Comment by pmoriarty | Hacker News Reader

Comment by pmoriarty | Hacker News Reader

"Digital audio workstations have become increasingly sophisticated, able to emulate with "plug-ins" the capabilities of sought-after analog studio gear of the past. It has taken a bit longer for virtual instruments to meet this same standard, but they may be nearly there."

Hardware is much more hands on, with real dials, switches, and other means of physical interaction which screens and mice don't come close to matching. Hardware does exist through which musicians can interact with computers, but it's usually MIDI, which has its own limitations compared to fully analog gear.

With hardware you generally don't have to worry about software or OS updates, and the hardware generally continues to "just work" for a long time.

A lot of uncommon analog hardware is very unlikely to ever be modeled in software, so the way it reacts when manipulated may never actually be matched in software (even when the potential to do so is in principle there if anyone ever bothered to create a high quality model of it in software). This extends to how it sounds as well, for the same reason.

That said, modular synth software like this is still great, and may be the next best thing to using the real thing. It's also a great way to learn and get a taste for what's possible before committing money to buying any hardware.

Yes. The phrases "menu" and "software update" are what kills all tactile electronic instruments of various kinds.

This is the same for electronic test equipment as well. There are so many features crammed in new devices, even if you manage to get enough controls for the basic cases on the front panel, a lot is still hidden in nested menus galore and they aren't necessarily sensible or discoverable or even work properly. This results in semi-religious fighting over which update is the least shit and people refusing to accept updates because they fear them.

The outcome, at least with electronics, is that you'll find a lot of 30-40 year old equipment nestling next to the state of the art with a sticker on it "for indication only". This equipment does a lot of the real work while the new stuff is used for validation and traceable measurements only. Due to rapid progress in technology, people cling to the last thing they felt any physical connection with.

This will pass in time as the status quo is accepted for younger generations. Fly by wire it is.

(I am incidentally a rather large fan of analogue and modular synthesizers and half built one in the late 1990s with 1970's electronic equipment, some of which I still own!)

The UX of digital equipment _really_ sucks. At first I thought it was because it was a new technology still in infancy, but things haven't improved since and we are still emulating analog controls. VST interfaces often have static background image (skin) and 'knobs' you control with your mouse. Actual hardware usually features rotary encoder or buttons to navigate the menu system with several layers of depth. This is tedious and hampers creativity. One positive example is Teenage Engineering OP-1, they seem to know what they are doing.

I'm currently building digital instruments for android phones and I refuse to add any UI elements until I figure out something that makes sense. Until then the interpreted Lua code is the user interface. I would be very interested in some new UI patterns for touchscreen that take good advantage of interaction between user actions and visual feedback.

It is completely baffling why software synths and routing software etc. feels the need to make itself look like a patch board and have rotary knobs and things. Can't we come up with a set of conventions for software-native synth UIs? Sure the physicality of real knobs and sliders is nice on real hardware, but on a screen - especially non-touch systems - there's no way it's the best option.

Exactly THIS!

> but on a screen - especially non-touch systems - there's no way it's the best option

There is a good option - but the current, dumb UX fad, rejects it whoreheartedly. That option is keyboard. Give everything a keyboard shortcut and let the user learn them (if at least by providing a visual overlay listing shortcuts straight on the interface).

If something is a tool - not a shiny cloud SaaS toy intent on getting audience to drive the hockey-stick graph straight into acquihire - users will learn keyboard shortcuts.

I always let my users click on a knob to direct enter a value with the keyboard. But keyboard shortcuts to increase / decrease a value -- that's just a pain. Mouse drag / mouse wheel is far faster at going between coarse and fine adjust.

Knobs are still the best option I've come up with, they act like sliders with the mouse, yet they are smaller so you can have more density. You can also draw a longer track for a given control size, 2pir > 2*r. They often get assigned to a knob on a midi controller anyway, so it makes sense.

My knobs look like software, I refuse to do screw heads, metal finishes, etc.

I haven't done much touch screen software, so no comments there.

There are software synths that don't follow the "virtual patch panel" concept: Audiomulch and Supercollider come to mind.

Audiomulch has a graphical interface. Signals are processed by constructing a graph where each line represents an audio channel that can be routed between blocks that perform specific functions.

Supercollider is more like a programming language, you can think of it as a DSL for audio generation.

Both of them can interface with hardware, of course.

I think there will always be a strong demand for tactile hardware interfaces, regardless of what people come up with on a laptop application. The kinaesthetics of making sound by manipulating something physical is so satisfying and intimate.

You probably mean software equipment, because there is lots of digital music equipment with good UX. See for example Mutable Instruments' Tides or Rings modules, 4ms Spectral Multiband Resonator etc. - no menus, no displays, one knob one function (well there are sometimes 'special modes' and easter eggs of course).

For those boxes, 'digital' is just implementation detail. User interface and interface towards other equipment is still completely analog. I agree that they look very clean and usable.

I don't think there's anything wrong with displays. They offer immediate and quick insight into internal state and operation. The issues appear when you introduce modal workflow, with some features only available in some modes.

Worst offenders in category of hardware equipment are probably guitar multi-effects and synthesizer workstations. They cram so much functionality with so little thought given to flow and usability.

Ableton Live in conjunction with the Push is a fantastic, very tactile, very musical solution to this. The Push is a very well-designed hardware instrument that, because of its tight integration with Ableton, has much of the flexibility of software. It's a really interesting approach and very productive from a music and creativity standpoint. This YouTube video gives a good sense of it:

https://www.youtube.com/watch?v=0CdMvkBOUgs

I believe quite strongly that for most people, utilizing hardware of some kind (pad controllers, keyboards, etc.) is essential for making music. Music is generally made by playing instruments. It is the small errors, the groove, the emphasis, the spirit/heart/soul of playing an instrument, that makes music feel alive and interesting and vital. If you're just clicking around with a mouse, you're programming it, and programmed music sounds programmed. Overly precise, inhuman, etc.

In Glenn Gould's debut on US television, conductor Leonard Bernstein talks about - and then demonstrates - the difference between rote, mechanical playing of a score (i.e. "programmed music") vs. a performer's interpretation of it (i.e. the heart & soul that I'm talking about). The discussion starts around here:

https://www.youtube.com/watch?v=9Nx09pigZRI&feature=youtu.be...

I would absolutely love to one day own an OP-1. Teenage Engineering does amazing work. All of their stuff has a certain tactility to it. They definitely put a lot of thought and love into their work.

For those that can't justify spending about $1k for one, their Pocket Operator series is pretty fantastic.

I also recently bought myself a Novation Circuit, which honestly has the feeling of Teenage Engineering's interface design (minus a lot of the whimsy). For about $300, it's a fantastic little groovebox that I'll be getting a LOT of mileage out of.

Edit: formatting.

Also worth mentioning is OTTO [1], an open source effort to recreate OP-1 on Linux / RaspberryPi.

[1] https://github.com/topisani/OTTO

I haven't tried it, but the idea is brilliant. I would love to be able to use the interface on a $5 machine...

I would also feel compelled to still buy an OP-1, though, since they obviously put a lot of effort into R&D.

I'm sitting in a room full of synths from 1971 through 2017 and I still think my OP1 is one of the most delightful, amazing things in here. You can find them used for around $600.

I haven't seen even the used ones go less than about 800 USD. Where are you looking? (I'd honestly love to know...)

Yeah i'm a fun of scrounging around for cheap little used modules like Mopho, ms2k, ms20 mini, mother32, mini/monologue, the Roland and yamaha recreations jp08, bass station 2 and various novas from Novation, emu proteus 2k. Roland actually has released dozens of rackmounts, some quite decent. Also kits: Preen, audiothingies, mutable instruments etc. It's astounding what you can buy for under $300 in many cases.

I do wish for a cheap knobby controller like the maudio axiom or novation SL mk2 that has reliable faders, pots and encoders. Those are the one instrument product i would say you should buy new and get the extended warranty from Sam ash or Guitar center. When the pots start shooting out lots of spurious midi CC's, deoxit's not going to save you. Deoxit's great for making door hinges stop squeaking tho

And importantly, it's a way to get into synths without breaking the bank on "fixed layout" hardware, even if it's supposed to be awesome hardware. If you're on a $10/week allowance, projects like these are fantastic.

I concur! As someone who has spent, probably, close to a quarter of a million bucks on synth/audio hardware over the decades, I'm extremely keen on VCV Rack because it lets me design my "perfect" Eurorack setup without any investment, other than the time to get it set up - and more to the point, it lets me experiment with my setup, refining it, without buying a single cable. So once I get my stack configured the way I like it, I'm quite likely to go ahead and switch over to reality and buy all the modules I've used, so I've got it in the real world too.

A truly amazing bit of software, and it has also been a very rewarding experience to be an early adopter of this suite, because the progress made so far by the developers is nothing short of astonishing - not to mention the uptake and adoption by a lot of the more cocktail/unique modular designers out there ... many of whom, seem to have taken VCV Rack as their new standard test-bench for working out ideas before hardware prototyping - meaning we have a huge plethora of modules to play with!

Anyone interested in synthesisers NEEDS to put a weekend or two worth of effort into learning VCV Rack and creating their own stacks. Its really a very, very valuable bit of software - and its free! And its open source! So, you interested-in-programming-synthesizer-algorithm folks can get started with very, very little fuss, already ..

Totally agree.

Reply on news.ycombinator.com