Skip Navigation

Posts
814
Comments
640
Joined
3 yr. ago

Currently studying CS and some other stuff. Best known for previously being top 50 (OCE) in LoL, expert RoN modder, and creator of RoN:EE's community patch (CBP). He/him.

(header photo by Brian Maffitt)

  • Sorely missing a com/mag/sub/your-preferred-jargon-word for theocho right now

  • What a save!

  • There's a balance since (at least in general) you can just accept slightly higher file size to offset the advantage of significantly faster encoding time. Some (most?) hardware encoders have "pretty good" quality levels so imo if you can get, say, 90% of the quality at same file size but then offset the loss with 110-120% file size to match "100%" quality, then that can be a pretty appealing tradeoff if it makes your encode an order of magnitude faster.

    The tradeoffs aren't just between quality vs encode time, they're between quality vs encode time vs file size. Of course, where each person's preferences are in making the tradeoffs will vary! But storage is cheap and it's hard for me personally to justify saving what might even generously be a cent's worth of storage costs by making an encode take a relative eternity for the sake of maximal compression efficiency.

  • Only selected apps connect with VPN protection. All other traffic is unprotected.

  • Man I sure hope not. The fact that it doesn't say they are alongside this announcement does make me feel a little better about it given how generally happy they've been to announce all their cost-saving measures lately.

  • I assume you're at least partially joking, but being able to make it work for a single tower (if the performance is good and it's not just marketing) is a bit of genuine engineering innovation

  • For most applications I assume you want to leverage AV1 hardware encoding to use it, and I imagine it would be the same for AV2 etc. We might've been a bit spoilt by H.264 being literally >20 years old at this point (and even VP9 and H.265 being >10 years old).

  • This was absolutely not on my bingo card

  • This was absolutely not on my bingo card

  • CECILIA IMMERGREEN IS GOING TO DIE, CONFIRMED ON FEDIVERSE

  • Justice doodles, my beloved

    Curious what the problem is that's simultaneously not that serious but requires overnight stays and multiple specialists!

  • Beijing claims its homegrown AI processors now match H20 and RTX Pro 6000D

    So do they? Last I heard (which was not that recently) they were still notably behind in performance and I'm surprised if they were able to catch up that quickly?

    The TH article doesn't really discuss the performance claim.

  • I recently had a dream where Bettel died and it felt so fucking weird

    Anyway I love these slow melancholic songs _*

  • I liked it (but didn't love it) when I first listened to it but for some reason I've come back to it and now really love it!

  • To someone who is/was interested in Proton's email service but not their VPN service this may still be a fair point though imo. I don't think it being a rehashed complaint here automatically invalidates it.

  • This just makes me think of 3D printing to customize aesthetics, which in hindsight should've been obvious 🤔

  • I actually think this video is doing a pretty bad job of summarizing the practical-comparison part of the paper.

    If you go here you can get a GitHub link which in turn has a OneDrive link with a dataset of images and textures which they used. (This doesn't include some of the images shown in the paper - not sure why and don't really want to dig into it because spending an hour writing one comment as-is is already a suspicious use of my time.)

    Using the example with an explicit file size mentioned in the video which I'll re-encode with Paint.NET trying to match the ~160KB file size:

    Hadriscus has the right idea suggesting that JPEG is the wrong comparison, but this type of low-detail image at low bit rates is actually where AVIF rather than JPEG XL shines. The latter (for this specific image) looks a lot worse at the above settings, and WebP is generally just worse than AVIF or JPEG XL for compression efficiency since it's much older. This type of image is also where I would guess this type of compression / reconstruction technique also does comparatively well.

    But honestly, the technique as described by the paper doesn't seem to be trying to directly compete against JPEG which is another reason I don't like that the video put a spotlight on that comparison; quoting the paper:

    We also include JPEG [Wallace 1991] as a conventional baseline for completeness. Since our objective is to represent high-resolution images at ultra-low bitrates, the allow-able memory budget exceeds the range explored by most baselines.

    Most image compression formats (with AVIF being a possible exception) aren't tailored for "ultra-low bitrates". Nevertheless, here's another comparison with the flamingo photo in the dataset where I'll try to match the 0.061 bpp low-side bit rate target (if I've got my math right that's 255,860.544 bits):

    • Original PNG (2,811,804 bytes) https://files.catbox.moe/w72nsv.png
    • AVIF; as above but quality 30 (31,238 bytes) https://files.catbox.moe/w2k2eo.avif
    • JPEG XL could not go below ~36KB even at quality 0 when using my available encoder, so I considered it to fail this test
    • JPEG (including when using MozJPEG, which is generally more efficient than "normal" JPEG) and WebP could only hit the target file size by looking garbage, so I considered them to fail this test out of hand

    (Ideally I would now compare this image at some of the other, higher bpp targets but I am le tired.)

    It looks like interesting research for low bit rate / low bpp compression techniques and is probably also more exciting for anyone in the "AI compression" scene, but I'm not convinced about "Intel Just Changed Computer Graphics Forever!" as the video title.


    As an aside, every image in the supplied dataset looks weird to me (even the ones marked as photos), as though it were AI-generated or AI-enhanced or something - not sure if the authors are trying to pull a fast one or if misuse of generative AI has eroded my ability to discern reality 🤔


    edit: to save you from JPEG XL hell, here's the JPEG XL image which you probably can't view, but losslessly re-encoded to a PNG: https://files.catbox.moe/8ar1px.png

  • Proton Meet is in closed beta testing.

    So possibly you can't yet

    edit: based on discussion here you can probably access it at https://meet.proton.me/ if you're a Visionary subscriber (maybe they'll have sent you an email about testing it?). For everyone else it just says "coming soon".