Keyboard Sound Check: consistent audio of multiple popular boards

[XMIT]

https://www.youtube.com/watch?v=5aMaetd ... XQQTqJjPtK

I'm hard at work on Round 2 prototyping for my Hall boards. I'll cross post about that later, or you can check here for details: https://geekhack.org/index.php?topic=90287.

All of these videos were recorded with a Nikon D600, using the built in microphone at a fixed gain of 16, about 2.5 feet from the keyboard. I recorded at 720p60 since that's the highest resolution mode offering 60fps on this camera.

This collection is really special because, since I have all these boards in one place, you can do a really good apples-to-apples comparison of the sound, recorded on the same equipment at the same levels! I'll have more to say about that clicky Round 2 board later. I'll work on adding more boards later.

The playlist is the best way to view them all - they will play in succession. Or, you can use these quick links for individual videos!

CM QuickFire with Cherry MX Blue switches

KBparadise V80 with Alps Blue switches

XMIT Round 2 with clicky Hall effect switches

IBM Model M with buckling spring switches

IBM Model F with capacitive buckling spring switches

IBM Displaywriter with beam spring switches

WASD v2 with Cherry MX Brown switches

hasu alps64 with Alps Cream Damped switches

Happy Hacking Keyboard with damped Topre switches

XMIT Round 1 with linear Hall effect switches

Topre Realforce with stock Topre switches

ITT Terminal with Micro Switch Hall Effect switches

Fujitsu terminal with clicky Fujitsu leaf spring switches

Macbook Pro Retina with rubber dome scissor switches

[Cattus_D]

Thanks! This is very useful.

[gjooro]

Whoa, those IBMs sound almost like steel drums

[seebart]

Very good work XMIT! I remember several people uploaded various recordings but a side-by-side list like this is what we needed thanks. Your Fujitsu terminal keyboard sounds much nicer than mine!

[zslane]

These videos demonstrate a very important truth: more than half of the noise a switch makes is from bottoming out and upstroke clack. The sound of the click leaf in nearly all those clicky switches is drowned out by the bottom out and upstroke noise, rendering the "click" element virtually pointless.

These videos also demonstrate just how awful the IBM Model M/F buckling spring switches sound. They make the most teeth-grindingly cacophanous din of any keyboard I've ever used. I honestly don't know what anyone sees in those wretched things.

The only boards I would be able to tolerate would be the Topre-based boards and the dampened Alps because they have far less bottom-out/upstroke noise, and mostly what you're left hearing is the switch mechanism itself, which is how it ought to be. I'm really looking forward to someday hearing a demo of the Hall Effect switches with some decent dampening built into them.

[DerpyDash_xAD]

These videos demonstrate a very important truth: more than half of the noise a switch makes is from bottoming out and upstroke clack. The sound of the click leaf in nearly all those clicky switches is drowned out by the bottom out and upstroke noise, rendering the "click" element virtually pointless.

These videos also demonstrate just how awful the IBM Model M/F buckling spring switches sound. They make the most teeth-grindingly cacophanous din of any keyboard I've ever used. I honestly don't know what anyone sees in those wretched things.

The only boards I would be able to tolerate would be the Topre-based boards and the dampened Alps because they have far less bottom-out/upstroke noise, and mostly what you're left hearing is the switch mechanism itself, which is how it ought to be. I'm really looking forward to someday hearing a demo of the Hall Effect switches with some decent dampening built into them.

zs bro the hall effect is pretty damn awful too
they are all cray terribad though even the fairly quiet ones

[Engicoder]

How about throwing another variable in the mix I believe the type of keycap also affects the sound.

[Daniel Beardsmore]

These videos also demonstrate just how awful the IBM Model M/F buckling spring switches sound. They make the most teeth-grindingly cacophanous din of any keyboard I've ever used.

The recordings sound like there's excessive reverberation in the environment, and the frequency response seems imbalanced: there's a cold, hollow, metallic quality to the sound that seems to vary between recordings. The Realforce and MX Blue recording both suffer from this quite noticeably — my Realforce had a much softer tone.

There's no "correct" recording, as even if the equipment is perfect, everyone's environment is different. If you want a clear idea of the sound of any keyboard type, you need to compare multiple recordings and attempt to mentally isolate the common characteristics of the sound.

Interestingly, modern Unicomp keyboards cut out a lot of that metallic tone and sound rather more subdued, or at least they did when I bought mine a few years back. I've come to prefer the sound of the IBM-era models though.

[XMIT]

So, this is a solved problem: the best way to get an authoritative recording is to use a recording studio with sound isolation and a professionally calibrated microphone, and to automate the testing process.

Everything is a variable. The material and shape of the case, key caps, PCB, and plate all play a role. Even the type of feet and the nature of the surface under the board play a role.

My goal here was to record everything I had quickly on hand in a homogeneous environment. I'm not a sound engineer (though I've known some).

Switch sound leads to a highly personalized reaction. Some folks love the sound of IBM buckling springs or the new click leaves in the XMIT boards. Some don't.

But, if you care for my thoughts: I rather like the sound of the XMIT click leaf, and I also rather like the Fujitsu board's sound.

[zool]

But, if you care for my thoughts: I rather like the sound of the XMIT click leaf, and I also rather like the Fujitsu board's sound.

I rather like the XMIT click too

[Engicoder]

Your right, there are many variables, but I like where you are going with this. For popular boards, having a sound recording under the same conditions, even if they aren't ideal, is worthwhile for comparison. Thanks for taking the time to make the recordings.

[zslane]

Yes, of course the environment makes a difference. So too does putting a soft material under a keyboard to absorb the unholy racket it makes.

However, notice that Topre boards and dampened Alps boards don't need any special acoustic treatment (like a soft pad underneath) to reduce the noise. That's because they make very little extraneous noise to begin with, irrespective of the environment. MX silent switches are similarly quiet and inoffensive.

[evoman]

I don't agree with zslane - the Model F sound nice to me. I use a Model M and there is a roughness to the switch sound that bothers me and it seems to be less apparent in the F. The Beamspring sounds especially nice to me. I hate linear switches because they just sound wrong to me, but for the tactile or clicky set tested, the blues sound the worst to me. There is a thinness that annoys me!

I am still looking for a keyboard that sounds exactly like I want, but it may not exist (I was hoping one of these would hit it - the Beamspring is close).

[Ander]

Sorry to post to an older thread, but I just saw this tonight.

Thanks for all the samples, José! Your XMIT boards sound great. (I'm typing on one now.) I can hear the added clicks in v.2, but the switches themselves have such a nice thock, it doesn't seem as necessary to enhance them as it would most other boards.

v.1 sounds a bit more open and resonant to me... Could it be that v.2's clickers absorb some of the energy of the presses, and/or dampen return sounds?

I've always loved that clear, musical ping of a Model F's, too. I can see how some people would find it annoying, but it's pure fun in my book.

BTW, José, why the rubber gloves? Were you filming a medical drama too?

[Chyros]

Btw, relevant: 100+ keyboards typing under identical setup: