what are .webp files and why has my online experience been plagued by them?
I don't know what a .webp file is but I don't like it. They're like a filthy prank version of the image/gif you're looking for. They make you jump through all these hoops to find the original versions of the files that you can actually do anything with.
Edit: honestly I assumed it had something to do with Google protecting themselves from image piracy shit
Jpg is ancient, and gif, holy shit gif is from stone age.
I dunno, if you're playing a video, you probably want x264 or better these days, no? For music, we use some variant of mp4 or lossless at this point.
Yet with pictures, for some reason we insist on the old shitty stuff.
Using jpeg or gif is like using mp1 for music and VideoCD for video. Come on now.
The only problem with webp is that there's quality loss if you convert an already compressed jpeg into webp with high compression rate, like some web sites do. That can suck, but I don't know how else to get people to use more modern formats. Otherwise we'd be using ancient formats into the 24th century.
It is more efficient. I thought it's obvious, that's why web sites use it, to save traffic and potentially storage. Hence my comparison to video formats. You don't see YouTube playing videos in Real Media format.
It's also more universal, combining features of jpg, png and gif. Gif especially is a dreadful format for what it's commonly used. It was designed for tiny clipart animations, not HD video clips. Something like x265 can actually be hundreds of times more efficient.
It stands for animated png btw. It was an extension. The benefit was that it always rendered something everywhere, if it didn't support the animation, because it would be read out as a regular (but suspiciously heavy) png in that case.
I brought it up because it's yet another old-ass .gif solution that didn't stick because people love the term "gif".
I think PNG is a good format even today. It's lossless compressed, so there isn't that much you can squeeze out of that with new algorithms as you can out of lossy formats with new and smarter approaches.
Sadly, PNG is being terribly misused on the internet too. What it's good for is simple drawn graphics, which it can compress to oblivion. So it's perfect for screenshots of say, your operating system's windows. I took a sshot as I'm typing this, and it came out as 190 kB. Not bad.
But what it's so commonly used for, is people taking screenshots of photos such as from Instagram, and then reposting them. So instead of a tiny and shitty 50 kB IG picture, you get a 1.5MB PNG screenshots. Some then recompress it to a 1.5MB JPG for "maximum quality" when they realise they can't upload PNG to photo sites.
I also very often encounter huge PNG photos with their extensions changed to JPG, and I don't know how or why that is happening.
people taking screenshots of photos such as from Instagram
This one really grinds my gears. Why do so many people insist on sharing text by taking a picture instead of pasting the text? Or better yet, just linking to the original? It's such a waste of bandwidth :(
I also very often encounter huge PNG photos with their extensions changed to JPG, and I don’t know how or why that is happening.
Probably people uploading to sites that limit extensions thinking they're clever by just changing the extension, or being straight up wrong in thinking the extension changing actually changes the file type.
The sites might not bother to check the metadata, and anything worth any salt that displays the image will ignore the extension anyway.
That sounds logical, but on most operating systems these days the extension is hidden, and/or you need to go through some hoops to change it. So I would think that most people who think that wouldn't even know how to change it.
But more importantly, where do those PNGs come from in the first place? Sure, some are clearly screenshots such as of IG or TT, but there are tons of large PNG images that are clearly photos from cameras that someone just took and resaved as png (and later, or someone else, then renamed to jpg).
I could understand that happening occasionally for a bunch of reasons, but I've encountered this so many times, it's pretty bizarre.
Btw it's something you might not even notice if you aren't using e.g. an image viewer that uses a different icon or background based on actual image type.
I suppose another solution might be that it falls under those lines, but some misbehaving services where they're uploaded are giving out improper filenames and not confirming the type.
Though I can't imagine many of those being incredibly popular, or, it's just that images are recycled for so long that eventually many of them hit such a site in their lifetimes.
After reading through more comments in this discussion, maybe I have my answer, at least partially. A lot of people here suggest to use extensions and other ways to convert webp from the web to... PNG.
And then that PNG gets shared further...
Oh gawd if that's why so many pngs are on the net... What a way to take a good idea and completely fuck it up. Now instead of a 2MB jpeg or 0.5MB webp we deal with 10MB pngs 🤦♂️
Just for you, I turned it into webp and made it 40KB. I could have also made it look less like garbage, but then it would have been 50KB, which is unacceptable.
Lemmy's official conversions is 124 kB. Next time I'm making a meme, I'll see how small I can make the size with webp. Should be better when I'm working from my original and not an already compressed jpeg.
It's still bad compared to modern lossless algorithms. PNG is very old and even though PNG encoders have evolved, it is still fundamentally a decade behind modern lossless compression algorithms.
For example: JPEG XL in lossless mode compresses at least 30% better than PNG.
Also, PNG is not actually lossless in many cases. PNG only supports RGB colorspaces. If you try to store anything that's not in an RGB colorspace (e.g. a frame of a video) in a PNG, you will lose some color information as colorspace conversion is not lossless. Example of someone running into this issue: https://stackoverflow.com/q/35399677
JPEG XL supports non-RGB colorspaces so you don't have this problem.
Okay but that difference is not as critical as with jpg, which are also more abundant. The bigger problem with png is that people use it for things it's not meant or designed for - frame of a video being case in point.
If anything, it just proves how lacking we are in other image formats, when we keep shoehorning clipart formats like png and gif into other duties. Well not lacking as in not having them, but not using them.
There's no real reason why you shouldn't use PNG for a frame of video. I'm not talking about using it as a video format, I'm talking about extracting a frame from a video and sending it off to an editor for inclusion in another video or image.
As a user, I would expect that I could use the most popular lossless image format if I want to losslessly share a frame from a movie with someone.
Of course I do agree that we need adoption of other image formats. We really should not still be cramming everything in PNGs or JPGs in 2023.
There is, exactly because png is made for strict rgb colorspace. Especially today when videos can be in HDR and with all kinds of color correction shenanigans, so which you won't get back once you try to put the PNG back into the video.
But I'm not a video editor, so I don't know what still format is best suited for this. I imagine real editors can deal with it, and for regular people who just make screenshots for memes, it's good enough. As I said, png is still a good enough format, but let's not use it for stuff like converting webp photos for further sharing.
I think we generally agree but I just want to clarify anyways. I'm not saying we should use PNG to store frames from videos.
What I am saying however, is that we should replace PNG with a modern lossless image format that is more flexible so users don't have to deal with these issues. All this colorspace stuff should be automatically handled and I shouldn't have to worry about it not being lossless. If I want to save a frame of video, I should be able to do it using an image format that everybody recognizes and accepts, it should not be a huge hassle and it should be fully lossless.
Yes. The problem is getting developers/companies to support anything beyond than just the basics. There's enough formats that could work, but if the big players only support these formats, everything else is secondary. Just like many music players still play only mp3.
PNG compresses like nothing else when it comes to graphs, text, UI elements, digital drawings, comics, screenshots from apps etc. And doesn't suffer from "mosquito" artifacts and other .jpg nonsense. It was never meant to be used for photographs and other statistically "noisy" images for which .jpg works much better.
But requirements for audio hasn't changed that much, and overall it's a much older and thus mature technology, that there isn't much left to figure out. Consumer CD format with 16bit 44.1kHz has been around for 40 years, and you don't need much better quality than that. So there isn't much left to figure out.
But images and videos are different. 20 or 30 years ago you didn't need to commonly send 20 MPix HDR photos and HD to 4k videos over the internet. Shoehorning formats that were made for 640x480 pictures and tiny silly clipart animations just doesn't make sense, especially with all the development that's been made in that time. Newer compression techniques can help, but you can only do so much.
Shoehorning formats that were made for 640x480 pictures
Err...nothing in the file format spec restricts jpg to a particular size. I would actually argue that this undermines your point -- bandwidth was incredibly limited in the 90s compared to what I see today.
Simple example: a 640x480 image is (at least) 307,200 bytes = 0.3M, so it takes at least 5.4 seconds to transmit over a 56k modem.
A 4k image, same color depth, is 16000000 bytes = 15M. On a gigabit connection (what I have), that takes about 0.02 seconds.
I was taking more about quality than size in this particular comparison. In 1993 you were happy to squeeze through an image in any quality almost.
It goes hand in hand tho.
If you can compress a 50 MPix, 16-bit, high dynamic range image from a modern high-end DSLR to a reasonable size with a better algorithm and format, you'd also have an easier time squeezing a crappy 640x480 pic to an even smaller size. We just couldn't do either so well 30 years ago.