@fcd72 Exactly the reasons I hint at in my original question above; the thing I’m looking for is simply a nice, somewhat scientifically sound, article summarizing these issues.
If you ask for such an Article in a DIY Forum…write it yourselves
I think the real big difference is when a track is made in the box with only soft synths in a sample accurate DAW or when it’s made with a lot of different devices mixed on a analog table.
But i also think there are a few plugin synths that do things you can’t get out of analog synths… Like Massive for example.
I would argue that it is the oscillators. It’s always the oscillators. I don’t care if the synthesizer is digital, analogue, or software. If the oscillators are not harmonically rich, the sound is already FUBAR.
Take for example all of the FREE VST software synthesizers. They ALL use the same oscillators - every single one of them. They may change around the filter, but no one ever changes the core oscillator algorithms. Or the FL Studio line of synthesizer plugins, they change the interface, but the oscillators still sound the same! A hardware example of this is the K3/K3m. It is a pretty cool synthesizer - especially the aftertouch routing, but every oscillator option (33 or so) sounds almost identical aside from major changes in amplitude. That is the main reason why I don’t own one. (If any one has single cycle waves from it they would be willing to share, that would be cool. :))
Basically, the filters are always over-hyped when it comes to this issue.
The other part is “warming up” the sound. Hardware is better for this because you can directly plug it into an old EQ amp without having to get it out of the box first. Just boost the lows, bring out some shimmer in the highs, and breath new life into your favorite VA synthesizer - pretty easy if you have the hardware. A great example of this is the new King Korg. The extra tube amp and distortion options actually bring out character in the different oscillators and filters that are not as easily audible otherwise.
However, for most accurate digital oscillators, the prize goes to the Shruthi - and probably Braids, but I haven’t gotten that yet. I use my JX8P as my Shruthi master keyboard. Why? The Shruthi “analogue” oscillators are based off of the Juno, and both my JX8P and Shruthi sound very similar - it also helps that the JX series is full of empty space to sit stuff on. The big difference is the amount of distortion I can run into the filter via the external input, but that is not a big deal if the volume is set properly. Of course, I am ignoring the aliasing issues, but it still sounds pretty damn good regardless - even though “weak” digital filters.
In summary, stagnate oscillators ruin a perfectly good synthesis engine.
I agree with audiohoarder ^
For example, a VCO connected to the external input of my SH-201 really changes the character of its sound. It sounds better.
I’d say a lot of it has to do with our innate inability to do fair comparisons - Or, to put it bluntly :
We hear what we want to hear.
If you know that the synth that you’re hearing is a VA synth pretending to be a minimoog, then you will subconsciously think less of it than if you were told that it was a minimoog.
This is of course the reason for double blind trials being so popular, at least if the result has to be absolutely correct, and not coloured by whoever did the study.
This phenomenon goes both ways, someone who likes the digital side of things will probably find the VA synth better, at least when told which one it is.
I have a great experiment you can do at home to really hit this phenomenon home! You will need a small MIDI keyboard - like a KORG micro control, a regular sized “professional” MIDI board, and a synthesizer that takes MIDI input. You don’t even have to hide what synth you are using, just concentrate on the keyboard you are using versus the synth that you are using.
Essentially the experiment should work like this - when using the less “professional” board, you won’t like the synthesizer’s sound as much. When using the “professional” keyboard you will prefer the synthesizer’s sound. Even if you have well-trained ears, this experiment still works. This is because it takes into account the overall experience.
As an addendum to my devil’s lawyer argument :
This works for audio applications as well, I tested it with my mates lovely setup, adn even when I told myself to dislike the more expensive cable : to my mind it still sounded like something was lost using the “cheap” cable(was still quite expensive)…
Compare this to the guy that made a bunch of self professed audiophiles sit in a room, and listen to some different cables, of which one was high end, one low end, and one a coat hanger… They couldn’t tell the difference, as long as they couldn’t see the cable being used…
Our mind is a weird thing
The difference between cables comes out as the length is increased. If you use a long, unshielded cable, you are more likely to pick up external interference like a radio station. However, both long cables will begin to remove the higher frequencies after 15 - 20 feet. Of course, the sentiment is the same and I agree that our minds are very weird. Cables just have a different set of rule to abide by.
These were high-grade “audiophile” cables, so double shielding etc were pretty much standard…
The feel of the more expensive cable was also much better (as I later found out), and overall, even knowing that the two cables should behave the same, I still somehow preferred the more expensive one… We’re talking something like 3-5m of excellent cables, sold at retarded prices
Yeah, double shielding still won’t help the drop in higher frequencies past a certain point. It would be around 6 to 8 meters. Also, coiling the cables will increase the amount of interference picked up despite the double shielding. I have to be careful because I live in a radio dense area.
Anyway, back on topic… the oscillators in VA synths are more prone to stagnation because most algorithms used to produce them are very very similar.
Talking bout cables - its time for the Cap Fairy to appear…
Regarding the OSC Algorithms - i bet most of them look like something like
for Ramp Up and
for Ramp Down and for Square the glorious
IF PhaseAccumulator < (FixedValueThatsHalfOfThePhaseAccsSize) THEN
OutPut = maximumlevel
OutPut = 0
If they would do it that way they wouldnt have these nasty spikes… and save zillions of cycles for not simulating something that produces these spikes
Lately my fixationwith vintage analogue has taken a nosedive, maybe its because My Nord g1 and Micro sounded so nice and WORKED that I was having fun tweaking and trying to play , same thing for all my MI gear , they have the feel and sound of the older stuff, but are MIDI and reliable ( well as reliable as anything I made can be) I think for my Analogue fix ill stick with the eurorack experimentation.
on a topical note, how would the Ambika compare to a Waldorf Microwave XT? are they even copmparable at all? I hope they are complimentary as i have one ( XT) on its way , I got it with my JX8P refund
The XT has its particular flavour of digital filters. So they are not really redundant.
This topic is widely talked about in every musiccreationforum I feel. While there always seem to be those who say software can sound as good as hardware, I say, until now, hardware (analog/digital) simply sounds smoother, fatter, bouncier. The harmonics are richer, delivering a distinct character.
A pretty good example of something I haven´t heard from any software yet is the sound of Trevor Pinch’s DIY Modular Synth
Btw these are apparently the schematics of that particular synth. Maybe you could make some use of it Olivier Schematics
Cool, I really like tangerine Dream , well until they went new agey , and listening to the demos of the Waldorf synths, they used them and the PPGs a lot more than I thought they did , learn something new everyday!