Timeline for Why don't most major game engines use gifs for animated textures?
Current License: CC BY-SA 3.0
18 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2017 at 12:33 | history | edited | CommunityBot |
replaced http://webmasters.stackexchange.com/ with https://webmasters.stackexchange.com/
|
|
| Oct 2, 2016 at 3:36 | review | Suggested edits | |||
| Oct 2, 2016 at 4:55 | |||||
| Sep 29, 2016 at 17:07 | comment | added | Kromster | @gsus Looks like, as with other drawbacks, this one can be worked around with using some uncommon GIF features. Could you provide a source please? I will be happy to add this into the answer. | |
| Sep 29, 2016 at 14:10 | comment | added | gsus | I don't think that generally has to be true (or I don't understand it fully), because you can create animated Gifs, where each Frame is a full frame, and not just the Diff-Image (don't know the official term for those Frames that only have the information that differs from the Frame before). Although ofc, the GIF ist much bigger then. But doesn't that mean you can have Random Access if the GIF is created this way? | |
| Sep 29, 2016 at 12:24 | comment | added | luk32 | @Agent_L The bullet clearly said "don't compress well". Which itself doesn't have any merit. Sorry, if I wasn't clear enough. I meant there are other problem with GIF compression, but this wasn't one. JPG wouldn't compress well either, nor PVRTC or any other HW oriented texture compression algorithm. Just because they are already compressed formats. Kromster understood me, so I guess it's fine. | |
| Sep 29, 2016 at 12:21 | history | edited | Kromster | CC BY-SA 3.0 |
added 229 characters in body
|
| Sep 29, 2016 at 12:02 | comment | added | Agent_L | @luk32 He didn't said that "gifs don't let themselves be compressed well" , he said that "gif compression is poor". And you don't use loseless compression in textures, you use lossy one like JPG, and then you measure difference not in percents but in times. | |
| Sep 29, 2016 at 12:01 | comment | added | Kromster | @luk32 the fact that they enforce specific compression is enough to be listed as a drawback. I'll update the answer, thanks for pointing! | |
| Sep 29, 2016 at 11:53 | comment | added | luk32 | My point is that the fact they are do not compress much is not a drawback itself, any texture storage using compression would exhibit this. It's obvious, one should not compress it. A drawback might be not being able to choose the method (or lack of one), which does limit usability a bit. BTW what are the much better compression methods? I've seen reports of 5-25% in PNGs deflate, vs GIFs LZW, it is not that much IMO, but sure it might be a serious limitation being stuck to LZW. | |
| Sep 29, 2016 at 11:02 | comment | added | Kromster | @luk32 this is a drawback because there are much better compression methods than old GIF compression. So you either use suboptimal GIF compression or get to (de)compress twice (note that compressing already compressed data generally yields larger size too). | |
| Sep 29, 2016 at 10:52 | comment | added | luk32 | Of course they don't compress well, they are already compressed. Why would it be a drawback? | |
| Sep 29, 2016 at 10:36 | comment | added | Kromster | @TomášZato True. Sorry, did I state otherwise? | |
| Sep 29, 2016 at 9:52 | comment | added | Tomáš Zato |
@Kromster Actually, resource savvy sites use custom animation mechanism for animated images, such as CSS background-image position animation where the animation is palced on one big image, which is then moved by offsets.
|
|
| Sep 29, 2016 at 4:16 | comment | added | Kromster | @Erik exactly that. Common GIF supports up to 256 colors, out of which just 1 can be marked to be transparent. Unlike other formats, where transparency typically has much wider range. | |
| Sep 28, 2016 at 21:53 | comment | added | Erik | I don't understand "one color for transparency." What does it mean to have multiple colors for transparency? Are you talking about varying alpha values like applying a transparency gradient across a texture? | |
| Sep 28, 2016 at 4:36 | history | edited | Kromster | CC BY-SA 3.0 |
added 1114 characters in body
|
| Sep 27, 2016 at 18:01 | history | edited | Kromster | CC BY-SA 3.0 |
deleted 1 character in body
|
| Sep 27, 2016 at 17:54 | history | answered | Kromster | CC BY-SA 3.0 |