Get Info: #script

Controlling local web servers using xbar

by Matt Sephton — Mon, 12 Jan 2026 20:57:00 +0000

I’ve released a native macOS app that does all this script can do and much more!
Read all about it: gingerbeardman.com/apps/localmost/

Sometimes I want to run local web servers for projects I’m working on. Usually more than one at a time, or at least over a short space of time.

So I thought it would be cool to have a controller for those local servers in my menu bar. Sounded like the perfect job for a little scripting and xbar, which is a great way to prove a menubar app idea quickly.

How it works

The plugin allows you to:

toggle servers on and off
open in browser
view ports
view paths
view log sizes
clear logs
edit config
and more!

Example config

In the config file we set the starting SERVER_PORT, followed by one or more SERVER_DIR for as many projects as you might want servers. We can temporarily comment out those server lines to prevent projects from appearing in the menu.

# .xbar_httpd_config
SERVER_PORT=8000
#SERVER_DIR=~/Projects/starchasers/
SERVER_DIR=~/Projects/serenity/
SERVER_DIR=~/Projects/point-cloud/

Source code

Python source code is available in the following gist:

Automating the cleaning of macOS-specific files on Eject

by Matt Sephton — Sat, 08 Feb 2025 20:45:00 +0000

I’ve released a native macOS app that does all this script can do and much more!
Read all about it: gingerbeardman.com/apps/driveaway/

Dot underscore ._ and .DS_Store files are macOS-specific metadata cruft generated for foreign filesystems (like FAT32 or exFAT) that are not usually needed for disks that are mainly used on other platforms. Digital cameras, music players, e-book readers, and handheld gaming devices can get confused when they encounter these odd files during file system parsing and directory listing. The problem is compounded if the devices naïvely process files by looking only at the file extension as they will then see the dot underscore version of a file as a duplicate and try to preview/play/open it.

For years I’ve used an app called CleanMyDrive to remove such files, but it was discontinued in October 2023. I continued to use it until it recently stopped working completely …so I needed to find an alternative solution. There are some apps on the Mac App Store that look like they’ll do the trick, but I don’t really want to spend the time buying and trialling multiple apps to find one that fits my usage habits. I can make one!

I already use an app called xbar for keeping track of my GitHub issues, itch.io sales, network/server status, and more. So I decided to flex my shell script muscles and put together an xbar script to do it.

The script adds a menu bar item that allows you to:

Eject (click)
Unmount (option-click)
Eject All (without cleaning, useful when you want to disconnect all drives from your computer)
All with a handy notification

That’s it! Straight to the point, no frills, functional software.

Automatically classifying the content of sound files using ML

by Matt Sephton — Sun, 13 Aug 2023 17:01:00 +0000

Following on from yesterday’s extraction of old sound effects, I quickly realised I needed an easier way to search them as they came out of Director as unlabelled, numbered files. I can use QuickLook or a media player to quickly audition them, but how could I easily find the sample that contains the sound of running water or a horse trotting?

I wondered if there was a way of using Machine Learning (ML) to automatically categorise sounds. It seemed like something that should be possible, especially given the recent explosion in “AI” (really: ML) tools. I quickly found Google’s AudioSet, which sounded like the perfect dataset:

AudioSet consists of an expanding ontology of 632 audio event classes and a collection of 2,084,320 human-labeled 10-second sound clips drawn from YouTube videos. The ontology is specified as a hierarchical graph of event categories, covering a wide range of human and animal sounds, musical instruments and genres, and common everyday environmental sounds.

But the data set is only one half of the solution. You need to use the dataset to create a model and then run that model against your own data to get the required results. Thankfully, I found YAMNet:

YAMNet is a deep net that predicts ~521 audio event classes from the AudioSet-YouTube corpus it was trained on.

I guess YAMNet is tracking behind AudioSet in terms of total categories, but it is good enough for me. Here is a list of all the classes of sounds it can recognise.

Let’s go

I used the script described in this tutorial as a starting point. I’m not a regular python user but using pip to install tensorflow, along with any other missing imports, and after that it …just worked.

Getting your files in order

According to the documentation all sound files need to be at a sample rate of 16000Hz. After getting some calssification results of “Silence”, I realised they also need to be 16-bit resolution. So I ran a quick sox command to create compliant copies of all my sounds. I’ll delete these when I’m done. Notice how I decided to trim sounds to a maximum length of 3 seconds. This helps speed things up and most sounds can still be recognised with such a short starting section.

find . -iname "*.wav" -exec sox {} -c 1 -r 16000 -b 16 {}_16k.wav trim 0 00:03 \;

Optimisation

Running the classifier works at about real-time, a few seconds per sound, but I noticed that it was leaving a lot of my CPU unused. This struck me as a prime candidate for parallelisation, which is pretty easy on the command line. I used the parallel command to scale up the classification to use all 10-cores of the M1 Pro CPU in my 2021 MacBook Pro.

find . -iname "*.wav" -exec parallel python3 classify.py {} ::: {} \+

As I type my computer is making short order of the task, whilst remaining perfectly responsive, if a little warm. Final speed for me is one sound every ~0.85 seconds.

Python Script

Creating my SFX Library

Soundly is a sort of iTunes for sound effects. It’s an app that enables easy, automatic organisation of files, quick searching of metadata, painless playback/auditioning, non-destructive edits, and simple exporting of the final sounds. The free version allows a local library of 10,000 files which is more than enough for my usage. I’m not affiliated with them in any way, but they offer a free version and a 1-month free trial of their paid version.

As you add your local folder of files it allows you to import a (semicolon-separated) .csv file containing additional metadata. It’s here that I point it to the file that was generated by the classifier. The categories are imported as the description of the sound, and are able to be searched. Perfect!

Extracting sounds from Macromedia Director files

by Matt Sephton — Sat, 12 Aug 2023 19:42:00 +0000

For my latest “quick” Playdate project—a remaster of a ~1997 web game by Thoru Yamamoto—I decided to add sound effects. In order to keep it as authentic as possible I decided to use only sound effects created by Thoru Yamamoto that were used in his other productions.

The largest collection of sounds I could think of were those included in his Macromedia Director web experiments which include everything from short animations, through games and interactive toys, to abstract slideshows. The problem is, these Director files are tricky to deal with some 25 years later.

Halt and Catch Fire

A bit of reading and some help from Misty De Méo led me to ProjectorRays which can convert a protected .dcr file into an editable .dir file, and also allows saving of all the individual chunks that comprise each file. Think of it as one chunk for each piece of graphics, sound, and so on.

Once you’ve converted the .dcr to .dir you can open the file in, say, Macromedia Director 2004 and use CastRipperTool to export sounds and graphics and more. But it’s a very manual process and the whole setup is prone to crashing. Crucially, it won’t open some older Director movies so this wasn’t a good enough solution for me.

Chunks

I looked to the chunks that had been dumped by ProjectorRays and some quick experimentation showed that the ‘snd ’ chunks contained raw PCM audio date with a bespoke header. Loading these into something like ocenaudio wave editor was proof enough, but I would need to figure out some of the header to see if I could get the correct sample rate of each file.

With a bit of help bouncing ideas off hikari_no_yume in a bug report on one of their tools I figured out the location of the sample rate. Interestingly, the values were unexpected and not the usual values 11025, 22050, etc. It turns out that classic Apple Macintosh used some slightly different sample rate values: 11127, 22254, etc.

Using the xxd tool I dumped the relevant section of the headers of 1163 sounds from 105 .dcr files. The breakdown in sample rates found was as follows:

Rate (Hz)	Hex	Quantity Found
7,418	0x1cfa	1
11,127	0x2b77	616
22,050	0x5622	31
22,254	0x56ee	515

The lack of any unexpected sample rate values was further proof that this was the real deal.

Shell script

Armed with this information I decided to write a short shell script that would do the following:

extract the sample rate from the ‘snd ’ chunk
create a trimmed raw pcm file excluding the 78-byte header
use sox to add a new WAV header using the correct sample rate

This script allows me to batch process all the audio and that’s much more to my liking: it captures everything compared to CastRipperTool and it’s a lot quicker.

So, dump your chunks using ProjectorRays and then call my script like this:

find . -iname "*.bin" -exec ./bin2wav.sh {} \;

Notes:

I’m assuming 8-bit, mono, unsigned PCM data in the ‘snd ’ chunks and have not found anything else in them.
SoX (Sound eXchange, the Swiss Army knife of audio manipulation) requires the .raw extension for raw PCM audio data, it will refuse to process the files without it.

Post-processing Animated GIFs

by Matt Sephton — Thu, 16 Jun 2016 00:00:00 +0000

This article was originally posted on pico8 BBS in 2016, and on Playdate devforum in 2020.

I was interested to see how easy/difficult it is to edit the animated GIF recordings. My goal was to trim some frames from the beginning and end to make a more succinct video.

After much trial, error and experimentation here are my findings using Mac OS X.

The game shown in the GIFs is Worm Nom Nom by ilkkke & kometbomb.

Overview

Animated GIFs can be:

viewed frame-by-frame using system Preview.app
manipulated using “gifsicle”, “imagemagick”, “graphicsmagick” command line tools
converted using “ffmpeg” command line tool

Notes

useful tools need to be installed, you may wish to use the “brew” command line tool or similar.

Viewing

Open the GIF in Preview.app and it will show you all frames.
Preview calls the first frame 1 (one), but other tools usually call it 0 (zero).

original/source animated GIF:

Un-Optimising

Warning: If you have previously optimised your GIF to reduce filesize, as in this other thread, and you want to edit the GIF further then be sure to first use gifsicle -U to unoptimise it, otherwise the editing commands won’t work as expected.

Trimming

required: gifsicle

then this is how you trim

gifsicle anim.gif "#212-238" > trimmed.gif

note: this makes a copy of the GIF and keep frames 213 to 239 (gifsicle uses zero based frame count)

Resizing

if you want to double size of the image:

gifsicle --scale 2 trimmed.gif > resized.gif

Captioning

if you want to add an overlay to caption the animation

required: imagemagick or graphicsmagick

then use this bash script:

#!/bin/bash
source=$1
caption=$2
: ${1?"Usage: $0 anim.gif overlay.gif [output.gif]"}
: ${2?"Usage: $0 anim.gif overlay.gif [output.gif]"}

fnsource=$(basename "$source" .gif)
fncaption=$(basename "$caption" .gif)

output=${3:-"$fnsource-$fncaption.gif"}

gifsicle -E $source
for f in *.gif.*; do composite $caption $f $f; done
gifsicle --loopcount *.gif.* > $output
rm *.gif.*

how to run the command

./caption.sh anim.gif overlay.gif [output.gif]

note: if you do not specify an output name, it will be named using original filenames, eg. anim-overlay.gif

overlay.gif

Converting to Video for YouTube

required: ffmpeg

to convert the GIF to MP4:

ffmpeg -i trimmed.gif -movflags faststart -pix_fmt yuv420p video.mp4

video uploaded to YouTube: