Analogue vs. Software - blindtest

[fluffy_little_something]

Interesting. So basically one might as well make a synth that doesn't have any LFO's, but instead, say, 4 oscillators, which one can freely use as mod sources as well since they are "superior" anyway.

In theory one could modulate the LFO frequency with keytracking I suppose.

Yes, I can imagine that when using several modulations at the same time, the amount of calculations increases a lot. Maybe it's not even clear which modulation to calculate first

Some of those sounds in your audio are not bad at all, especially those with few higher frequencies.

[aciddose]

Mostly the aliasing only kicks in at the highest notes above 3k or so, if you stick around 1k you usually won't need over-sampling.

That's because the product is going to be something like 4x (at most) of the source component frequencies.

Non-linearity or distortion usually generates an unlimited number of harmonics, just as they get higher the amplitude drops off.

When you're modulating things however the result is usually the product of the modulator + carrier components.

So say you have 1k for both. If the operation involves a polynomial function of some sort, the number of harmonics generated is related to the degree of the function. Or in other words the number of multiplications (or powers).

https://en.wikipedia.org/wiki/Degree_of_a_polynomial

The starting point is to look at AM or ringmod, where there is one multiplication.

Carrier * modulator.

The highest frequency this can produce is the second harmonic of the highest component. You might have heard of "sum and difference", well the highest sum you can get is:

Carrier + modulator.

So if both contain a 1k sine wave, the result is 2k at most. If you're using a high sample rate this can't produce aliasing!

Most sorts of modulation are maybe second (x^2) or even third, but generally not higher. This means the number of harmonics generated is limited.

So you can get away with reasonable amounts of oversampling like 4x quite easily.

This is why using audio-rate modulation to produce "a little dirt" is often a lot better choice than using distortion. You can often control the number of harmonics generated directly.

Actually the same applies to distortion but since this depends upon amplitude and the functions are often high degree (hard clip = infinite) it is more difficult. Try placing a limiter before a distortion though for example. It allows you to control things a lot more.

[Ingonator]

In theory one could modulate the LFO frequency with keytracking I suppose.

Key tracking of LFOs seems to be possible in many synths AFAIK, especially those with a mod matrix and something like "key", "note" or "key tracking" as a mod source and "LFO Speed" as a mod destination.
LFO key trackings seems to be quite useful when doing PWM sounds.

- FWIW the upcoming Tone2 Icarus wavetable synth has a key tracking knob (bipolar) in the LFOs and seems to be the first Tone2 synth that has such knob in the LFOs (while in that and some other Tone2 synths it is possible using the mod matrix for this).
- Waldorf Largo has a keytracking parameter in the LFOs too (up to +/- 200%), same in Blofeld. In the Pulse 2 this is available from teh mod matrix ("key" routed to LFO speed).
- In U-He Diva you could use "Key tracking" as a source for the "Rate Mod" and/or Depth Mod" parameter in the LFOs.
- Arturia Matrix-12V includes tons of modulations including assigning "keyboard" to the LFO Speed and/or amount.
- In Xils Lab Syn'X 2 "Keyb" could be assigned to 3 LFO parameters using the mod matrix.

In the Minimoog which was able to use Osc 3 as a LFO with a switch you were able to switch of key tracking for Osc 3 which was useful when using Osc 3 as a LFO or doing sounds that use an oscillator with constant pitch (which is ueseful in some cases and i used this for a few patches i made with Waldorf Pulse 2 but also with some other synths).

Just checked Sylenth, they go up to 144Hz.
I remember TAL's free Elektro, its LFO's go much higher. Just checked their site, 400Hz.

To me it seems that audio-rate modulation sounds a bit unmusical for the most part, especially with high amounts. But it varies from note to note

In Tone2 Saurus 2 the LFOs could go up to 440Hz and it has a dedicated Filter FM knob (while it does not have a dedicated Osc FM). In other Tone2 synths like e.g. Nemesis and the upcoming Icarus this goes up to 440 Hz too.

UVI Falcon so far does not seem to offer crossmodulation except inside the dedicated 4 OP FM module. Anyway it seems to be possible to do audio rate modulations using teh multi stage enveloeps as those could go up to 1000 Hz. You could use teh MSEG as a waveform and tehre are even some presets that emulate certain waveform shapes.

[fluffy_little_something]

A reference note button for tuning might also make sense. Was just playing around and noticed the pitch to keyboard relationship was more or less gone.

440Hz on Tone2 synths, probably no coincidence, a famous audio frequency.

[zyguli]

I guessed (1) ,(2) and (4) right and I preffered Hw. In (3) I was wrong with guessing, but I preffered sound #2 which turned out to be coming from sw, but the timbres were too distinct so the challenge was slightly off. (5) is interesting case, because I preferred #2, which I thought was hardware, and even after knowing it was coming from sw, I still liked it more.

One note though - while this is fun I don't think these kind of tests are of much use. To really tell how good sw or Hw is, one needs to hear it in context of a multitrack - Hw really shines when used on many synth parts at once - sum of multiple SW tracks doesn't give as much clarity.

[Ingonator]

I just noticed that the OP had mentioned using an "Eurorack modular" synth. It would have been nice to know details about the modules involved. Along the huge amount of Eurorack modules there should be good and bad ones like with software. A comparison of a Eurorack system only seems to make really sense if comparable modules are used than in the softsynth that it is comparted to.

Concerning ACE (also called "Any Cable Everywhere") the difference to a hardware modular system is that there is no difference betwen audio and modulation signals so you indeed could plug any output to any available input. This means you could do audio rate modulations where this is usually not possible in other synths, including hardware. The filters in ACE do not seem to have the same quality as those in Diva in terms of replicating real analog filters while the ACE filters sound good in their own way.

If doing such comparisons a synth with fixed architecture seems to be better and even better a synth where a dedicated emulation exists. Anyway i had given up on posting such comparisons of hardware and software a while ago as it does not seem to make much sense and it will not change the opinions of those who prefer software anyway, no matter how good the hardware sounds.

Personally while i got tons of plugins i prefer having at least a few good sounding hardware synths which should include at least one real analog synth. At the moment besides the real analog Pulse 2 i also got the hybrid Korg DW-8000 (which actually could sound more analog than some VCO based synths...) and Ensoniq ESQ-1 synths besides some fully digital and/or VA hardware synths.
FWIW as VA and/or emulation plugins i got e.g. U-He ACE, U-He DIVA, most Xils Lab plugins including e.g. XILS 4, Tone2 Saurus 2, NI Monark, Waves Element 2, TimewARP 2600, Poly-Ana, Arturia V-Collection 4 and more...

[HardSinc]

Here's a interesting question...

How many 'prefer' software but then use analog emulations of instruments or fx or better yet, nebula?

Kind of arguing the imitation is better than the real thing if so.....

[tasty tatsyn]

https://www.youtube.com/watch?v=S1h07ws--CM

it's not a blind test but an awesome honest review imo and preety obvious to me a couple things
1 - A pineaple is a pineapple
2 - ... and an apple is an apple although there's a lot of apple varieties
2.1 - and even in the same variety there's diferent trees
2.2 - even two apple from the other side of the same tree can taste diferent...

so is it better, worse!? i guess it depends on each of us... on some things it might be better one, on others another one

2.3 - the tastiest apple out of the tree is not always the best to make a specific apple desert....

*I'm not sure i've explained myself correctly

[Cinebient]

https://www.youtube.com/watch?v=S1h07ws--CM

it's not a blind test but an awesome honest review imo and preety obvious to me a couple things
1 - A pineaple is a pineapple
2 - ... and an apple is an apple although there's a lot of apple varieties
2.1 - and even in the same variety there's diferent trees
2.2 - even two apple from the other side of the same tree can taste diferent...

so is it better, worse!? i guess it depends on each of us... on some things it might be better one, on others another one

2.3 - the tastiest apple out of the tree is not always the best to make a specific apple desert....

*I'm not sure i've explained myself correctly

I remember later he also compared the Moog Model D app against the hardware.
I also remember he said that the app has the best filter section of all of the softwares.
And i would agree (beside that there is also the same filter in P900 which i find even much better). Personally i also find the feedback/saturation in the app the best (again, beside P900) as well as the included FX which i like most in the app as well (well, beside.....you know).
https://www.youtube.com/watch?v=ARxrOAj5GzU&t=748s

[tasty tatsyn]

https://www.youtube.com/watch?v=S1h07ws--CM

it's not a blind test but an awesome honest review imo and preety obvious to me a couple things
1 - A pineaple is a pineapple
2 - ... and an apple is an apple although there's a lot of apple varieties
2.1 - and even in the same variety there's diferent trees
2.2 - even two apple from the other side of the same tree can taste diferent...

so is it better, worse!? i guess it depends on each of us... on some things it might be better one, on others another one

2.3 - the tastiest apple out of the tree is not always the best to make a specific apple desert....

*I'm not sure i've explained myself correctly

I remember later he also compared the Moog Model D app against the hardware.
I also remember he said that the app has the best filter section of all of the softwares.
And i would agree (beside that there is also the same filter in P900 which i find even much better). Personally i also find the feedback/saturation in the app the best (again, beside P900) as well as the included FX which i like most in the app as well (well, beside.....you know).
https://www.youtube.com/watch?v=ARxrOAj5GzU&t=748s

~~~~~~
in his words.... "reaaaallly impressive..... Amazing isn't it?"

if i even prefered the app in a lot of things, what i ment in previous post was that just in sound wise + hardware hands on inspiration (can be a thing too...) from a personal perspective, the overdrive and FM alone might be enough for someone to prefer the Model D

Quick question, the P900 is just for mac, correct? also sounds awesome

[BONES]

The analog sounds much better. I mean.. it's not even close in my opinion.

Did you listen first and then find out which was which? I bet you didn't because you couldn't afford to have your illusions shattered so positive reinforcement was your only option. For some reason my laptop will only play them through it's own speakers, not through my I/O device, and through those tinny things they sound absolutely identical. So say "not even close" reveals your bias because they are very close.

The giveaway for me in this test is the quality of the resonance, and in the bass track, the low end fullness is like night and day.

My favourite filter resonance comes from software, hands down. Nothing has ever come close to Wasp, one of the synths included in Synapse Audio's Orion.

I liked the software better in the majority of them but the difference is so subtle it's not even worth caring.

Exactly. None of those sounds are particularly exciting and no-one is going to notice the difference in a mix, much less care.

Anyway, I can't look at the cheat sheet, it seems the site is down, so here are my picks, totally blind -
Sample 1 - I can't honestly pick one over the other.
Sample 2 - I prefer the first one, it has a more interesting range in the resonance.
Sample 3 - I like the second one better, it feels like it has a little more resonance dialed in.
Sample 4 - I prefer the first one, the other sounds a little lifeless/flat.
Sample 5 - I prefer the first one, it sounds ever so slightly fuller.

So, how did I do?

[tasty tatsyn]

i prefer the second sample in all 5 examples
without knowing wich is wich because it's not available!

So, i think i'm not helping much
...granted,.... except in example 5, the 1st sections could get close to 2nd using a compressor after the synth

nevertheless... i still have U-he Ace and there's something really nice about it...
although i haven't been using it much (simply because i use other synths) i don't think i'll ever want to lose it! .... every now and then i open it up and...wow!

[BONES]

For example, I have a Boomstar 4075, which was about $1,000. Sounds great, super flexible monosynth. I also have Ace, which was $200, and then I have to use my $1,200 computer which is optimized for audio, and a $300 USB soundcard to even make a sound. Since there's no standalone version, I also have to use a DAW as part of the package.

No, you don't have to spend all that money, you choose to spend it. I can play every song off all five of our albums on the five year old Surface Pro 2 I bought on eBay for $200 last year. And, in total, I reckon I spent less than $400 on the software that we used to make the first four of those five albums. ASIO4all works well enough that you don't even need an audio I/O device.

I find it somewhat fussy to have to boot the computer and load the DAW and do the rest of the process, so it can be a couple of minutes to get to work.

Well, think how much worse it would be if you had a hardware studio. You'd have take off all the dust covers before you could start up the half-a-dozen or more machines you'd need, some of which might take longer to boot than your computer. And you have to do it in the right order so everything can see and talk to everything else. It is way, way less immediate.

I realize that the Boomstar's real price is even cheaper that it seems because right now I have a unit I could sell for about $650 (based on the asking price online), and historical trends suggest that it will hold at that price or even appreciate in value, whereas the prices on used computer equipment and sound cards tend to drop much more rapidly. Laptop computers from the early 2000's are worthless, for example, whereas even analog synths that had huge production runs, like the Prophet 5, still command good prices. Software drops in value even faster from what I've seen.

This is also bullshit. I used to spend thousands of dollars a year on hardware, I only spend hundreds now on software. I buy a new laptop every year, so my old one is usually still worth good money. In fact, this year I expect to sell last year's laptop for more than I paid for it, as prices on eBay are about $300 above what I got mine for new. Software depreciation is irrelevant because I never have to sell my software so I can afford to buy more, which means a synth I used on a song in 2000 is still around if I want to perform that song again tomorrow. It saves so much time.

So for the monosynth function, Boomstar was about $350 (price-value) whereas Ace is about $1700.

How about you compare apples with apples? RePro-1 is only $149 and it comes with RePro-5 for free. So for less than half the turn-around price of your Boomstar you get two synths, one of which is polyphonic and both of which not only sound every bit as good as your hardware but also have way more functionality. They never break down, you don't have to obsess over them to keep dust out of the pots/faders, you don't have to find yet another power outlet for their power cord, you don't have to waste money on stands to put them on and you don't have to spend even more money on a mixer so you can play them with the rest of your stuff. Yes, you need a computer, so if all you want to do is play a single instrument then it might be a relevant expense, although in that case any old computer would do the job. i.e. Nobody buys a computer to play a single VSTi, they buy a computer to run a full production suite of applications and plugins. Think of it s the same need in hardware to have a mixer, which you can spend many thousands on if you want to chase the quality of a software production environment.

I'm sorry, is that sorta the opposite of the way we're supposed to think about this? You won't be able to take away my Ace until you pull it from my cold dead hands, but I'm not under any illusions about the price or convenience advantage it supposedly has.

No, you are completely deluded on that front. Either that or just another clueless idiot.

[teilo]

Bones, you realize you are responding to three-year-old posts, and so are basically taking to yourself?

[johnwoo]

I think is how any of us is used to do things . I am in love with my Yamaha EX7 , my Korg Triton and my Korg M3 . Is something magic when I fire all three and start to improvise . The same thing happened with the vst software but with less magic . Software is good but is always , play something , record , overdub , load another vst , play again . Hardware is there , touch and play . I don't have Moog or Alesis Andromeda (wishful thinking) , but what I have has been good to me and give me beautiful moments .