Making Nintendo LABO Waveform Cards Using SNES and Vocal Samples
HTML-код
- Опубликовано: 26 ноя 2024
- That last video really took off, so here's a more in-depth followup!
You can download SVGs and STLs for all 18 of my waveform cards here: www.hunterirvin...
You can download a free mp3 of "I Like You" just for checking out my patreon page: / hunterirving
I don't tweet a lot, but I'm on twitter if you want to follow me there: / thehunterirving
Thanks for watching, and stay tuned for more projects!
As a vocal synthesizer enthusiast, you have no idea how happy seeing the switch do primative vocal synth made me.
as a vocal synthesizer developer
no one has any idea
because all the dickheads at kvr like to make money and be famous instead of help people
but yeah holmes, formants are gonna scale with pitch, you want them to stay fixed.
teh tingting is, your "physical waveform card" lends itself well to being modulated, eg. make a card with two layers that slide, then modulate the shape during performance, and you'll get diphthongs (eg. "ow", "oi" "u")
if you like DIY, try making a latex larynx you can squish around, then use a buzzer/speaker for the voice box.
@@atomictraveller ah!
Well, I have researched into formant synthesis and some basic psola and lpc based synths. I also have a niavie implimentation of NPSS and Dennis Klatt's synth running. I havn't a complete understanding, since I am self teaching, but it is rather exciting getting to mess with with all the technologies. OH! And I also sorta figured out how hts/hmm works, i oatched the hts_engine so I can feed it f0 as I don't have enough data for accurate prediction.
Ah I just wanna mess with them all
@@Digits0801 my most advanced implementation uses 7 or 8 formant bands (BP filters) in teh classic method, thus extending past 12k or so, this can sound relatively authentic in best cases.
iirc in dafx 18? (17? 19?) someone published an antialiasing method (related to a polygonal oscillator) that only handled discontinuities (eg. teh jump in sawtooth) and would give say 30dB drop of aliasing for cheap first order processing, which means you can use whatever method you want to "draw" a waveform and get decently aa results... mystran published a scheme ("antialiased saw - pulse oscillator" or something) on kvr maybe 2014, i've (xoxos) published a slightly diffrent code, dsp forum.. but these are awesome for any kind of psola similar method.
i'm sure ai methods will predominate in teh future but "i wish things had been different," eg. in some of my last posts there i talk about cepstral processing, how simple it is to do, when you know how, and how long everyone obscured it so they could sell inferior products. the cepstrum allows you to derive the spectral contour of the formant without being locked to time data.. convolve with signal, its formantised. (or simply reverse the cepstral decomposition, replacing the source component with whatever you want).
reason is.. if you use phase vocoder (phase correction took me some years to do right) then it would be trivial (eg. the type of dev i do on a weekend/evenings working another job) to build an app that lets a user record their voice, then define their own formant set, then export it and share it with the world for free, as well as use it instead of paying balls for stock voices that don't formant shift or anything.
instead of all the melodyne or whatever (i don't keep track) super expensive exclusive products that work for shit, everyone in the world could exchange their own vocal sets for free, could have twenty, 25 years ago.
using pvoc would skip a bunch of possibly fruitful experimentation but, quality results.
tbh an analysis app could probably churn out coefficients for a source-filter model just as well.
i didn't go much past trig in education, spent years of my life reading ccrma pages.. truthfully, all bitterness aside, the industry leaders (profiteers) are so landed in academia that they lack the pioneer spirit thatdiscovers interesting processes, because they have to be so "legit". they will sell you dumb synths that use serious power churning out minbleps or whatever but just make the same boring oscillations.
my vst "black" and "angle" for example, use two different methods of extending oscillation to more dimensions, doing very simple things to the information, then generating amazing results that NI and u-he will never market because they aren't predictable or "academic sounding".
here's an early audio file from one of them that soundemote ran thru his then-in-development "prettyscope" so you can see the 3d component (overlapping with visual/"oscilloscope" synthesis)
ruclips.net/video/kzIA45Qyh_c/видео.html
i didn't dig up mystran's method (teh difficulties) but you can glean my "transition point antialiasing" from this page if its useful, you gotta scale it by the transition. digging aorund for more on polygonal oscilators fromd afx ought to return multiorder forms that are steeper. prolly find his easy if you want to read dumb posts all day.
www.kvraudio.com/forum/viewtopic.php?f=33&t=522701
@@atomictraveller atomictraveller oh wow! That's lot to analyze, i'll have to look into it all over time
But it all sounds interesting.
I understand your points on expensive software, melodyne can easily be replaced with the world vocoder and a neat ui (or now the llsm2 vocoder)
The math part of it is a little above me, but slowely I have been understanding it. The spectral domain is still a bit iffy to me, you may be interested in the flexibility that SynthV has for it's voices. The system is able to create glottal effects like growel by modifying the source track (it runs off of llsm2 and some machine learning)
As for vocal distrobutions, my interesting in singing synthesizers is due to communities like that. There are hundreds of people that produce their own voices and even write songs. Some of them preatty popular in the community. Although they all rely on concatinative methods.
These being Utau(has multiple backends), DeepVocal, and NiaoNiao
(there are probably others)
(you could technically use Festival + Mbrola but the singimg capabilities are limited currently)
Plogues Alter/Ego is actually a HQ formant synthesizer, although vocal development is left to only 1 company now due to... an earlier conflict that shut development down
EmVoice is a samoler that runs on melodyne, personally i don't like it, it's to flat and to expensive to me, limited genre range
I still to like me squeaky and crunchy voices like earliy loop based vocals or retro voices like in Plogues ChipSpeech
Thank you for shareing all that with me!
@@atomictraveller thank you so much!
I'll have to note this all down
I never really thought much about compilers themselves, I usually just use the gnu compilers.
Windows alway feels too much. I really only have windows so I can game. I use linux for everything else.
It never occured to me the 32bit to 64bit transition would have been an issue, but again, I was in middle school when xp stopped support, so I couldn't have understood that.
Did you just make your own vocaloid? Out of cardboard and plastic?
LOL! Kinda!!
Durdge-o
That's fair
And a switch.
More like an UTAU instead of a Vocaloid ;)
Darkoli1987 / MikuHatsune i'd actually argue more like Garage Band Macne Nana, Utau's synth is /slightly/ more sofisticated than that. (Especially when using moresampler)
Using those vowel sounds to mimic the human voice was wicked cool
Joseph Rasbold using wicked like that... you must be a new englander
Jeef Squaa haha I live in Boston currently but I'm originally from Pennsylvania! Too much New England must be rubbing off on me :P
this is really neato
WHAAA HOW COME YOU ARE LITERALLY EVERYWHERE I GO!!!!!!
Cyranek hi dude
Oh hey my dude how's it going
Cyranek
Why not surprised your here
I’m just disappointed
hey
Oh my god you can do vibrato by moving the switch that’s awesome
It's not really vibrato, it's glitching out, but it's a useful glitch :P
Richard Smith No, it’s a feature, talked about even in the labo piano overview. How could it even be a glitch?
It happens because when shaking, the ultra-red camera precieves the waveform slightly differently. It's awesome how people still find new ways to use a cardboard piano...
@@mechamind9569 No, that's not true. The Waveform is scanned at the beginning, then not scanned again until removed. It is a feature built in to the game, even outlined in the trailer.
@@mechamind9569 Thank you. Obviously Nintendo manipulated physics though, as they would never have accidents.
your leaking how a high quality rip is made
>2 months later
bruh
h
@@KatzRool h
what the fuck why are you here i love you
I NEED to listen "Still Alive" from Portal. PLEASEEE
Armando Berrios DIY
oh, I didn't thought on that! I'm flying today to a country that sells nintendo Labos and 3D printers
Armando Berrios don't worry buddy, i got ya covered ruclips.net/video/Y6ljFaKRTrI/видео.html
@@TheLazyEyebrow YOU FUCKING MADMAN
@@An74res while I understand your point, see as you have internet access and international shipping is a thing. It's possible
The speech synthesis just needed a noise layer for the "S" sounds, and it would have been like a really nice Vocaloid for the Labo. XD
Hatsune Labo
Tip: use c700 vst for the snes instruments instead of getting a soundfont. It's more accurate, easy to use and can rip samples from literally any snes game (so you're not limited to all the ones the community ripped for you). It also shows a waveform in the plugin itself, though I'm not sure how useful it would be due to how the way it previews the waveform of the sample.
Another option that would not only eliminate the usage of a DAW, be more accurate and flexible and can preview the sample without any problem is a program called SPC to IT, this program converts snes songs to the .IT tracker format, which can be opened in a tracker program like OpenMPT. The waveform of the samples ripped can be easily found in the sample section and can thus can be screenshotted for use. This not only eliminates the need for something like FL Studio but it also removes the need for Soundcard Oscilloscope.
Great tips, thanks!
Seth everman with hair shows us how to 3D print cards to make different sounds on a cardboard piano
Your voice reminds me of the English translator for sakurai
I thought the same thing! especially when he said "please take a look". 😂
Diabloid Yeah! Especially when he says “oh, and one more thing,”
Oh my god it's exact
Like 200
I felt like I was watching a Nintendo Direct
I don't have labo or a 3D printer to mess around with this myself, but this is incredible! Please make more of these videos if you can!
Thanks so much!! Just a note if you do end up getting Labo sometime down the line - You don't explicitly need a 3D printer for this. I have files on my website for use with traditional desktop printers as well that you can print and cut out of cardstock :-)
Good to know! Will check them out if I ever get labo!
cool ! i was wondering it it would work without being 3d-prited =D thanks ! this piano is the only reason why i want to get nintendo labo
my laserjet prints come out a bit smaller than a real slide. doesent sound right either, sounds like the saw wave >_>
I like how this video is gaining traction again after so many years
Your vocal samples were basically just animal crossing voices!
also sorta like vocaloid
skeleguy john
They were from vocaloid
You do know what they're speaking right? When they're talking ther saying the letters of the word real fast
I like you
You like me
Let's-a-go
The next smash hit on the billboards!
I like you
You like me
Letsa go
Play the wii
I like you
You like me
Let's a go!
Play the Wii
Wait a minute
I gotta pee
my new ringtone
@@EnderTails ok good
now play with me
That Smash Ultimate mockup made me happy inside
ChaseDaMLGLord Gaming
It's funny how recognizable the song is, and it's been out for less than a month.
i mean TECHNICALLY the melody was originally revealed in the nintendo direct ( ruclips.net/video/3tfjcsMvN8M/видео.html ruclips.net/video/NsFna-Z1MjU/видео.html 33:45 ) about 3 months ago.
Huh, i can hear it faintly in the background of the smash reveal...
Your profile picture makes me happy
Chiz
Dude, so glad youtube is recommending people I don't know to me again. Because for every 10 channels I'm not interested in, I find 1 that's absolutely amazing.
Thanks for watching! :)
I can’t believe this showed up randomly on my feed 5 years after it was uploaded but I’m glad it did xD
I thought the shout out to your best friends was really sweet and it was my favorite part of the video.
The song at the end kind of sounds like Miku's singing it.
Dragonite905 Hi Dragonite! This- is - Buizel! :)
Huh... It kind of does
It's cause Miku is a Synthetic Singer, she's not a human, but a sound bank put together to make sounds similar to speach, which is easier in Japanese as the language is easier to mimic authenticly with synthetic tones, as apose to English which comes of more Robotic, and dead pan like Siri
I'm sure you already knew this to some extent though Miku is kinda popular even here in the West for a Japanese Icon.
Proto Propski wow, and all the while I thought that Japanese is very hard on the pronunciation. If I made even a single mispronounceiation, the Japanese would simply slap me.
Lol, yeah that's native speakers for you, Grammar Natzi's lol
Im surprised that you aren't running with this idea and releasing renditions of other popular songs using Nintendo LABO. I'd love to hear how E1M1 from Doom would sound.
Seriously tho, i imagine theres an audience out there who'd love to hear LABO song covers.
This is one of the best videos I've seen on RUclips. So many videos are filled with garbage to reach 10 minutes, creation is fueled by greed for the most part. This video, however, is packed to the brim with different and interesting topics. It's very admirable for it to have been made this way - a greedy content creator might take material equivalent to this video and split it into 6 different 10 minute long videos.
Keep doing what you're doing, watching is a real pleasure. I'm glad to know there are still good creators on this platform.
PokeMC,
Thanks for writing out this comment.
One of my main goals for this channel is to avoid putting out formulaic, low-effort content - I want to make videos that I myself would find interesting.
Obviously this is a lot more work, but it means a lot to hear that that work is appreciated :-)
Thanks again!
There are absolute tons of them, you just gotta know where to look
Smells like ignorance
>Starts with earthbound
*Oh cool*
>cant do stuff wit earthbound sounds
*gets disappointed
I'm not in EarthBound fan but i feel bad for all the Earthbound fans why would they release Earthbound Beginnings but not release Earthbound 3 I'm confused
@@bland9876 are you referring to mother 3?
@@pequod6660 yes mother three. I should have said they only localized mother 1 but why not three?
@@bland9876 they localized they first two. Im playing rn the third and im seeing stuff that werent a lot appropriate for western people. But thats just an hypotesis.
I can be happy with the "DEEEEEEEEEEE" alone lol
It's a wavetable synth. Artists have been using Wavetables since the end 80s. Tangerine Dream, Allan Parsons Project, Aphex Twin, Skrillex, they all have used them
日本語で失礼します。
Nintendo LABOのピアノでこんなこともできるなんてすごいですね。
どうもありがとうございました！
おけすか？
うん、ニンテンドーラボは本当に素晴らしいですね。
Чего?
Hunter Irving NANI??
That intro was very good man, I really appreciate you recalling what most people thought.
This video was just recommended to me and I clicked it, but you sold me the video in the first place. I usually don't comment, but wanted to let you know that the intro alone impressed me and that I look forward on watching the rest.
Ridiculously cool!! The people behind LABO must be so happy to see people like you using it in such creative ways. Really inspiring work!
4:54 did he just... ssbu us?
5:39 Yes
What’s that for a song?
1:54 que the people saying "that's from undertale!"
Undertale sampled a ton of Earthbound sounds
@@littlesnowflakepunk855 Toby Fox being the creator of the Halloween Hack for Earthbound... I think he composed on what he already knew how to use.
AMALGAM MOMENT
3:56 Now Nintendo should hire you because you just played mumbos mountain from banjo kazooie.
Thanks for that. Another useful video with lots of handy info!
Thanks for watching! Good to see some regulars start popping up in the comments :-)
Always really interesting to see people get experimental with Labo. I'm surprised how accurate some of these turned out!
It would have been good to hear a side-by-side comparison of the original sounds and what they sound like in the labo.
The part I enjoyed the most of this video! (5:27 - 5:37 / "The Song of Time" from OOT)
I KNOW RIGHT
I though it was from mm
"The Song of Time" is in both Ocarina of Time and Majora's Mask!
I love POT and flipped my bricks when I saw this song be played on cardboard
team up with that guy making high quality labo piano covers named geniway
I would love to collab with him!
I _immediately_ recognized the Past sample when you played it, and I'm so glad to know its origin now.
Love the smash and Luigi's mansion themes!!
Nice video man. You obviously put alot of work into it and followed up with the attention you got from the previous video. You got my sub.
I see SMRPG, I click.
I see cute boy, I stay.
Combining Nintendo and music nerdship, I actually pay attention.
Hotel Trivago
I think what the fans want is you playing old school music with this, it could just be me tho.
that human speech part reminded me of Nier Automata
The music you made sounds like it belongs in crash bandicoot
Hunter: *starts playing Earthbound "The Past" sample
Me: Hol' up one hot second there that's not what I think it i--
Hunter: 1:50
Oh sh- undertale?
Dude! This was soo good! I'm very interested in these designs and the theroy behind this. I hope you keep making videos like this!
Also, I love the pocket operators on your wall!
1:30 hearing this literally shook me
SOUNDS LIKE A [[BIG SHOT]] TO !!
When I get a switch I'm definitely doing this kinda thing. I'm a musician, and I like technology and video games. If I had a switch right now I'd use this for a project in my comm-tech class. This is a good way to make some really funky beats. Actually, I have to go do schoolwork right now but everything is online. I'm gonna work on comm-tech today because of this. Awesome.
oh hey this popped up in my feed again. I'm at college now studying mechanical engineering lol. Also I have a switch. Might have to mess around with this.
1:55
smells like amalgamates
2:16
smells like barriers
Undertale I presume?
There is a VST called ' C700' for FL where you can drag in any SPC (SNES Music File) and it will emulate the SNES sound chip and let you use the instruments. It's more accurate than many fan made soundfonts.
They sometimes require adjusting the settings, for things like vibrato, fading out or fading in.
So... Earthbound's "The Past" song uses the same instrument as Toby Fox's "Amalgam?"
The other way around, but yes.
Toby Fox's first game was actually an earthbound fangame!!
I like the character you've played here.
BTW
N.Labo was made for this kind of project that surpass the original functions.
You've made it brightly.
You can be proud of it. Thanks for sharing.
Thank you for watching!
Hey! I made it into the video! *grinning ear to ear*
Besides that, thanks for making such a good follow-up video. I really enjoyed it and wouldn't mind learning more.
From my experience editing SoundFonts, you might want to stick with a loop which can match the ends so you don't get the "buzzy", "poppy", or "sawtooth" sounds from your cards. If your loop can't match the beginning of the wave, to the end of the wave, audio distortion as described earlier will occur. It's a great idea though and I can't wait to see what more people do with it.
That's right, where's all of that? I suspect this is Nintendo BS , everything sounds different to what he intends
I know this is an old video, but do you know if the soundbytes from those websites also work with any other DAW? (Pro Tools, Ableton, etc.)
Good question. I haven't used Pro Tools or Ableton, but .SF2 soundfonts are a pretty universal format, so I wouldn't be surprised if they support it!
of course it works
Dude that was awesome. Would love to see you make a song using several of those wave patterns and using the record feature then playing the full recording at the end so we can listen to it if at all possible.
Just starting playing Ocarina of Time yesterday and when you started playing the song of time it got my attention
4:15 - You missed a golden opportunity to record this with the 'switch' click sound in place of the snapping fingers of the original!
(great video btw)
Sounds like probably what splatoon characters sound like
4:09 *Addams family theme plays*
Me: *snap snap*
This sounds like a good way to replicate the FM synthesised waveforms the Sega genesis made.
bro has EXCELLENT music taste. Cocoa Cave is one of my favorites from Kirby Superstar Ultra.
THE ADDAMS FAMILY THEME! We have a disk of some of the T.V. show.
...1:58 I was recognizing some samples and then hearing Here We Are made me go YES!
Also, Earthbound samples Monty python?! wow, that makes me more interested in playing it. which I will. eventually.
I'm thinking in a year or two, there's going to be some *crazy* stuff to found online done by this little thing.
You have a very likable demeanor, if you want I think you can really take this RUclips thing a long way. Whatever other endeavours you choose to make, good luck with them :)
This is pretty amazing. Thank you for your hard work.
I find it really interesting how you're able to shake the switch around to give it a kind of vibrato
Homestar:
"Loading screen, it's the loading screen,
I can't believe this cartoon is just all the loading screens."
Oh my good*ness*, you have to make more of these wave forms. Would love to hear a SM64 sounfont on Labo! Keeep it up. I'm saving up for a Switch, so yeah, this made me jealous.
I just realized 4:10 is the addams family opening theme XD this person is a man of culture (Geez... That show is OLD)
Benjamin Knapp yep
“Hey I’m interested. What instrument do you play?”
“Nintendo Switch + Cardboard.”
That final song was super cute!
"NiNtEnDo LaBo Is FoR kIdS1!!1!" - this video is really awesome
5:12 *Kaito kicks in*
I saw the OG vid years ago why am I only just now seeing this one??
I’m sorry that I don’t know this, but what is the song at 6:29? Also, AMAZING JOB!
It's Lovely Angel from Katamari Damacy! Also, THANK YOU!
Hunter Irving Thanks! I knew it had to be from something good. Keep up this great content.
a liitle side question could you not 3d print the cardboard parts so its more dense? that would look really cool i think
I guess adding some attack might help create some consonants. That being said I know nothing about the labo; just spit balling
Insane, actually. The speech part blew me away
I knew that I heard that sound before (the past ) it's also in Undertale :D that makes it even creepier
More than anything I'm impressed at how skilled you are at playing the Labo piano
That e3 reference killed me so hard ooof
It's a shame LABO never really caught on. I know it got made fun of a lot with comments like "lol Nintendo is charging us a bunch of money for cardboard" but it really was a cool and unique concept. I think they tried to hit a middle ground between the "maker" and "gamer" audiences and ended up not making either group happy.
I remember building the fishing rod and having a lot of fun doing it, but after like 2-3 hours of folding cardboard the actual fishing game was so barebones that it took me around 20 minutes to see everything it had to offer. Same with the bike racing game and the virtual pet house thing: the builds were fun but the "games" were incredibly simple and felt like afterthoughts that didn't justify the time and effort to build the kits. Imagine if the fishing game had some kind of actual progression where you could fish in different bodies of water and collect rare fish, like a fishing version of Pokemon Snap; or if the bike game had a career mode where you could compete in tournaments against NPCs and modify your in-game bike to change its attributes; or if the pet house was an actual virtual pet that could grow and change into different forms depending on how you care for it day after day, like a Tamagotchi. It also seems like a huge missed opportunity that they never made a Wario Ware LABO game where you build a few simple kits and then play micro-games with them.
It feels like Nintendo believed in the idea just enough to R&D all the kits but not enough to put actual game design muscle behind it, and I think that's what tanked it before it ever got off the ground. It seems like they tried to do that a little more with the vehicle and VR boxes but it still wasn't enough, and by then it was already too late to turn things around. Anyway, sorry for the long ramble on a 5 year old video, this just reminded me how much potential LABO had and how Nintendo failed to capitalize on it.
I love how he shakes it to make a wavy sound
The other video was cool but this is some high quality production!
Thanks so much! I put a lot of work into it.
6:01 kinda sounds like a Super Mario World soundfont
instantly subscribed when he played the song of time and as im subscribing i hear the freaking smash bros theme... and luigi's mansion
*BRUH WHY DOESNT THIS HAVE MORE VIEWS??*
*AND WHY DONT YOU HAVE MORE SUBS???*
It was not the previous video that made me subscribe. It was this one. ^^ Very nice stuff!
Thanks for the sub! I know it's been a while since I posted a new video, but I've got TONS of ideas - stay tuned!
And so came a new genre of music known as
*labowave*
So you basically remade the Voder.
Or a version of it, at least. Sweet!
Yes! I knew this was possible! Thank you.
The end Sounds like some kind of victory menu or game theme love it
I have like 15 of that cat. 1 for every new roll of filament or new printer.
something neat about vowel sounds is that vowel frequencies or "formants" are just different harmonics that are present in normal speech in specific vowels. obviously its gonna be mildly different from person to person, accent to accent, and region to region, with each different vowel sound, including ones outside of the regular aeiou, having a different group of frequencies, but its relatively the same frequencies for each individual vowel sound, and you can add vowel and vocal resonance to pretty much *ANY* sound you wanted to. technically, there would be a particular group of different frequencies present in the R sound too, but i still have yet to discover what those R. (badum ts)
Gotta sample the synths from Aquatic Ambience in DKC
by the way, i forgot Who it was (i wanna say Katy Perry) but there was a big promotional stunt for Nintendo Labo, where [insert Talk Show; mightve been Jimmy Fallon] had [insert Musical Guest] on for a song; but the Band played Labo Instruments, just to say "yeah, you can technically put on a Concert with this software"
6:15 dear God you've created an analog speech synthesizer
i listen to vocaloid and even i would struggle to call that human speech. its moreso the power of suggestion. if you listen and never look at the subtitles you wont hear it at all. much like how people think they hear words in midis if they think they should
Thank you Hunter very cool!
Hey hunter on 1:58 that's from undrertale and that part is in a song call amlgam