Digital VS Analog with Ruben Tilgner, Elysia | Superbooth 2024

Great discussion about re-creating transients in the digital domain. Sheds a lot of light on why analog sounds so good

Perfect interview, I'm looking forward to the follow-up!

Some may perhaps find a little frustrating that digital emulations are still not that far as expected. I'm not that affected by this, since my favorite playground is early digital pre-DX sound, which would actually be some further material for discussions, as it's a little noisy but doesn't cut through any less than fully analogue machines. Concretely I mean YAMAHA's GS series.

Nice comments, I think that a lot of the bit hit songs, were and will be produced with analog and reduced recourses, Creativity seems better than having an unlimited supply of plugins, Also the skills to use analog needed are higher. Thanks Terence

It's like we are evolved to hear analogue generated sounds. It's always going to be a challenge to make digital sound the same or "better." Certain parameters yes like noise floor but the character of the sound ismore tricky. The transient discussion was fascinating especially the character of the beginning of the sound. Like we evolved to hear a twig cracking in the forest.

Anthony you call them instruments because you're an awesome artist that understands music thoroughly. You connect with your synthesizers (and don't forget the pedals) on another level you're awesome dude thanks for another video. ✌️🎶

I can't believe there's a sign that says the Water is not drinkable / potable

Who would drink from a fountain hahhahaha

this digital vs analog sounds boring. Use whatever you have available! when they are listening the music they have no ideia if you used digital or analog or that plugin or this plugin etc.

Digital now sounds the same or better. It’s no longer a question or a challenge. you can get a 5000 USD synth in a 100 USD plugin, but now you need a 2000 USD macbook also. The only company that is gaining is Apple.

In some analog synths, you can hear the transient larger or a little more inconsistent than a digital synth. I've always said a lot of the analog vibe comes from the transient. Sometimes I'll use transient shapers or expanders on my soft synths to make them feel a bit more analog-y -- if they call for it.

If digital sounds good enough to be used by Hans Zimmer in big movies then I think I'm fine with it as well.

Thanks so much for these interviews and interactions over there! They are so interesting 🙏

😎

Fantastic ❤🔥🙏

When I was a kid in the 90s just getting into electric guitar, it was the golden age of “hype.” By the year 2000 people were hyping up stuff like using silver solder (don’t do it! You get colder/weaker solder joints). Luckily my father was an electrical engineer. He taught me a lot about “hype.”

The most important thing he taught me was that the human ear uses the phase relationship between high and low frequencies to tell us where a sound is coming from. It’s how we know if someone is talking next to us or behind us. Digital processing will always have a bandwidth. At the high and low ends of that bandwidth, the frequencies essentially get chopped off. This messes with the phase relationship between high and low frequencies.

People talk about digital audio sounding more 2-Dimensional. And to an extent, they’re correct. I remember the conversion from vinyl and cassette tape to CDs. I don’t think it’s at all coincidental that as soon as digital became the standard medium for music it was immediately followed by the invention of “surround sound.”

💐💐💐

Just love the new inventions of these beautiful instruments best wishes.❤

I think if everyone could own a moog or a prophet they would .. i always said i dont need no analog synthesizer and that it was over rated till i purchased a moog matriarch and then i was like okay i get it now.. and im addicted .. i have a OB6 on the way 😊

I hoping you cover THE FAIRLIGHT synth/emulator from the 1980s. I know they are rare big hunky machines and if anybody can do it your channel can.

❤ clear difference between the artist and the technician. Playing is not producing for music industry.. I have some early synths, playing, and can compare to test records on internet, 33rpm, CD, and I have software instrument like Korg polysix and monopoly, wave records, a Roland Jupiter 50, based on samples and numeric, one Behringer Deepmind 12D, wich is Analog and numeric effects that I mostly bypass, for me the pleasure to play music is lot more than technology, and the better pleasure is clearly to enter in the harmonics world, not the technical passiv listening. By past I tuned a real piano strings, good school for harmonics treasure. That's why also analogic is better as playing because this is only an association of instruments, voices, and notes, and your proper touch in that... complex technology is oht of price for me and constantly in evolution meaning you change your material, it is cancel culture, you are lost is wanting what you have not. My old Roland SH 2000 have secret sounds and presets, an amazing after touch, and this is a pleasure to play. I will buy for it an analogic effect, I use actually old digital, Behringer Rack box, it is in the same line, 50 years old technology facing the actual mess changing every six months, pretending being better than the previous. Note I have a clarinet and recorders, and we also must count them analogic, guitars, sounds of Nature, this is more a philosophy than a technical professional technical issue. What's makes me happy is that, actually, we have easy access to rebuilds of early analog instruments, as players. Personally, I don't care about industry, obsessional rythmic of gigantic cities . I live in Nature, far from that. 😂so I'm certainly not representative ❤

I think I've just come up with a kind of explanation of effects discussed in the video. When two waves interact nonlineary, they usually create so-called sidebands with frequencies equal to sum and difference of initial ones. That's why we can get non-audible frequencies out of audible and at the same time sidebands of this non-audibal frequencies can appear in audio range. For istance, two waves with frequencies 22 kHz and 23 kHz can create a 1 kHz wave

Ruben is such a great guy! I would love to see a discussion between Ruben and Mark Barton. Digital is great when it comes to recording (AD conversion) and limited in producing complex signals. That's what I said before about the difference between recording an analog synth performance vs playing the same performance on a plugin synth. Analog signals are too complex to faithfully generate digitally.

Best analog processing company in my eyes because of its transparent and open communication, but assembly and QA done in-house. And as an European I don’t need to buy products from overseas or outside EU imports, which makes the price very achievable. Have Xpressor, Filter, and Nvelope - all of them are super fast to use with instant great results but they are also deep enough to grow with them. And these processors don’t lose the slightest detail when putting a signal through it. Highly recommend

Very interesting! This confirms my feeling regarding plugins and digital synths, it's like wav vs mp3. It's also interesting that there is a group of people who clearly hear and feel that something is amiss in digital sound, and another group who can't hear it, or don't care.
To me, it's like when someone says stones are grey and water is blue.

That's one of the best explanation I head. Thanks a lot Anthony for sharing this in my opinion very important topic. There's a reason why I use a lot of hardware instruments, hardware post processing, record an edit with digital plug-ins in my Daw at 96KHz. This interview explains it so well

I love the comment about digital being limited when it tries to emulate analog. The same could be said for analog when it was first introduced; so many presets were created back then to emulate acoustic instruments. For the most part, they’re not great emulations. However, when analog is used as its own instrument, it’s limitless.

Awesome discussion. There is so much fun with hardware analog gear! Digital synthesizers for me usually are workstations. But for software, Cherry Audio 🍒 really rocks on Ableton Live!

glad this conversation is happening, but I don’t think it matters. What matters to me is songwriting and musicianship. I’d rather hear good music played on a diddly bo through a transistor radio than generic pop garbage on a Strad violin. That’s why I LOVE digital - so many tools, so cheap! Then comes the hard part - musicianshiplitude

In a mix it doesn’t really matter that much. I have a full rack of Elysia and it is luxurious. Analog is more about immediacy, feedback and irregularity than it is about a clear edge in sound quality. Recall and speed in digital matter too. Especially if you are working with multiple artists.

Digital : 0 or 1

Analog : 0 , all the values in between , 1
Example : 0.0001 , 0.0012 , etc etc
All this values are missing in the digital world

I was putting the prophet rev 4 against repro 5 and it wasn’t even close.

I take from this talk that Analog is pure, and seems way to complex for a digital path to really capture pure analog or improve in any way. But it’s different tools for different things


I’m always gonna love an Esq1 the same as I love a minimoog digital / analog bring different flavours and that’s what I love about this synth world

That was a really interesting conversation! I got a lot out of that one. Thanks, Anthony!

I love digital for its convenience, it's a must. But I fell in love with hardware long ago, there is just something so special with some of these pieces that digital can just never touch. I have to have a chain of hardware as an insert on the 2 bus, otherwise the mix sounds lifeless in comparison. In the end digital/analog is a marriage made in heaven as they are both so complimentary when used properly together.

I appreciate this gentleman's point of view and no doubt some aspects of what he has said have some relevance but for the most part it's not an issue. The issues he's talking about with emulation of digital processing have a number of usable workarounds and solutions that render equivalent sound quality. For example most 64 bit digital synths and fx use much higher sampling rates in their internal processing.

Love your energy, Anthony. The way you engage with others is deeply empathetic yet authentic to your own knowledge, great for communicating these ideas to the rest of us!

I love how deep your analog digital question turned out to be, this is an excellent series. Each video has taught me so much. I like how you can liken audio resolution to image resolution, digital can only replicate so much of the "sound image"

I really appreciated all the info.. Went to check on the site and I see a compressor for 10 grand. How are musicians supposed to be able to access these kinds of units when Spotify are paying nothing peanuts? I mean, what do you want me to say? I went from being enthusiastic and wiling to sacrifice things to just feeling like Im being trolled by these prices. I wont sound as good as rich people but I'm gonna make better music than all of them now I see the perspective .. pay to win

It's weird how many people make subjects like this so needlessly complicated. VOLTAGE is the difference. Sure, non-linearity plays a roll if you're summing through an analog mixer, but voltage is the key and also why there are different stages of analog goodness. Consumer gear is 9v (Behringer), pro-sumer gear is a little better at 18v (DBX), and the pro gear is best at 24v and higher. Rupert Neve was so clear about this when he constantly explained why he used 48v rails. Pro mastering desks use 120v rails. The higher the voltage, the higher the headroom, and that means depth... done!

The point about transient response is interesting, but I don't think it's actually kind of 2 different points.

For something like an envelope for a filter or vca, digital sample rate should be totally fine as often those only go down to 1ms minimum attack (which is 1khz).

The problem comes with reproducing the much much faster transient changes in something like a square or pulse oscillator, because those harmonics are going to become very high frequency very quickly. A true square wave has all the odd harmonics of the fundamental frequency at a level of 1/harmonic number.

So, already at 1khz fundamental, you have -9.5 dB at 3khz, and then 5khz, 7khz, 9khz, 11khz, etc. Then you're reaching 20khz at -33dB approximately (1/20). So, that's already close to nyquist and at still an audible component of the signal.

If your fundamental is a little higher at say 2khz, then those harmonics are twice as high frequency and you are already passing nyquist at the 10th harmonic, which is -30dB. etc etc.

So, very quickly you start to lose ability to represent the square wave without aliasing.

The same is true for sawtooth, but they also contain all the even harmonics as well as the odd.

So yeah, he's totally.correct about that especially when it comes to modulating those pitches. One place you can hear a lot of aliasing is vibrato effects of higher pitches if not properly handled.

Emulations are still like animations

The perception of brain is very delicate. It is a question between imagination and impression...

If we have no problem on playback reproducing symphonic orchestra with 48k SR - why it would be a problem for synths? For transients? The math says there's no problem. No problem with "dots" and high freqs - the only solution to equation is the real sawtooth/square, not that we cannot reproduce them up to 1/2 nyqwist - we can and we do, perfectly accurate.
The problem with digital synths is that most of them does not use full SR for "CV", for LFOs and EGs — that is the problem. 
Nonlinear processing? — the only problem is it takes a huge resources to built a good model of analog nonlinear to be precice enough to fool our perception. No other problems. We have a good digital nonlinear, we just have so much more analog stuff no one cares to model. Plus we love to touch things. Digital + analog + physical = the real deal.

I think he is wrong about the part about not emulating the input stages of analog gear though, most companies do this.

Well that escalated quickly! Great conversation, read my mind

hey whateup dudeo. i'm Saul Pickett. I discovered a thing. so if you get a big rock, and like, some chisels , and like cut every vertex off such that it becomesa "wheel" of sorts.... well. you put the newly shaped rock on top of a synthesizer, and now you have something calleda gramophone or turntable. The fact that the turntable or reel to reel platter has MASS is a very important thing . No matter if you record or repro, the playback speed will be different on any device, analog or digital. BUT. crystal oscillators , such as in cellphones. or recording 24 tracks at once versus 1 track during an overdub... This consumes Slight but "significant relative to clock stability" amount of power the supply can generate. Ever wonder why when you punch in, saya bass overdub on digital.... "you can always tell it was dubbed somehow..." thats because the pitch. the pitch when you were tracking 24 channels live is different in the converter than if you do the overdub, because less power is being consumed during the dub or whatever. if your converters are plugged directly in, you have nothing. every single audio channel, midi channel, whatever, anything connected to the converters has to be transformer isolated period, that is a whole nother discussion. analog tape has its own problems. it always sounds "rectified". it cannot "listen to negative voltages." if you were to run your music through a full wave rectifier... well. that is exactly how tape records audio. the difference is... the TIME aspect, is often not very good but sometimes nearly perfect. verusus digital, the clock on digital is terrible terrible terrible in comparison to tape. as far as ... lengths of time and all this. tape / turntable is 100% reactive. digital is like a vocoder with a couple hundred poles. you can end up with fundamentals and overtones that did not exist during the performance, with digital, as in. the notes and times and tempos and pitches themselves change. of course it will on any recording, but analog , if exactly perfectly setup. i mean agonizingly perfectly setup. can produce more accurate information in terms of When events happened in time, more stable downbeats. just cause a computer has 5 decimal points doesn't mean it is accurate to 5 decimal points, and it might stray by the time you start and end the recording. just saying. people said analog is all about tone, brah, distortion, etc. nah man. analog is about clean, its about downbeat stability.

It's like trying to measure a continent's oceanic shoreline it's technically infinite granularity so we cant truly ever hope to get a fully accurate simulation of every possible nuance or possible variable that actually influences the sound that ultimately is what hits our ears. I do think one of things is the air that an amplifier moves that in turn moves the mic diaphragm and how that interaction is like the idea of causality with the butterfly flapping it's wings. How much each molecule of air being pushed needs to be simulated to make a simulation of that to be impossible to discern a simulation from the real thing?

digital and analog will never sound like well, analogues of one another for a very simple reason. analog circuit design starts off as pure chaos that you gradually refine into something usable on a signal. digital audio is an empty box where you try and trick linear systems into being chaotic. i don't think the amount of samples in a signal has much of anything to do with why transients are better represented in some cases in analog. interpolation can capture most of what analog does at most sampling rates. that is obviously not the issue.