $1000, it has 2 oscillators, one filter, one proper lfo and one lfo/arpeggiator, no effects and very few modulation possibilities. Any synth freak would consider it the worst thing to spend your money on. Yet people still buy it and keep it in production. You want to know why?
It’s the sound that’s usable right after you plug it in. There’s no wasting time, no complicated way of getting a specific sound. You have a very narrow palette of good sounds. That’s exactly what any producer wants. He doesn’t really want to spend around 20 minutes with each complex synth he owns just to make one patch at a time for a project that involves 20 audio tracks. He just wants to fiddle around with something for a very short time before coming up with something musical. And that’s why the Nord Lead is successful. That’s why people love it and that’s why I love my Pulse. I just fiddle a little with it and BAM, bass. No compressing, no eqing, no tube processing, no distortion, no harmonic enhancer. Just hit record and play. I find usable stuff at every small tweak – and you can’t do too many, although it IS a complex synth when we’re talking about competitors.
So why did Hans Zimmer choose the Minimoog for a bass patch in The DaVinci Code? The same reason I would have chosen the Pulse: usable good sound with no effort or time wasted. Can I use my Korg Z1 or Blofeld to make a similar bass? Of course I can. But the Z1 sounds rather wooly and the Blofeld requires some careful tweaking. Then it would take a bit of touching finishes and enhancements before finally getting the desired sound. But anything that can get you in a hassle will eventually get you into one. So I can understand the love for a Jupiter or any other synth like it: it’s hands-on, with good sounds at every tweak of a knob.
I’m not saying software lacks that, but it’s certainly not hands on and it doesn’t inspire you to fiddle around. A Juno 106 has only a few sliders which makes it easy for you to fiddle with it and get an idea of what it can do. With soft synths, you either need experience with tangible interfaces to know what you’re doing or experiment, but it usually takes a lot more time.
Now, getting into a more realistic point of view: comparing raw saws is pointless. It really is. Nobody in their right minds uses raw saws (unless they’re making music with synths that don’t have parameters or something.) If you have something that’s complex enough or if your programming is good enough, you can fool the analog purists easily. Getting patches to sound the same is possible. However, the more parameters you tweak, the more the accuracy drops, simply because small differences come together (slight differences in filter curves, resonance, drifting etc.) I can’t accept the fact that you can’t remake a patch on something else and nail it by, say, 90-95%. A static patch is doable. Something that brings out high resonance and drifting, that’s a bit harder, but you’ll still be 90% there.
What I’m saying is that producers work with patches and programming, not raw oscillators while raving “yeah, feel the purity of the sine wave!” (I remember reading somewhere that Depeche Mode used a $20000 modular monster for a/some sine wave. Seriously folks, it’s just a sine wave.) Will Gregory of Goldfrapp used an MS-20 to pitch-track Alison’s voice (creative way of using it!) IDM producers use dblue Glitch for…well, glitchy beats.
Splitting hairs is the most wasteful way of spending your time, honestly. You’ll probably ask me why I eventually bought a few hardware synths when I was happy with all the free softsynths. Honestly? Not enough CPU. Getting complex layered sounds requires a lot of CPU power so that’s why I have the Blofeld (Largo clearly demonstrates that I’m better off with the Blofeld, heh). Physical modeling needs even more CPU so that’s why I have the Z1. The Pulse? Mainly for bass and monophonic leads and also to see what’s all this analog commotion about. You’ll probably tell me all about freezing tracks and other CPU saving methods. Whatever works for you, do it. I’m a bit older so I still find it appealing to have a tangible interface and watch the lights flickering in the dark.
Right now, the only thing is how much oversampling is needed before the most micro-fine ear fails to perceive any difference between an analog filter sweep and a digital filter sweep. I think it’s all a matter of time.
I am slightly in favor of software. First of all, you save some money. Secondly, you have less dusting to do. Third, your language will feature less swearing from all the cabling and recabling. Lastly, less time loading, more time recording.
Needless to say, I will still try to integrate all my gear into my work. They each play an important part in every project that I do.
The last thing I want to say is that good sound is what sounds good to you, whether it’s analog or digital/software. Take your time when choosing, because in most cases you will be able to save some money.
Over and out.
