Most people don’t know your photos can be cross profiled and identified by the unique noise signature of your camera.
I’ve never heard of it being used in practice though. There’s a github repo somewhere if you’re interested in trying it yourself.
Partly true but Amazon has been known to make dupes of a product and sell it at a loss to drive the competitor out of business. I’m sure that threat lingers in some peoples’ minds.
They probably also promise the management team “you’ll still have creative control after we buy you”. Then turn around and replace each of them one by one
This has been such a depressing trend over the last few decades.
Fresh bright-eyed startup with a passionate creator develops some interesting and innovative product.
Gets bought out by FAANG, turns to shit and stops working properly or gets discontinued, and primarily functions as a spyware device.
This is amazing. Thank you!
This is fine if the post is something insanely low effort.
But I do worry if this ends up being too aggressive.
One of the things that made reddit so awful is how over moderated it was.
I don’t really take issue with dozens of posts by newbies asking the same basic question over and over. I used to be one and am occasionally back there again if I start a new hobby. Hopefully newcomers don’t get pushed off by overly sensitive moderation.
It would be helpful if you could provide a hypothetical example of what is considered a “low effort” post.
Join lat-team private tracker. They have everything you’d ever want.
Your first suggestion is a clever one.
I can imagine writing a small script on the host machine to listen for subdomains, forward them to pfsense to update the aliases, and possibly set them to expire after a few days for security reasons.
Surely something like this exists. How to find it…?
There’s a few apps I need to split out. Top priority is the signiant app which according to their documentation requires various AWS subdomains as well as their own. Specific subdomains are not specified and are implied to change regularly/on demand.
In an ideal world I would do my split tunneling on the device itself, but I don’t trust Windows and thus I run my VPN at the router level.
This isn’t a problem for most things, but I need to utilize my full bandwidth to transfer large files to clients in a timely manner, and a VPN becomes a massive bottleneck.
Pfsense lets you alias by domain name (I believe it regularly resolves down to an IP and uses that for filtering), but again, you need to supply the exact subdomain.
Just wondering if there’s an alternative solution to this issue. If it’s external to pfsense that’s not the end of the world.
Worst case scenario, I would set up a dedicated Linux box or maybe even a VM which could share access to the file transfer NAS and split tunnel the entire box around the VPN. Definitely less convenient.
This looks promising. I’m going to investigate further. Thanks!
The question is whether they were using infrared to see through normal sunglasses. IR blocking sunglasses prevent “night vision” from seeing through the lenses. Under infared, you can see through normal dark sunglasses like they aren’t even there.
Got any further info about these more advanced methods?
I haven’t seen anything beyond the feature distance models. I have seen the models that essentially recreate you entire anatomy in 3D, place it in a database, then use that profile to match to in the future–almost like a 3D match move artist would do for visual effects. Not sure if this is just a proof of concept though.
I wouldn’t be surprised if the millimeter wave scanners at airports have been collecting 3D models of us for this database over the last decades.
IR blocking sunglasses are the simplest and most practical solution.
Facial recognition systems compare the distance ratios between your eyes and nose primarily. Hiding your eyes is very effective towards fucking that up. A mask alone is typically not enough.
That immediately came to mind. But is there a way to have it reroute to a different platform if the video is on an alternative platform? I follow a few youtubers in grayjay that I only recently discovered mirror on peer tube. Grayjay didn’t give any hints.
I’m not sure whether you’re asking why to use a frontend vs YouTube rawdog or conflating Odysee/Peertube with a YouTube frontend. I thought they were frontends as well for a long time.
If you’re asking why to use a frontend proxy for YouTube, there could be a few reasons. The obvious being privacy concerns, but other people prefer the less cluttered interface, no ads, no YouTube premium or sign in with google popups, no manipulative algorithm.
This is actually a brilliant idea.
To play devil’s advocate, I could argue the monetization of YouTube has lead to all of the quirks of the platform people like us hate.
The list goes on.
Although I agree there should at least be some way for creators to recoup the expenses they put towards producing videos. An ad-free patreon style donation system seems the most practical and balanced from what Ive experienced.
Fair point. I’m sure many would disagree with me, but for web video anything more than HD is pointless except for very niche content. But even HD streaming at scale is taxing and expensive.
Airlines make the majority of their money from a small percentage of flyers paying business and 1st class. I think there’s a world where this principal can be applied to something like peertube hosting in some form.
Do you mind sharing your channel? EDIT: Nvm found it. Added the new hyprland video to my watch list.
And can I ask how you find other interesting channels?
I’m sure you weighed the pros and cons of Peertube vs Odysee. What made you choose peertube?
I agree with you for the majority of “content creators”. But I think there’s a sizable number of people who aren’t interested in making videos for a profit and I imagine there’s a fair overlap with people in this community and the fediverse at large.
If I were to create videos I would make them on either peertube or Odysee. I wasn’t really aware of either platform other than vague whispers of them until recently, and I find it difficult to gauge the community sentiment on which of these platforms would be suitable for finding interesting content as well as posting it, hence this post.
I believe this is the original whitepaper: https://ws.binghamton.edu/fridrich/Research/double.pdf
And here’s an implemention I found on github: https://github.com/andrewlewis/camera-id
With that repo you should be able to test ways to obfuscate the noise signal.