this post was submitted on 07 Oct 2026
38 points (100.0% liked)

videos

23364 readers
128 users here now

Breadtube if it didn't suck.

Post videos you genuinely enjoy and want to share, duh. Celebrate the diversity of interests shared by chapochatters by posting a deep dive into Venetian kelp farming, I dunno. Also media criticism, bite-sized versions of left-wing theory, all the stuff you expected. But I am curious about that kelp farming thing now that you mentioned it.

Low effort / spam videos might be removed, especially weeb content.

Content Warnings must be placed before the rest of the title.

There is a cytube that you can paste videos into and watch with whoever happens to be around. It's open submission unless there's something important to commandeer it with at the time.

A weekly watch party happens every Saturday (Sunday down under), with video nominations Saturday-Monday, voting Monday-Thursday. See the pin for whatever stage it's currently in.

founded 6 years ago
MODERATORS
 

Nebula Original Videos

I have uploaded a selection of Nebula Original Videos to Tankietube.

Short Film

The Dinner Plan

When he’s pressured into attending a dinner party hosted by his boss, a man who suffers from a rare food allergy must navigate the most stressful night of his life with one primary objective: just don’t throw up in front of everyone.

Indetiteaze

Two strangers wake up in a mysterious white room with no way out or memory of their past. The two must unravel their identity while navigating a corporate metaverse that controls virtually every aspect of its employees’ online reality. A story about finding yourself.

Dracula's Ex-Girlfriend

Two former rivals reconnect over drinks and a quick bite. One is compulsively running through men and can’t kick the habit of drinking their blood, and the other is working through her trauma and settling down with a new career in monster counseling. Things get a little uncomfortable when the conversation turns to the same toxic lover that connected them — Count Dracula — one of whom is still entangled with him.

Documentaries

Boomers

Old people run the world, with octogenarians in power and wealth disproportionately flowing to the elderly. Life opportunities once available to today’s seniors—affordable housing, cheap education, livable wages—are slipping out of reach for younger generations. Boomers, the debut feature-length documentary from Tom Nicholas, follows a trail across New York, Washington DC, Florida’s The Villages, Paris, and London to uncover the rise of the gerontocracy and the questionable legacy of the Baby Boomers. In a time of intense intergenerational tension, Tom investigates how a generation that seemingly had it all has left a complicated mess for those who come after.

17 PagesChoose your starter pokemon: (just for fun, don't overthink this)

Watch Watch

Revisiting one of the largest scientific controversies of the 20th century; the dispute that began with an ethics question over a single biology paper caught the attention of the national media and US congress, and had lasting repercussions on how scientific misconduct cases are investigated even 40 years later.

Series

⭐ Abolish Everything.

On Abolish Everything!, comedian-abolitionists roast a pet peeve they want banned from society, while a panel of improvisers dubbed “The Political Establishment” argue on behalf of the status quo. The audience chooses their champion each episode, each of whom will come back to compete in the season finale.

Amy’s Dead-End Dreamhouse

Welcome to ‘Amy’s Dead-End Dreamhouse,’ where Amy and her household appliances navigate adult issues like mental illness, job hunting, and impossible beauty standards with help from some very special guests.

Essay Channels

Theater

The Prince

Written by and starring Abigail Thorn, creator of Philosophy Tube, The Prince is a transformative new play that has everything: sword fighting, lesbianism, Hamlet, and disappointed parents.

you are viewing a single comment's thread
view the rest of the comments
[–] Kajika@lemmy.ml 1 points 15 hours ago* (last edited 15 hours ago) (2 children)

The first 2 links are just claims without substance. the third one is more interested and shows how, even in 2026, this is far from being ready.

  • The paper is cleverly not trying to embed information in pixels but in a latent space representing frames (sequence of frame to be precise, hence the R 𝑇𝑧 ×𝐶𝑧 ×𝐻𝑧 ×𝑊𝑧 -> time/channel,height,width) that's why it is "good" against pixel degradation from encoding. One issue is that model needs to be trained on the video itself (hence the "Prior" in the title) so every single video upload would trigger a model training and each training needs to be good (don't think that training sessions always end well) and models saved independently (basically saving each model and related code with every video).
  • An other issue is that the payload they are encoding is only 8-bit and the final discussion talk about needed to work on longer payloads for deployment. In the paper it is stated that test with 12 bits gave 94% accuracy already. How many accounts Nebula have? Also it means that the identification needs to be quite stable and scalable. Usually companies like using UUID but we are talking about 128 bits so we are very far here.
  • This kind of embedding alters the video quiet a bit, not that it is perceptible to our eyes but any comparing/mixing 2 separated marked video would destroy the system. Which might not make you feel safe but that mean in a court this is a one shot : if they use such a system once, people will notice and with the magic of having 2 accounts will protect themselves.

The cost of implementation is definitely not worth it. Nice paper though, the fact that this is state of the art of 2026 makes me reassure that no deep-learning based method is used now apart from GAFAM who can have more secret and advanced processes under their sleeves.

[–] cypherpunks@lemmy.ml 1 points 12 hours ago* (last edited 12 hours ago)

The cost of implementation is definitely not worth it. Nice paper though, the fact that this is state of the art of 2026 makes me reassure that no deep-learning based method is used now apart from GAFAM who can have more secret and advanced processes under their sleeves.

The fact the field is still advancing does not mean it is not mature; this stuff has been in use for decades. Yes, that 2026 paper i linked is about new techniques and not what is deployed today, but (as the other links explain) what is deployed today is quite capable.

This paper Dynamic Traitor Tracing was published in 1999, and includes descriptions of what was deployed even then in the broadcast setting (which imo is much more impressive than doing it on the internet). See also Dynamic Traitor Tracing Schemes, Revisited (2018) and Survey on Decentralized Fingerprinting Solutions: Copyright Protection through Piracy Tracing (2020) for links to lots more.

[–] cypherpunks@lemmy.ml 1 points 13 hours ago

The first 2 links are just claims without substance.

Huh? The first two links are describing the systems that are actually deployed today. Facebook admittedly doesn't say how they're doing it but they say that they storing >64bits in a way that survives not only transcoding but also editing; there are a lot of things Facebook says where I'd assume they're liars but I doubt they're bluffing in this case.

The second link helpfully explains how even encoding a single bit per segment (a segment might be a few seconds of video) is plenty to allow embeding a large unique ID over time. This approach obviously won't survive in short excerpts or other editing, but can easily survive scaling and transcoding and it is plenty sufficient for identifying the source of complete episodes:

Instead of generating a completely unique video file for every viewer, the system prepares two slightly different variants of each segment: an A version and a B version. A unique identity is then created by selecting between those variants in a specific pattern as the stream is assembled for a given user or session.

This approach matters because it balances scalability with forensic value. It allows a platform to create individualized outputs without having to store or process a fully separate asset for every viewer, making it far more practical for large-scale streaming environments.

Implementing this kind of watermarking system is cheap and trivial to do. Assuming that only GAFAM-scale actors can do it is naive.