My adaptation of a dummy battery 3D print model from thingiverse - https://www.thingiverse.com/thing:6467582 ; I'm replacing the six AAA batteries of my label printer with a DC socket for 9V.

One bummer though: The printer only has volatile memory for storing settings, so once I disconnect and reconnect the power, it goes back to factory (language, font size etc.)...

I want to see for low volume recordings, using a relative inexpensive USB audio interface (Roland Rubix44) makes a big difference with regard to microphone preamp self-noise in comparison with a relative expensive USB audio interface (RME Babyface Pro FS). I test both with an expensive microphone (MKH 8040) and an inexpensive one (OM1). Since I plan to use the Rubix44 with the OM1 in a piece, I'm interested mostly in the difference for the OM1, which can be heard towards the end as an A/B background noise comparison (gains were adjusted with a reference signal to have exactly the same signal level).

1:31 - preamp A with MKH 8040

Correction: weight of Rubix44 is 2 kg (still a lot heavier than the Babyface).

I have a small mic "booth" in my studio comprised of three EQ Acoustics L10 tiles, and a carpet below. I want to know if adding a fourth absorber to close the top or "roof" makes a substantial difference with regard to perceived reverberation on the recordings.

0:00 with open top

To make any judgement, you'll have to wear headphones. I found the difference not as pronounced on the recording as perceived when standing in the "booth". Towards the end, there are pitched down versions which of course emphasise the reverb durations.

Comparison of a few phantom-powered and a few plug-in powered microphones, with respect to quiet sounds and self-noise. See also https://diode.zone/w/2KzpGD6D1Yd1goJqZTdLTw

1:23 - Røde Lavalier II

I did not use any compression or equalisation (the metallic sounds are sometimes clipped by limiter which I used to boost the overall level in the demo). The relative gain boosts are indicated, which are used to match the microphones in overall loudness. The whispering parts are boosted by an additional 6 dB. This is the test where you can see how the Sennheiser stands out. Of the others, I would say the Røde is the noisiest, followed by the Edutige and the DPA, with the Line Audio and Audio Technica performing relatively good.

I didn't use a pop filter, and I had to cut out a few fragments of DC wind pops (you'll notice when the spoken words seem strangely cut).

Comparison of a few phantom-powered and a few plug-in powered microphones. See also https://diode.zone/w/xoQVzmnzd4ZyocA3ekN2Gt

1:27 - Sennheiser MKH 8040

Note that the MKH 8040 and AT 825 are Cardoids, all others are Omnis. I did not use any compression or equalisation, and as I am speeking only ca. 10 cm away from the microphones, the MKH 8040 and AT 825 exhibit a proximity effect (boosted bass / attenuated treble), which makes them difficult to compare.

Overall, I was slightly disappointed by the exaggerated brightness of the ETM-001 (might also be slightly worse in terms of distortion), but then again it is the cheapest of them. The OM1 appears quite neutral and pleasant. I want to add another test that compares the background noise, and yet another test that runs them on a cheap preamp, as well as running the plug-in power microphones on a very cheap preamp. The video indicates the relative gain boosts I applied to each microphone to make them appear roughly equal in loudness.

The recordings were done in a small studio with little acoustic treatment, but using some absorbers around the microphones. A pop filter had been placed between microphone and speaker.

Quick demonstration of https://codeberg.org/sciss/dotterweide-exp-metals - using the Metals Scala language server with the Dotterweide mini-IDE / editor. New Debian 12 screen recorder seems to have bad colours and no mouse pointer, sorry about that.

Building the wasm (WebAssembly) target of SuperCollider synthesis server (scsynth) in July 2023. This required a change in the wasm branch (https://github.com/Sciss/supercollider) to change git:// submodules to https://

Just me trying out the CNC drill toolhead of the Snapmaker 2 machine, cutting some boring holes and rectangles into an acrylic disk and making a monologue on the thing.

Unless mouse is constantly moved, the actually perceived frame rate is low and uneven.

Here is why it's a bad idea to use proportional width for arabic digits for UI / system fonts.

Shown: Firefox audio elements, OpenJDK application with native look-and-feel, VLC