deskthority

Introduction

This is a comparison of typing sounds produced by three of my full-size Alps-switch keyboards that all sound good to me. Before I had tested them side-by-side, I thought they all sounded more or less alike. After a more detailed comparison, I realized that the sounds are clearly different. I would be interested in your opinions on which one sounds best to you and your reasons why.

Materials and Methods

I made recordings using a Samson Meteor mic suspended 13.75 inches above the "H" key on each keyboard using a boom and radial shock mount not connected to the desk. I used a Windows computer with Audacity software for converting to MP3 and producing waveforms and spectra. Recordings were made in stereo mode and collapsed to mono for waveforms, spectra, and MP3 files.

The three keyboards were as follows:

1. Leading Edge DC-3014. Switches: stock Alps SKCM Blue. Keycaps: Stock doubleshot ABS. Modifications: internal Orihalcon-Soarer converter and a flush-mount micro-USB connector.
2. Northgate 101 US ANSI. Switches: stock Alps SKCM White (Pine). Keycaps: stock doubleshot. Modifications: internal Orihalcon-Soarer converter, flush-mount micro-USB connector, top-case painted black, spacebar switch modified with Matias Quiet Click slider and tactile leaf, spacebar shock absorbers replaced with new self-adhesive ultra-thin foam, bottom case lined with 1/8" art foam, self-adhesive ultra-thin foam applied to contact points of PCB with bottom case.
3. Zenith 163-73-1. Switches: Alps SKCL Green hybridized with top switch housings and click leaves from Alps SKCM White (Pine). Other modifications: internal Orihalcon-Soarer converter and flush-mount micro-USB connector.
Typing consisted of typing 3 times each the following sequences:

F6 F7 F8
7 8 9
Y U I
H J K
N M ,
Backspace
Backslash
Return
R Shift
R GUI
Spacebar

Results (Scroll to bottom of page for links to videos with audio)

Leading Edge DC-3014, Alps SKCM Blue (blue waveforms) Amplitude (arbitrary linear units) vs time (sec)
Amplitude (dB) vs time (sec)

Northgate 101, Alps SKCM White (Pine) (red waveforms) Amplitude (arbitrary linear units) vs time (sec)
Amplitude (dB) vs time (sec)

Zenith 163-73-1, Alps SKCL Green / Alps SKCM White Hybrid (green waveforms)
Amplitude (arbitrary linear unis) vs time (sec)
Amplitude (dB) vs time (sec)

Amplitude (dB) vs Frequency (Hz) for all 3 keyboards:
Videos with audio recordings
Leading Edge DC-3014 Alps SKCM Blue
Northgate 101 Alps SKCM White (Pine)
Zenith 163-73-1 Alps SKCL Green/Alps SKCM White (Pine) Hybrid

Conclusions

This is still an assessment in progress for me. The waveforms for the three keyboards looked very similar to me. The spectra showed definite differences, but these were not straightforward to interpret (for me, at least). The audio playbacks were the most revealing.

Based on the audio, I would rate the sound of the three boards as follows:

LE DC-3014 > Zenith 163-73-1 > Northgate 101

My preference for appearance:

Northgate 101 >> LE DC-3014 >> Zenith 163-73-1

My perception of build quality and solidity:

Northgate 101 >> LE DC-3014 ~ Zenith 163-73-1

What are your impressions of these three boards?

All three of them sound great but the LE3014 stands out the most. Those blue alps are deeep.

Great. Regarding the spectra; it would actually be more informative to plot a spectrogram (which is basically a magnitude spectrum over time), to see how the frequency content varies over time. Just a thought (unfortunately it seems like this is not available in the version of Audacity that I have; strange - because it's a common tool to use).

the_marsbar wrote: Great. Regarding the spectra; it would actually be more informative to plot a spectrogram (which is basically a magnitude spectrum over time), to see how the frequency content varies over time. Just a thought (unfortunately it seems like this is not available in the version of Audacity that I have; strange - because it's a common tool to use).

Thanks for the tip! I will look into this, and when I can find some time again, I will see if I can generate some spectrograms. It might also be good to produce some difference spectra in order to detect the frequencies that differ in intensity between keyboards.

EDIT:

@the_marsbar: My version of Audacity has the spectrogram feature. Do you have suggestions for the various settings? Currently, I have used settings that enable seeing fairly distinct frequency bands as well as resolving the separate keystrokes -- essentially a balance between the time and frequency domains. If I then keep the settings the same for the 3 keyboards, the most apparent difference among them is the color of the bands: the LE and NG bands are blue and the Zenith bands are red. It is my understanding that this color change reflects differences in loudness; if this is the case, then it might be more informative to adjust the gain and range dB settings for each keyboard to enable the frequency bands to be readily seen, although this would mask differences in loudness. Advice welcome!