#spectrogram

[ourladyofomega]

Spectrograms found inside Aphex Twin, Plaid, Venetian Snares, and Nine Inch Nails albums.

[graphicnoises]

Wheezy/whistly laughter is so cool looking on spectrogram.

[ch4ng3l1ng]

bird songs

[kreuzaderny]

Self-supervised Anomaly Detection for Narrowband SETI

 The Search for Extra-terrestrial Intelligence (SETI) aims to find technological signals of extra-solar origin. Radio frequency SETI is characterized by large unlabeled datasets and complex interference environment. The infinite possibilities of potential signal types require generalizable signal processing techniques with little human supervision. We present a generative model of self-supervised deep learning that can be used for anomaly detection and spatial filtering. We develop and evaluate our approach on spectrograms containing narrowband signals collected by Breakthrough Listen at the Green Bank telescope. The proposed approach is not meant to replace current narrowband searches but to demonstrate the potential to generalize to other signal types.

[personwithatophat]

Genloss Logo

chatting about the logos because there are so many

these are the main 5 types of logos that are shown and labled on twt + the logo for the genloss website (fun fact: the website logo was drawn with a brush made OF the logo) #1 and #4 being the most used currently with #4 being the pfp for the Genloss twt and the Genloss YT channel but #1 being well associated with TSE through TSE promo and media you can find the #1 logo if you dig a bit and if you want to know about hidden logos we got hidden logos here's the spectrograms

Track 1: Generation Loss title Card (long)

Track 2: Showfall media

Track 10: Hetch

Close up for track 10 ^

Track 16: Generation Loss title card (short)

Track 24: Live or Die

Track 25: The holding Portal

track 25 close up ^ All of these examples are from the official soundtrack, and is also part of the reason I don't make spectrograms for the Youtube videos since most of the time the logo is there due to it being part of the music. Huge thank you to Nathan Hanover who made the soundtrack for being so active on the genloss discord, for throwing bricks at the genlosers, answering our questions at 2am, and for confirming that he sampled the among us stinger for Track 23: The button.

yup anyway im posting a part two since there are so many more hidden logos

-Tophat

[klaufir]

selected spectrograms from my forthcoming album:

https://klaufir.bandcamp.com/album/miniature-ecosystem

[bombfuse]

the kris audio

[psychicmyst-blog-blog]

Here are five entities that came through an experiment via spectral waves. The head of the woman on the right is not on her body - perhaps indicating that she was decapitated.

[lavenderskimask]

8 DECOMPOSED FILES

by salem woods / Selma Soul

april 2025

[superlittlecannibal]

spectrogram

[latetothegames]

[wormnoodless]

okay i went and put the audio from the new Generation loss video into a spectrogram and while you can definitely see a bit of symbols I don't know what any of it means. Putting this out there for people to see if they can figure it out. Please ignore the windows watermark in the corner, I have no money.

[babybluesquid]

Some spectrogram fun! I like singing "She'll Be Coming Round the Mountain" in a particular way, one with glottal stops in the word "mountain" so it's pronounced like so: [mɑ͡ʊʔn̩]. I really accentuate that glottal stop because I love singing with emphatic stop consonants. It just satisfies me.

The spectrogram is gorgeous. Here's the last part of the song, "she'll be coming round the mountain, she'll be coming round the mountain, she'll be coming round the mountain when she comes." All those really wide gaps preceded by increased amplitude? Glottal stop!

Since the phrase repeats three times it's really easy to see what's going on. The first repetition has different frequencies than the second and third, in which I sang the same notes, which is really obvious on the waveform even though the waveform depicts the complex wave!

[sssspectrogram]

carter scholz, kaleidophon (2000)

[genuflectx]

Guessing Whether a Song is AI or Human (guess for a 🍪)

I think I've found the general (albeit, unlikely to ever be 100% accurate) way to guess whether a song is AI or human via its spectrogram- something you can look at easily for free with Audacity.

I made an image with AI on the left and human on the right and think it makes a handy guide to compare other spectrograms to. Comes with normal, reversed, and a black and white (contrast slightly increased) version. Feel free to steal it and use.

AI songs tend to have more purple/orange at the top and less bright yellow at the bottom, i.e. "muddled" looking. Out of 10 AI and 10 human sampled songs of varied genres this pattern holds up relatively well.

Let's see if we can get this to hold up by looking at unlabeled spectrograms and placing them either as AI or human using these images. I tested this on two friends and 1 got it right and 1 only missed 1 >:3 If you get them all right you get cookies ;3 I'll put the answers in the replies after a while. Click read more to play :D

hint: look at the bottoms of the spectrograms, they tend to provide more accuracy than the tops. Compare them to the provide charts.