Hacker News new | ask | show | jobs
by nlh 548 days ago
Congrats on launching!

Limiting this to work only with folks who have MIDI keyboards attached to their machines probably cuts your audience down by 99.9999%. Especially here on hacker news.

I think you’re going to find it tough going to get any real feedback with that requirement. You might find better luck launching in a community of folks who are more likely to meet your hardware requirements.

7 comments

I disagree. I believe it is a good idea to focus in your "customer" and ignore the rest.

This is the best advice I was given. People are always asking for you to unfocus and disperse.

Most people that is serious about learning chords have a MIDI device. It will make your life way harder to add non MIDI devices and the people that will use it are not really that committed. MIDI devices are so cheap today that anyone that wants one could get one.

BTW I have several MIDI devices because I create my own games too.

If you're smart, you'll put an affiliate link to a MIDI keyboard on Amazon
Or find a virtual midi keyboard. Technically we already have a keyboard, right?
Sounds like an interesting idea but honestly I have no idea how many people will use the game.
I agree with you functionality wise, but a video showing it in use would be a good idea so those of us away from our midi devices can at least see it in action.
"people that is serious"
Just got my MPK61 hooked up this week… I’ll have to give it a try.

The set of people that have MIDI capable keyboards and want to improve musically is vanishingly small compared to those without, true, but it’s clearly a lucrative market if you do it right: there are tons of (music) keyboard classes out there online.

And a lot of the hardware and software are pricey enough that I imagine ads could be pretty lucrative too.

I went to look that keyboard up, and incredibly there's no picture of it on Akai's product page. I mean... WTF? https://www.akaipro.com/mpk61
Thanks for you feedback. It is really valuable for me. Actually, my main idea when I started this project was helping to remember the chords and common progressions on a real keyboard. Like making a muscle memory, so personally prefer to keep it for MIDI devices, but helping users there to find a virtual MIDI device could be also a good idea.
I have a MIDI keyboard and an interest in learning, but it's not hooked up to my computer most of the time.

But if this is strictly for learning fingering and not for identifying chords by ear, I guess it's reasonable to require a keyboard. Otherwise you can't use it anyway.

Related question: Some people are suggesting keyboards that are only MIDI controllers; but that doesn't seem sufficient, because in that instance where will the sounds come from? Are there default sounds that will be triggered if you hook up a controller-only keyboard?

Agree - this is something I'd be interested in trying out, but I don't have a MIDI keyboard.
If you really want to try it, you can use a virtual MIDI keyboard. I made one (BloopKeys for iPhone) but there are others around.
They're a minor investment now a days with some being under $30. It's just buttons at the end of the day
Can you provide a link(s)?

A super-brief spin through Amazon doesn't show anything like that. And <= $30 is within the 'impulse buy' range, so I'm very interested to know more, if you're up for sharing :)

Akai LPK 25 is like $50 brand new, so is LPD 8 if you're more interested in pads (MPK Mini gives you both for $100 new). There's cheaper options out there, but that's the cheapest I know from a well-established brand.

Even if you have no interest in music, get one. MIDI is super easy to read, you can DIY something like an Elgato Stream Deck for a third of the price, physical volume knobs included. It's up there with a Raspberry Pi for the coolest cheap gadget.

You can even get halfway decent touch sensitive ones on marketplace for sub-$200 if you’re not fussy about it being a full 88 or weighted keys.
okay, but.. any path to connecting this to phone mics so that it can function anywhere? whats the state of the art in tone and chord recognition models?
No need for any ML models here, you can "just" do an FFT on a conventional signal and pick out the peaks
The waveform produced by sounding a note on most physical instruments will often not exhibit a peak, or will not exhibit the strongest peak, on the note being sounded. Rather, most instruments will instead produce harmonic overtones and our brains fill in the gap of the pitch that’s intended to be sounded.

You can still absolutely deduce the fundamental with great accuracy via an FFT, but the approach is a bit more involved. The relevant research area here is called ‘fundamental frequency estimation’.

For an example of this, you can see this app I built that lets you give keyboard and mouse inputs via playing notes on a bass guitar, which are recognized over the microphone: https://github.com/codyd51/offkeyboard

(i mean yes i agree but) so ... why do i feel like ive never seen anyone do this before?
If you have a guitar handy, you can try out the live mode on https://www.fretboardfly.com . It works and people use it: Using your acoustic guitar to play the fretboard learning game via the mic.
I believe this is how Yousician works.