/tech/ - Technology★

stop using commercial solutions and support the open source ones.

>>1004668 (OP)

Video: Kdenlive

Audio: Audacity, LMMS and Bitwig Studio

Btw are there any DAWs for BSDs?

>>1004672

There's Fruity Loops on macOS :^)

>>1004668 (OP)

Is AV Linux or that one av-oriented ubuntu still a thing? Used that back in the day.

As I recall, at one point in time "Hollywood ran on Linux"

>>1004681

I remember that, Hollywood used their own in-house proprietary applications though

>>1004668 (OP)

I used to do some music production as a hobby, but most commercial plugins are distributed as magic proprietary DRM software that only has a 20% chance of running through WINE, and unless everyone magically starts using GNU/Linux, it's unlikely any of the companies in charge of those are going to do ports.

Audacity is ok for pure audio editing, but I have had issues with recording with pukeaudio. LMMS runs a lot better than on Windows, but it still has some issues that make it a pain in the ass to use (beat sequences suddenly disappearing in the GUI; time signature changes resize the entire timeline; parameter automation is awkward as fuck). I've only ever tried running other DAWs while running Arch but none of them seemed to work properly.

Kdenlive is about the closest thing to working video production on Linux, but even it has some stability issues. Everything else I've tried is shit. Openshot seems to expect you to slice clips using your mouse alone, and it has random memory leaks that make my computer freeze up after several hours; Shotcut likes to crash for no apparent reason; and both of them have amazingly horrendous performance when it comes to manipulating large clips.

For what it's worth, there's a modified Gentoo distro designed for audio workstations, but I've never tried it:

https://gentoostudio.org/

>>1004668 (OP)

Check out supercollider, pure data, vcv rack, foxdot, csound, pyo if you want to make music on linux.

I believe coding music is the linux way of making music. DAW's is the windows/fagos way.

>>1004727

I put most of the Gentoo Studio packages on CloverOS and I'm considering making an optional kernel configured for audio. I might do that today. If there's anything I'm missing or shouldn't overlook, reply.

>>1005068

>I believe coding music is the linux way of making music.

I agree but I'm not a big fan of existing solutions. Out of the ones I've tried, most of them were built for live performance and the ones that aren't are wonky and clumsy at best. I'd like something like a MIDI file but hand-writable, with more features and some conveniences like macros or reusable sequences. I've been thinking of the features and syntax I'd want for over a year; maybe I'll implement it some day. In the meantime, I'm thinking of using Lua's table syntax to build a toy prototype.

L = instrument{comment='Lead', voice="dave's cool piano 7", ...}
R = sound{comment='Rythm', voice='nailsonchalkboard.wma', ...}
B = bpm(230) -- or, roughly, seconds(0.26)
T = tuning(440, 'a-major')
nice_melody = {T, B*2, T+4, B*2, T+4, B, T+5, B, T+4, B, T+3, B, T+1, B*2, T+1, B*2, T+2, ...}
song = parallel{
	{L, nice_melody},
	loop{R, B},
}

The tables are flattened where possible, lengths of parallel sequences are checked and a state machine zips through the thing. Hitting an instrument changes the instrument, hitting a tuned note plays the current instrument, hitting a sound just plays the sound, hitting a time value moves forward in time. This is obviously an incomplete example but it shows the gist I think, e.g. there could be objects that cause notes to timestep automatically to reduce the timing (B, B*2...) punctuation.

>>1005112

Peak autism, a site to behold

>>1005126

*Sight

>>1005126

I'd be curious to know how you end up on /tech/ if you think elementary programming is peak autism.

>>1005143

On second thought, keep doing what you're doing. This is better than a lot of solutions out there. Given the embeddable nature of the language you are working with you could incorporate this notation into a VCV Rack module. That would very a neat thing indeed to have live coding going on at the same time as analog-style patching.

>>1004668 (OP)

I have never found any music production software on any operating system that I would consider acceptable so I just do everything in Python with numpy/scipy. The reason is that I don't like equal temperament and note frequencies are often hardcoded in audio software. In the few that aren't hardcoded they still don't support changing your fundamental frequency mid-song which is quite limiting.

For the little I need I use gimp for shoops and inkscape for diagrams (or tikz/latex depending what I want). I've heard blender is good with video but never used it.

>>1005130

/tech/ is for anyone who likes technology, not just programmers.

>>1005206

Sure, but not in the sense of "I like showing off to all my friends", but in the sense of using and understanding technology. And you can't do much with technology without programming it.

>>1005159

Blender is most certainly NOT good with video.

i still use it though because I cant be bothered to lean something else.

>>1005159

>>1005211

Blender is good with video. The problem is you have to operate it in the ways of Blender. You can work very quickly once you've been trained in it but the training effort is quite significant.

>tech nerds

>being "creative"

kek

I am an amateur composer and video producer. I've been composing my music using LMMS. For video, I've been using Blender.

>>1005214

There's this one guy called ZUN and he has a huge volume of work.

>Ctrl+F

>no DaVinci Resolve

https://www.blackmagicdesign.com/products/davinciresolve

Now supports some major Linux distros.

>>1005159

>gimped

it pains me there's no good PS alternative

krita is for drawing, pinta is in mono hell (and without the godly paint.NET plugins it serves little purpose) and gimp lives up to its name

>>1005247

update your gimp maybe?

GIMP now has the MyPaint brushes integrated, so it's now more like Krita++

>>1004727

Gentoo studio works fine, I've been using it for about 9 months. I'm mostly using it with Renoise and Supercollider.

I had to tinker a bit with jack but didn't run into any problems, it was easy to install for someone who never used Gentoo.

The distro stays really focused on audio, it comes with software for everything music related from sheet music editing to mastering without taking a general multimedia approach like kxstudio.

>>1005213

>blender is good with video

is it as unstable and incomplete as Kdenlive? I'd be willing to learn the Blender Way as long as it was a stable and featureful video editing platform

>>1005323

I'm an amateur trained by Youtube so I don't know how well it compares to anything. I don't know what is the professional grade language and techniques for video editing and production. All I know is that Blender video editing is sufficient for little video advertisements, the kind where you have a few video clips clipped together and you have text flying in an out.

>>1004668 (OP)

I've made music using Ardour (DAW), Hydrogen (midi Drums), and Guitar Rig (VST fx). It all worked great; the PROBLEM was JACK audio. It never saved the configuration properly, so every session began with 5 minutes of routing ports. It was a major impediment.

My Razer Blade 15" came with a free copy of FL Studio and I haven't looked back. It's ridiculously fast and polished, and just a 10x better experience.

>>1004668 (OP)

>audio

i have no fucking clue

>NLE video editing & 3d models

blender

I use Ableton for my music production and I refuse to switch to anything else.

>>1005331

honestly JACK sucks, and there should be MIDI bussing in pulse, the problem with that is pulse sucks and alsa sucks and all audio is a hot pile of shit.

I tried LMMS and wanted to kill myself so bad that I pulled out an old machine, installed win 7 + FL Studio

>>1005733

So are all sound servers made for Linux shit?

audio sucks dick on linux, everything sucks dick on linux, kdenlive is the most stable and capable video editor and it's a piece of shit that crashes if you look at it wrong. everything fucking sucks and i refuse to use windows

>>1005745

correct. I don't think pulse will be fixed, the next generation of pulse is pipewire, also redhat, and it's supposed to be a multimedia server, so just running ffmpeg on the root x window to stream desktop to whatever output will not be possible (debatable if this is actually good, and it's needed for any output on wayland) but I haven't seen anything regarding supporting digital multimedia workstation software so soon starting JACK will take out your webcam too.

>>1005746

I think the bad audio and colour APIs are the main reason why adobe isn't on linux. Sucks, but that's just how it is on this gay earth.

>>1005746

>>1005745

sndio is pretty decent

>>1005843

Only on BSD where it can run without ALSA

>>1005257

<3 renoise

I kinda dream of making linux a renoise based sequencer and a ardour based harddisk recorder. Like, no plugin bullshit, just midi sequencing, orchestra hit samples, analog synths and lexicon outboard.

>>1005112

Where can I do something like

>while(i != a# || cb) if (j == c||a) new harmonicMinor(A).set(1,4,rnd(parentScale),1,5,1,6,7).insertAt(typeof minor || major);

>>1005992

Try pyo.

>>1005987

It would be awesome to have a ncurses based tracker with vi bindings.

OBS for recording, Kdenlive for editing, and Handbrake for uploading (only used if necessary).

I use Bitwig Studio and Milkytracker alongside some external synthesizers/effects, and about 10 GiB of drum machine samples for music making. Bridging VSTi's over to Linux is not very reliable so unless you work primarily with samples and/or external hardware you probably won't enjoy have a good time.

>>1007073

I remember when I was distrohopping I experienced same issue. Try to remove/install pulseaudio.

>>1007073

>Why does audio on linux always crackle when I change the volume on a video?

If you're on PA, maybe change your module-suspend-on-idle timeout or just disable the module altogether (but read its documentation first either way). If that solves nothing, maybe some settings are improperly set. Sub-optimal buffer size or latency? Not sure why those would cause your problem but check them anyway. On PA, you would look at the buffer and fragment settings. Otherwise, it might just be poor buffer handling on the application's side, especially if you don't use said application's primarily targeted sound server.

>>1007073

Post your specs so I can tell you about how you need better hardware.

Dual boot with an offline Windows 7

>>1005068

more stuff like what this guy says, i too believe coding plus gnu plus linux go hand in hand. someone make a blog style post pls

>>1005112

also incorporate this guy in blog post, idgaf about live performance

>>1005157

BASED as fuck, ur also included in /tech/s music blog post

>>1007255

>>1007256

>>1007257

In summary, pls do this guys. I play only acoustic instruments (piano and guitar) and every now and then I go through a spurt of wanting to step up my game and start composing. But I already spend enough time on the computer, and changing my workflow from analogue acoustic instruments to having my nose back in a GUI based computing experience makes me want to vomit. But I want to create, so please come up with a sufficiently comfy solution.

>>1007259

Lastly, I love Boards of Canada, William Ballinski, etc.. So for most of what I’d like to create, electronic fluency is a necessary prequisite. But dear god I just want screen time minimized, and time spent using a shitty GUI interface counts doubly so imo.

>>1007259

I think my issue is that the vast majority of DAWs have really clunky interfaces (i.e. click, or worse, double-click to place a note, click to extend a note or worse, click on a tool to click on the note to extend the note). I'd be interested in a DAW with a vim-style interface.

I done a few simple videos in openshot

The interface is simple and crude.

>>1007266

>>1007259

If you can afford it, build a separate workstation just for your DAW sessions and use a big ass touch screen monitor built into the desk; Ableton and FL Studio both come to mind as having touch screen modes.

Short of that (or in addition to), get a nice MIDI controller. You can perform most recording tasks without ever touching the keyboard or mouse.

>>1007288

i already have a midi, and i’d love to use it. but the software is so atrocious. i’ll look into what you recommend, thanks for the time.

>>1007305

>midi controller software sucks

Yes I had a similar problem. There is a nice piece of software that will let you arbitrarily map and route your MIDI data. inb4 shill https://www.bome.com/products/miditranslator

I had to use this software to get my Presonus Faderport working with FL Studio.

>>1006787

Yes. If it sequences a Moog Modular you could set it up on a screen as big next to it.

There was something like that though (for modern linux), it fails to run on my machine though.

>>1007288

If you can afford it, go analog

>>1004668 (OP)

>any hope for commercial solutions/VSTs that only support Windows/MacOS to get a Linux port?

Probably not, but airwave is a pretty good bridge for windows (wine) -> linux with jack audio.

I am using pianotech 6, myself and it works really well with carla.

https://www.mediafire.com/file/jts36oycyahlxnx/pianoteq_stage_setup_v622.exe/file

I originally posted this in the "private tracker" thread.

>>995929

Anyway, jack audio is your best asset. ardour is pretty good, etc nondaw is pretty good too. carla is good, everything is good, really, just needs more support and polishing.

>>1007550

i think this is, ironically, the only tech approved way to go here. all the software options are dogshit and my gut tells me they always will be (just my opinion).

what are some cool analog tech things? maybe tape recording some ambient field noises? lol

>>1007902

i mean other than synthesis and acoustic instruments

>>1007902

>>1007903

i really wanna see more musicians workflows examined up close, from people like Boards of Canada and William Ballinski

I'm getting babbys frist guitar in a couple weeks. I can't wait to not install guitarix or any of the other Linux software amps/plugins because jack and pulse absolutely ravage my system any time they come into contact. >>1007550 is right.

>>1005327

Where can I learn the ways of video cuts using Blender? I'm a VirtualDub migrant and for the last couple years I've ended up using ffmpeg to manually and imprecisely cut video by time stamps after tracking in a video player. I want my frame-precise cutting back, but last time I tried Blender it was so confusing I gave up.

Just donate to WINE so they can create a decent layer where to run your Win32 software. Everything else is a pipedream so far.

>>1008270

>WINE

Not really viable for live monitoring because of added latency. Best to just have an offline Windows box at that point.

>>1004672

>Video: Kdenlive

I can vouch that this thing does a better job than Adobe Premier. It supports a wider range of formats than Premier's very narrow support for anything that isn't used for TV/Hollywood. KDen may be unstable, but fuck it does such a better job that Adobe Shitsmeer.

Overall I'm quite pleased with the art software available for Loonix:

>Music

Musescore for music notation

Deflemask for a chiptune tracker

>Drawing

Krita

AZPainter

someone please post or link me a video tutorial or example workflow. I am interested in the styles demonstrated in these post:

>>1005068

>>1005157

>>1005112

>>1007550

>>1006787

>>1008280

>added latency.

How so? Just don't use poetteringaudio

>>1007073

Have you tried unplugging it and then plugging it back in?

Sometimes static can ruin audio.

>>1009787

It's all lies you doofus. Except the tracker one. That's a legit autist who thinks ncurses is some kind of usable UI, priceless stuff.

>>1009789

If you're playing a live instrument through an audio interface to record a track in your DAW, through an FX plugin, then routed out to your headphones (live monitoring), you need to have it under 10ms latency or your ear will notice it is out of phase, and you won't be able to play along with drums, and your performance will just generally suck. It's hard enough to get low latency without a virtualization layer, and pretty much impossible with it.

>>1009968

not true (other than the ncurses autist), and it turns out supercollider is my wet dream

>>1005157

hey idiot have you tried supercollider?

>>1005157

Interesting. I used a guitar tablature program back in the day an experimented with adjusting the pitches of notes a few cents here or there to try different termperments. It was ridiculously time consuming and more of a math exercise, but a lot of fun. Haven't heard anybody talk about this kind of stuff before or since I did that and brought it up with that community.

>>1010538

can't you do this in supercollider?

Probably a retarded question, but is there something like a plugin for either GIMP or Krita that lets you open SAI files? I want to ditch wine+Paint Tool SAI and use only foss tools for art but don't want to convert a shit ton of them to PSD.

>not editing videos with a hex editor

pleb

>>1010595

https://tools.suckless.org/blind/

For most people, kdenlive will do the job. If you think you need Hollywood shit, there's DaVinci Resolve

LISPFAGS BEHOLD

>https://www.cs.cmu.edu/~music/nyquist/

Program your sounds in a Lisp!

>>1007288

If you can afford it, >>1009787

go analog>>1009787

> example workflow

https://www.youtube.com/watch?v=sfrZQzoC6zs

>>1012484

I can’t, but I think someday I’m going to try to make a stand-alone synth with a Raspberry Pi (or a more performant, lower latency sbc) with custom functions from Super Collider or Nyquist.

The only analogue equipment in my price range is a MicroBrute. But I think I’m happy doing some Lisp hacking in Nyquist for now.

>>1004668 (OP)

>Not writing hymns to music based in DIVINELY ORDAINED RNG in TempleOS.

GITDs I swear...

>>1013249

I hated the cacophonous tunes in the Temple OS demos the most.

But I realize that this is a joke post, so yeah.

>>1004668 (OP)

how hard is it to make a supercollider/csound/nyquist powered raspberry pi semi modular lookin synth box

>>1013646

it also has to have a patch bay

>>1009491

Its lack of a cache and preview slowness makes it hard to use.

FOSS trackers, synths and ffmpeg allow you to do basically anything. Just glue stuff you find yourself using with shell scripts.

>>1005207

> pro·gram·ing

> the act or process of planning or writing a program.

Nowadays, you can do pretty much whatever you want with just configuring and,the occasional scripting. Writing actual programs isn't needed amymore in most fields.

In fact, web developers make good buxk coding pages without having to ever program any browser or framework themselves.

>>1009491

>>1013714

I barely do any video editing, but for the few hours I've used kdenlive, I didn't have any problems on ubuntu.

the more hardware you have the better; i.e. if you work with real instruments, mics, synths, drum machines, effects, etc.

If you treat linux like a pro-tools studio before everything was VST there is nothing holding you back.

If you work with audio more than midi you're in luck, for beat-makers this means using lots of loops and sample packs.

If you don't work this way you have to be old school or forward thinking/experimental.

It require lots of time to find a workflow, don't expect the ableton/reason/garageband experience.

The selling point for lunix audio is modularity, but then to have persistant sessions you need a session manager like LASH/LADISH, which I've never got to work.

ARDOUR and REAPER are the most mature and familiar traditional daws, best support for mixer controller hardware and OSC.

qtractor: best minimal DAW, qt interface can be confusing.

LMMS: acceptable for midi, but uses the retched fruity loops 7 interface, mixing sucks but good piano roll, step seq is shit.

seq24: passable, alsa only, no metronome, some quirks, needs session management.

non-daw: meh but makes sense in theory just seems cumbersome

hydrogen: it kinda sucks, but is close enough to what I want.

calf-plugins: great quality but stability issues, many crashes

guitarix: one of the better audio programs on linux

TAL Noizemaker: sadly one of the best soft synths.

zynsubaddfx: the best softsynth, but the editor sucks, there is a new version worth looking into

misc stuff I use: oxefm, obdx, TAL plugins, synth1 wine, drumkv1

soft samplers suck in general and my biggest gripe.

TLDR; if you're an electronics/beat maker you need to be advanced and do everything by hand.

>>1009566

Also there's libresprite for pixel art

FL Studio through wine or virtualbox and the same with Sony Vegas, literally only sane options

<m-muh LMMS!

>open LMMS

>samples actually midis that sound like from some fucking 1$ chink kids toy

>can't add any song to a playlist, cut it couple times and stretch it (i.e. to make vaporwave)

>looks like dogshit

<m-muh audacity!

>make your music literally blindfolded today! (can't do shit without these "set value" windows, aka in real time or something)

<m-muh Kdenlive!

>works like ass, GUI is better than most video editors on Free Software side (OpenShot, or Pitivi, god these looked like diarrhea done by some sumo wrestler after a weak of not-shitting and eating just extra-spicy mexican food)

<m-muh ffmpeg! lets be minimalist!

>Try to make montage parody or ytp in this crap, even if you somehow manage to, then there's always the horrendous rendering time

so yeah

>>1018236

If you're going to go for proprietary software, look at Reaper as well. They offer support for Linux users on most versions, historically through Wine but also more recently as native packages (though I've not tried them).

>>1018236

>FL Studio through wine or virtualbox and the same with Sony Vegas, literally only sane options

Wine is absolutely not sane especially when you work with external VSTs. Stability goes out the fucking window then. Bitwig is far better than Fruity Loops and is native on Linux. Additionally, plugins are sandboxes in case one of them does take a shit.

>>1004668 (OP)

I unironically use audacity for recording spoken word. Blender is good if I need to whip up a quick render of a product or send off for a prototype mold. Gimp has also served me well in the 2D space.

what's the best audio synthesis tutorial?

I tried synthorial but it's only windows, and dunno if I can run it with wine.

>>1021142

Thanks for reminding me about sunvox, reminds me of the good ol days of pc software.

The audio world is 99.9% sales/marketing, .1% R&D.

Even in the FOSS world evangelism has made bloated crappy software more popular.

>>1004668 (OP)

>Linux port

Desktop linux is such a clusterfuck that no business will provide commercially produced software for it.

Not only does each distro fuck up the default paths, like /tmp for temporary builds might be under ~/.package-name-here instead, but the real kicker is that the packages themselves, the actual software is so poorly maintained, designed, and implemented that there is no way for a commercial supplier to provide competent support when something wrong happens.

It's easier for large software companies to focus on the big two, Windows and Mac, and leave the rest of the OS's to fend for themselves.

This is why Linux or BSD's will never be number #1 on the desktop.

>>1013692

niggerlicious, delet this

Did avisynth/vapoursynth ever get a proper Linux port that's been promised for over a decade now? Is WINE support any good? I really enjoyed using scripts to edit video back in the day but had to leave it all behind when I gave up Windows many years ago. I've yet to find anything as powerful as avisynth for video work.

>>1021238

Why would you say this? It's not entirely true and, when it is, isn't something exclusive to Linux. Just take your file path example, for instance. You do realize that Mac and Windows both have plenty of clusterfuck software attached to them that do the same thing, right? Add that to the fact that all a company really has to worry about is targeting Snaps/Docker and I really don't see what case you really have.

>>1021260

You are not making any argument.

With Windows and Mac, most programs do not need to be compiled, so there aren't a bunch of temporary build files laying around randomly consuming disk space.

Look at config files on Windows and Mac, especially Mac, are located neatly in a familiar area.

With Linux, you don't know if it's in some _hidden_ folder or if it's in /etc or /opt or wherever.

Snaps/Docker are not targetted nor designed for desktop users.

>>1021265

10 albums of noise music are still 10 albums of trash

>>1021262

What is Linux Standard Base and freedesktop defacto standards?

>>1021238

>>1021262

There is commercial support for Linux, but it's typically catered around Ubuntu. This is a sensible approach, as it allows developers to focus on a predictable target while utilizing an open desktop's greatest attribute, which is open collaboration community support.

If you're unable to adapt the software to your system, you ought not to be tinkering in the first place. At the end of the day it's still better than wasting your time with canned answers from Pajit, because let's face it, the "big two" are AWOL or trying to be something that no one really wants.

The commercial offerings that do exist are great and these things take time. I honestly don't expect macOS to be around much longer as long as Cook is still in charge, and Microsoft is too busy trying to be Apple and spreading themselves extremely thin.

>>1004671

>stop using commercial solutions and support the open source ones.

The problem with that, is at the current stage, no open source solution can be used for actual work.

Consider Blender. Some professionals use Blender. So due to this, money comes in and work gets done to make Blender even better, and it can be used for work. If you want to get work done, a case can be made for Blender. There are people for which Blender is better than say, Maya. But there are no people for which Kdenlive is better than After Effects. Zero. Not a small minority, fucking zero. You could only make a case for Kdenlive as "you REALLY love open source, like you have such an autistic passion for it it's a fetish at this point". That's the case for Kdenlive.

As long as that's the case, money won't come in, and development won't be done.

>>1005211

>Blender is most certainly NOT good with video.

Can fucking confirm, god fucking dammit what a load of crap the video editor is. The rest of the software is awesome though.

>>it pains me there's no good PS alternative

this. my only need for WINE at this point

>>1021265

suit yourself. i've had someone jack off to a song i made in audacity. i find non-windows, non-apple more customizable and therefore can make it convenient and comfy, which is creatively great for me.

>>1021446

>What is Linux Standard Base and freedesktop defacto standards?

Not important enough for every distribution to adhere to.

>>1021457

Those commercial "options" you discuss aren't targetted to Linux as an ecosystem but only focus on one distro, Ubuntu, effectively acting like an additional competitor against Linux itself like Mac and Windows. Even Android is a competitor to Linux at this point.

inb4 >but it use duh linux kernel

Can you use an Android apk on a fresh LFS install?

Community contributions aren't that great. This is why Linux is in the mess that it is, because community contributors just do enough to get it to where they like it and expect someone else to fix everything else for them, which quite often doesn't happen until years afterwards or the project dies.

>I honestly don't expect macOS to be around much longer as long as Cook is still in charge, and Microsoft is too busy trying to be Apple and spreading themselves extremely thin.

Holding on to hope, I see.

There will still be iOS, and Windows will be around for decades to come purely for legacy support.

You should worry about whether desktop computers will still be around in the next 10 or 20 years and less on the operating system.

>>1021262

>Look at config files on Windows

I don't know about Mac but I've seen Windows config files end up in roaming/local (devs often don't understand the semantic difference), the user's home folder, the user's documents folder, the public documents folder (which is a major faux pas), in the program's own install directory, in Application Data, in some random folder in C:\ (rarer nowadays) and others. Sometimes it's hidden but you'd be silly to have hidden folders actually hidden on any OS. It's usually faster to start procmon to find config files than to read the documentation or to look through those locations yourself, unless you know where they're gonna be (e.g. I know that Flash saved object files are in the user's appdata/Macromedia/...). On the Linuxes and BSDs I use, the package manager can tell me what files it installed where with a single command. Additional, run-time configuration is rare and usually ends up in the user's folder (rarely anywhre beyond a dot-file in the root or within .config) and when it doesn't (e.g. for services), manpages are available. There's many oddities to Unix-like environments but some exceptions withstanding, configuration is very sane.

>>1021544

Not until there is an API (syscall) available that allows programmers to write and read configuration files without worrying their FS location, there won't be any sane configuration method on Linux.

>>1021142

Is that a fucking Buzz clone anon? Also, someone needs to write a *nix VST wrapper already, I mean shit, I lived long enough to see X-Screensaver ported to windows, VST to *nix shoud be any time now...

>>1021730

I thought there was news about this a while back. Not that I bother with music on linux much now. I wasted too much time compared to just dual booting and doing my shit there.

>>1021557

What are you even talking about?

>>1021537

I had already said focusing on one distribution is the most sensible thing to do. It's too commonplace for an entity to just give up before they even start because of the vast array of distributions. It's better to support one district well than to halfass support across the board. Porting hasn't been an issue for me at all (I don't use Ubuntu but I use commercial software that normally comes in a deb file). If you can't make it work for your distribution, you're better off using something like Windows/Mac anyways.

Your comment about Android was out of left field. No idea why you included that.

I'm not worried about the desktop going anywhere. People have been posting videos about the death of the desktop for a decade now, all from their video editing suites on desktops. Integrated devices with ecosystems are black boxes that enterprise won't even touch. I have seen some DJs with iDevices and have even been to one ameteur studio that used only iPads. Even still it was for capture and automation and not post production. I have yet to see a professional studio without a desktop.

>>1021750

A simple syscall in Linux that opens the config file.

Instead of opening it manually and having to put in the path, which could be anywhere, the kernel handles creating the path in a standard place and returns a file descriptor or pointer to a FILE struct.

At least it would create conformity through a central source instead of depending on the community, multiple sources, to adhere to standards and practices, which many don't.

>>1021890

>I had already said focusing on one distribution is the most sensible thing to do.

But that's not Linux. That would be writing software for Ubuntu or Mint or Gentoo or whatever.

That just ignores the issue without providing any solutions to the problem.

>Your comment about Android was out of left field.

Android is a distribution of Linux, yet you cannot use apk's on non-android distributions.

Android is now a competitor to Linux just like Windows and Mac are.

Very simple to understand.

>I have yet to see a professional studio without a desktop.

You are using anecdote as a generalization.

Desktops are already losing market to mobile devices. Look at the numbers. More people access websites and the Internet through mobile devices than desktops. That only took 10 years to happen, considering from the release of the iPhone in 2007.

Within the next 10 years, manufacturers will either reduce or eliminate their products for desktop markets while bumping up their production for mobile.

There is too much money in eliminating the desktop and locking consumers into mobile devices.

>>1021922

no professional is gonna work on a fucking tablet, and if you dock that shit to use certain input devices you might as well use a desktop that gives you more performance at the same time (without haviing to worry someone is gonna steal that shit)

i use vmware with photoshop and premier.

Fuck the haters, it just works.

>>1021925

>no professional is gonna work on a fucking tablet

They already do.

>>1013692

saved

>>1021922

You're conflating grandma's Facebook experience with the AV industry?

At this point you seem like you're in the wrong thread. You're painting broad strokes based on internet polls which do not represent the topic industry in the slightest. Consumers may be willing to lock themselves down with iDevices, but again, that just isn't happening in the professional world.

Nobody plugs a preamp into a tablet, and nobody does mastering in their living room on integrated speakers.

>>1022230

>Nobody plugs a preamp into a tablet

The tablet IS the preamp

the rest of the post is fine though

>>1022230

Professionals; e.g. doctors, lawyers, chief management officers, these people use tablets and prefer them because they allow for a small form factor "computer-like" device that accesses their records quickly and effortlessly.

The A/V industry will fall in line when they are forced in line.

Even program development will be moved over to tablets and walled gardens.

Do you think I want to see the desktop computer die? Hell no, but I am sure not going to delude myself after reading the writing on the wall.

>>1021142

man thanks again for the suggestion, I've been using it a lot over the past couple days.

This is almost exactly what I've been looking for, while it has quirks its a great paradigm

>>1022291

Checking out their github

>not available in my distro's repo

You've failed me for the last time, Arch.

>>1022811

It's an appimage, who cares?

>>1009965

Never helps, it's literally only a thing when I explicitly change the volume though, for example on a youtube video.

I just installed gentoo. My wacom tablet worked immediately. No drivers, nothing. It just werks. I'm impressed.

>>1023240

Never mind. Error. LiveDVD. Gentoo seems to take quite a long time to ACTUALLY install. Real no brainlets allowed software.

>>1013692

>People are remixing the extracted mid files from TempleOS (go to archive.org)

I was thinking of doing this also a few months back, but I enjoyed Terry's schizo pieces as they were. Good work here though.

>>1022993

I tried Radium a while back. Got it working on windows but for some reason it didn't really click with me. Now tbf I was trying out a bunch of DAWs because I got this idea that using Renoise for like 8 years had limited my musical ideas. I still think that a bit but I couldn't find another DAW that I liked even half as much.

>>1023465

You've got shit taste and also autism, m8

>>1024268

Nah, it's just 96dB of dynamics (in reality more like 90dB, because one of the bits is basically just sign, which doesn't increase amplitude), while the difference between the threshold of hearing and pain (let alone hearing damage) is at least 140dB.

>Why then, do we need dither when downsampling from 24 to 16 bit? The dither itself is described as a little bit of white noise added to the audio, wouldn't that be even worse than aliasing your volume a little?

Read about a phenomenon known as quantization noise (it's an issue precisely because 16-bit is not nearly enough to make the transition to "silence" (i.e. below the threshold of hearing) smooth enough.

>>1023500

Way to miss the point, faggot. Spamming a bunch of noise in dozens of channels doesn't make "music". It's impressive for idiots, just like complexity in code impresses them, or talking about 144 Hz HD monitor, when I'm fine with a fucking plain old CRT and 320x200. The irony is YOU'RE THE AUTIST WHO NEEDS ALL THAT COMPLICATED BOTNET BULLSHIT.

>>1024284

We've got amazing capabilities in our hands now, regarding sound synthesis. You don't have to buy analog synthesizers and circuitry anymore, if you don't have the money or don't want to. There's so much good stuff available for free now, times have never been better for someone interested in making music. With that comes complexity, but I fail to see how that's a bad thing. DAW's are easier to use than ever before and there's so much resources if you want to learn.

Reducing sound synthesis to the least common denominator isn't going to make music more well understood. If you want to reduce input for the listener to a minimum, read some orchestral sheets and tell me how that's bad and overly complex music. Everything about that is impressive - from the overall orchestration, to recurring themes, to the skill needed to play the individual instrument, to the mind creating such a piece, to the sheer mass of instruments. Yet I don't see how being impressed and moved by that makes me an idiot.

>>1024294

Yeah okay faggot. Enjoy getting fucked by the cianiggers. Some people can't be helped. Go fuck yourself.

>>1024284

Thanks for proving me right, but on a functional board your post wouldn't have survived long enough for me to read it

>>1024297

Do you produce anything or are you just here to shit up the thread? Assuming good faith, there's plenty of ways to digitally compose without resorting to FL Studio Cloud™ AND without having to go to cheaply produced chips from the 80s (or emulators thereof). While I personally happen to enjoy the sound of chip tunes and find the medium's restrictions conducive to creation, you'd have to be way up your own ass to think it's the apex of audio. For example, SuperCollider or ChucK are interesting modern takes for live performance. There's also plenty of free and Free DAWs and DAW/tracker/etc analogues like

>>1024303

Oh nevermind you're either brain damaged or trying to troll.

Look you fuckhead! >>1023500 directly insulted me, so WHY THE FUCK ARE YOU MAKING EXCUSES NOW WHILE PRETENDING TO BE ALL FUCKING SO MORALLY SUEPERIOR! YOU FUCKING SHITHEADS ARE ALSO HYPOCRITES! SHITBAGS!

IF YOU CAN"T FUCKING HANDLE AN INSULT< THEN DON"T DISH THEM OUT, FUCKEHADS!

>>1021557

I DEFINITELY don't want that in Linux. You could do it like with $PATH but what you're suggesting is simply bloated and unecessary.

>>1031818

It is not bloat if it is necessary, and it is definitely necessary given the nature of how shitty programmers are with """"""free"""""" software and adhering to de facto standards. It simplifies the creation, modification, and removal of configuration files instead of needlessly relying on the install or make script to print out where the files are along with the possibly thousands of other files placed wherever they feel in the file system.

>>1031822

>placed wherever they feel in the file system

Not really. The system wide stuff is to be found in /etc and the user specific setting in ~/.*

And sure, some applications provide examples somewhere in /usr/share but they usually say that in the manpages.

>>1031826

These are places you _expect_ to find them. Nothing requires them to be in those places however. Look at how many different directories X.org checks for configuration files before defaulting. If the kernel would supply a simple interface for configuration files, it would make everything simpler and easier to use.

>>1031822

that sounds like the windows register to me. dont want that on linux

>>1031847

It has nothing to do with Windows "register". All it is to provide a simple interface for programs to create configuration files, modify them, and remove them as needed. You can still modify the config files with whatever text editor you want. You can still browse to the folder and fuck shit up if you so desire. What it doesn't do is create additional complications for new users when moving to Linux.

>>1024268

Dithering takes place below the precision of the least significant bit. Think about it like this: without any noise, a value x<0.5 always rounds to 0, x>0.5 always rounds to 1. This is deterministic, so you lose most of the information in that range. What if you add statistical noise? A value like x=0.1 will have a high probability of rounding to 0, x=0.3 less likely, and so on. Although this low order bit has become noisey, it's not humanly noticeable and you've actually preserved more information. If you boost a quiet section of a 16 bit recording with no dithering, you will hear a buzzing quantization noise, which sounds similar to clipping. The noise masks this, by preserving more of the perceived dynamic range below the LSB.

You might be a bit confused about sampling rates. Due to nyquist theorem, a sampling rate of 48khz can represent at most 24khz. Think about sine waves; if you sample a sine wave at the same rate as it's frequency, (once per full revolution,) you see a constant value, which is heard as silence. You need to sample the sine wave using at least double its frequency in order to capture a peak and a trough. This way you'll get alternating positive and negative values.

As far as downsampling goes, removing the frequencies above the nyquist limit is exactly what we do, not just because of the limits of human hearing, but to prevent artifacts. In the 96khz -> 48khz case, we need to remove as much of the frequencies above 24khz as we can and then discard half of the samples (96/2 = 48). If we do not discard these samples prior to removing frequencies above 24khz we get aliasing noise. This is a consequence explained by nyquist theorem; when you try to represent frequencies that are greater than the rate of sampling, they will create 'aliases' that appear to be of a lower frequency. Think of a video of a tire or propeller spinning so fast that it seems to stop (reaches the video's nyquist rate), and then appears to start rotating in the opposite direction (the rotation speed exceeds the nyquist rate). In audio, high frequencies are reflected back into the audible range, so a 25khz signal would introduce a 23khz alias. If you had a lot of loud signal above 24khz, you would end up with a loud mirror image below 24khz which sounds extremely unpleasant.

But ultimately, downsampling is just about removing high frequencies so that the signal can safely be resampled at a lower rate. To remove frequencies we need to use filters, and there are many kinds of filters which all have different properties. Some are fast to compute but produce imperfections (simple IIR filters and small convolution FIR filters), others are expensive and highly accurate (large convolution FIR filters using a lanczos/sinc kernel.) There are a wide variety of problems that can be introduced (latency, phase/time-alignment problems, subtle aliasing, and so on) so it's just a question of what trade-offs you're willing to accept, hence the large number of different algorithms all suited to different use cases. If you're creating a synthesizer with 2x upsampling, and you need to downsample in real-time, using a 4096-sample FIR would introduce a delay of about 45 ms which would be unacceptable for a musician to perform with. You'd be much better off with something faster and lower latency, like a butterworth filter. On the other hand, the FIR would produce highly accurate results for offline rendering, so you could switch between the two depending on the context.

>>1031878

So if those frequencies always mirror like a sinewave does that apply to the data before during storage as bits on a recording medium, or during playback as it passes through the physical hardware to the speakers?

Where are those high frequencies being introduced is what I am trying to get at. Is it at the recording phase, during transportation from physical to digtial via cosmic rays/other interference, or during playback from digital to physical via audio hardware/drivers/software?

If those high frequencies are only being introduced by changing the bitrate of the stream then why fuck with the stream at all and not just leave the data on the disk at the original bitrate and get better audio hardware for the sake of pleasent hearing?

Or why record at say 24 bits if all audio hardware supports 16 bit so that nothing has to fuck with the bitrate?

doesn't this mean you could hide messages in the whitenoise of the audio that gets left from imperfect sampling then? Like morse code via 1's and 0's in the white noise or something worse then that?

>>1031945

You always want to record and mix with the best quality possible and then sample down the end product so the source material is richer, has more of everything. Just like you wouldn't record to MP3. Also frequencies below or above human hearing can often still be felt.

Producers often refer to 128khz as more "tape like and analog feeling". Artifacts reach an acceptable level when they're inside the nyquist condition (twice the frequency) but when you increase the sample rate even more, there will be even less artifacts.

Ok let me rephrase the question. Let's say I as a audio producer want to produce and send the highest quality audio undithered and unchanged to a persons speakers. Should I on the software side record at the bitrate the speakers will output at? Say my recording equipment/software supports 16/24/32 but the output speakers the listener will use only supports 16. Is it better for the listeners experience to record at 16 and send it that way or to fuck with dithering/sampling/container formats in software and record at 32 bits or whathave you?

By the way is there anything that will phsyically affect the audio during transportation and storage that will demand it be dithered like cosmic rays or defects in the silicon of recording/playback hardware or some such shit?

>>1031945

>Is it at the recording phase, during transportation from physical to digtial via cosmic rays/other interference, or during playback from digital to physical via audio hardware/drivers/software?

All of the above, minus the cosmic rays. Audio recorded from an amp or microphone or any analog input has no theoretical limit to how high the frequencies can go, because it is a continuous signal. Of course natural sound sources tend to become quieter the higher the frequency range you look at. Dogs and other animals can of course hear things well above 20khz, so it's those kind of things that you're protecting against by filtering an analog signal before sampling it to get a discrete signal.

>If those high frequencies are only being introduced by changing the bitrate of the stream then why fuck with the stream at all and not just leave the data on the disk at the original bitrate and get better audio hardware for the sake of pleasent hearing?

idk, depends what you mean. If you're saying everyone should buy $200 sound cards to playback a massive 100MB 96khz wav file, then I guess you're right. In reality I'm pretty sure the most pragmatic thing is what we're doing already, which is to downsample to something reasonable like 44.1khz or 48hkz and distribute that. It's smaller, easier to process and playback, and it's not like anyone was going to hear those higher frequencies anyway. The sound engineers couldn't have even heard them.

By the time you've got that file, the majority of the high frequencies have been removed. The only remaining reason for resampling is to put the audio in a form that can be used by your sound card. Some files are recorded at 22khz,32khz,48khz but you need to get it in a form that the hardware will accept. The average sound card only supports a fixed set of sample rates and bit depths.

>doesn't this mean you could hide messages in the whitenoise of the audio that gets left from imperfect sampling then? Like morse code via 1's and 0's in the white noise or something worse then that?

You might be able to do something like this, but I'm not sure how you'd get the morse code back out. You could do it by subtracting a higher bit-depth copy of the source audio to recover the noise, but then it would be pretty trivial, basically a one time pad. Can't think of any practical and/or devious applications for that one.

>>1031963

Short answer is that it's not really important.

The difference between 24 and 32 bit is practically indistinguishable, and is basically lossless for all a human listener could care about. Even 16 bit is pretty good, and you are only going to hear some buzzing if you turn up the volume REALLY LOUD on a section that is REALLY QUIET. (to the point where the signal is grazing those least significant bits.) No one will ever be able to hear the effects of dithering listening to a piece of music at a sane volume.

Best bet would be to distribute 24-bit dithered. If you knew for certain your listener had a sound card that only outputted 16-bit, it would technically be better to send them audio dithered for 16-bit playback, but negligibly. Dithering in itself is entirely optional, and kind of autistic once you get beyond 16 bit. More importantly, if somebody has a 16-bit only sound card and they listen to a 24-bit sound file, there going to lose virtually nothing, and it's possible that the drivers might do some dithering in the conversion anyway. On the other hand, going from 16-bit to 24-bit means that you're not taking advantage of a full 8 bits of precision. Using 24-bit is the best option for both cases, and it preserves the most information.

And dithering definitely won't protect against cosmic rays and all that, if that happened (pretty rare I think), the damage would be a lot more noticeable.

>>1031969

And just to make things make a little more sense: dithering is done by engineers for archival purposes, it's not for enhancing the listening experience. Basically, it's only valuable if you need to re-master the mix without re-doing the entire thing. It just theoretically reduces the decay in quality over multiple iterations of remastering archived material.

>>1005112

>I'd like something like a MIDI file but hand-writable, with more features and some conveniences like macros or reusable sequences.

Lilypond suits this type of musical autism perfectly. http://lilypond.org/

You can programmatically generate both sheet music and MIDI files using a LaTeX-ish syntax.

>>1031963

I suggest you use the highest profile since the equipment has the most power over the quality than the software "stated" 24/16/32. Who knows what kind of workarounds they do behind the hardware or the software they provide? Just use the highest one and start from there.

Think of trying to put huge cheese inside a small container. If you tried to eye trimming such cheese before putting it in the container there will be gaps but if you just shove the container into that cheese you'll have less gaps. Why use a cheap 16bit sound recorder in the first place? There will be more room for error margins caused by such weak equipment.

That also happens in video/audio.

Most android phones sound shit and only produce up to 16bit 41kHz but if you listened to my cheese example some of the bits will be 'dropped' because the hardware is too weak to handle that not to mention other audio streams happening inside an android phone just steals more of the quality room. However if you have a 32/24bit 192kHz soundcard phones like Sony, LG, or Samsung you will enjoy much better audio quality just like the cheese example. The hardware can also handle lots of audio processing and have lower latency.

Try playing back a flac on a 16-bit 41khz cheap noname soundcard on a noname android phone and you'll hear the dead ass difference after trying out Samsung or LG flagships.

Recording at 16 will just ruin the quality of the master. You should record higher regardless of "muh dog hearing" and believe it or not MP3 or at best AAC is still the king for lossy/downsampling and it is well-tested that you won't have 'gaps' from continuous gapless playback while other shitty coding formats with their respective decoders do this insane "rounding off milliseconds" not to mention how these shits like OPUS actually fuck up the bpm around the end of the song when played side-by-side with the master,AAC, and MP3 format at the same time.

>>1032034

>Who knows what kind of workarounds they do behind the hardware or the software they provide?

Well if I am using hardware I have the schematics to and software that is open source/I can audit, I do.

>believe it or not MP3 or at best AAC is still the king for lossy/downsampling

Holy shit I just went and tested it and you are right. Starting a song in libopus-1.3 made and ran music vs the original the beat got off by like a tenth of a second towards the end of the song. It wouldn't have ever been noticeable to me if I hadn't started the original and the opus one at the same time, which I usually listen to opus because of. I still think opus is better in every other metric besides that one point though for its compresion/size vs quality ratio.

>>1032057

I updated to the latest ffmpeg and audio playback software available for me, even going out of the way to build it by source, and it didn't fix it. REEEEE THIS IS GOING TO BOTHER THE FUCK OUT OF ME NOW THAT I KNOW OF IT. FUCK YOU AND FUCK OPUS.

>>1032068

Opus is a meme popularized by cuckchannel.

>>1032072

For the purpose I am using it for its not a meme, that is to save space on low space/memory devices. But it bothers the ever living fuck out of me the beat is off now and I am thinking about just saying fuck it and only listening to originals and screw mobile music with high quality audiocard/headphones.

**also opus for internet traveling sound is better if its created on the fly i.e chat/voice clients/servers.

>>1032075

Using Opus for non-music-oriented Intenet radio, audiobooks and podcasts is also a great application. TL;DR; Opus is okay for non-music purposes.

>>1032392

VS

>>1032072

name one better lossy audio format

>>1032392

Not sure if it has support for C++ but I use mono-develop for my c# programming and that has a pretty nice ui

>kdenlive

This is good but the catch is it's basically windows movie maker in terms of functionality.

It works, it's simple, and it gets the job done. However, for anything that requires more heavy editing it's absolutely terrible.

I like to use blender's vse as i find that there is a lot more you can do with it then then kdenlive.

>>1032392

>VScode

>Electron

>500 MB with nothing open

nice meme

VST support on GNU/Linux is shit so I use hardware to use with Ardour insted.

>>1005112

Have you tried Renoise yet?

>>1010794

yesss, i clicked this thread and hoped for this

>>1032072

you mean cuckchan audiophiles and discord

>>1032068

Why use opus when vorbis can have around similar quality and file size? Even opus has less volume than vorbis which has less volume/dynamics than flac raws.

Also opus on video is kinda fucked up and have this crip-rip sound caused by that off bpm-fucked compression.

But if you don't care about the loss of volume dynamics and that bpm shifts, opus is good for you. "you won't even hear the difference" they said

>>1004668 (OP)

try

-Wine

-Mixxx

milky tracker for LibreChiptunes

>>1009491

I wouldn't say it's that good, although it is good. The effects & transitions often don't work, which limits its use for me.