I just uploaded a new article describing how a modern video game texture is stored in graphics memory, why it's done this way, and how it may differ between platforms.
It is called Anatomy of a Texture.
Spent way too long trying to get a rusty metal texture just right in Unity once. The PBR pipeline is no joke.
Really nice article, thanks for writing it! I recently dabbled in some of this in connection with modding a game and there encountered the hierarchy you describe of mips, tiles, blocks and texels, so very relatable. Great write-up!
Really enjoyed this, refreshing to read something passionate and hand written. I had no idea how much complexity existed in something as innocuous as handling textures between platforms.
Also pretty cool getting a glimpse into the work 505 does helping developers get their games in front of more people. I had no idea 505 was the publisher[0] for so many recognizable titles.
Suggestion: Adding dates to your blog posts would be helpful. The first sentence mentioned "last year" and I was confused what that meant since the post was undated.
That's a great suggestion. I'll probably add that tomorrow.
Why not compress the texture using NTC, VAE-VQ or some neural codec into a general embedding space of uint8 and then just expand and create mipmaps and similar using the target hardware standard?
The easiest answer is that most tools and hardware in use predates those type of embeddings.
The advantage with block compression is that it is purpose built to minimize bits used, it's local to small blocks, decoding and encoding can be done with fairly simple code[0], and it doesn't require any specific AI hardware.
I am unfamiliar with NTC, but at first glance it seems to require an Nvidia card with RTX features, meaning it wouldn't be supported on Xbox One and PlayStation 4 (AMD GCN), Xbox Series X|S and PlayStation 5 (AMD RDNA2), nor the original Switch (Nvidia Tegra X1).
[0] https://learn.microsoft.com/en-us/windows/win32/direct3d11/b...
NTC requires GPU for real-time decoding, but you can use it as an compressor only to have a portable representation. It's actually the common use-case today, you can on-device offline decode directly to BC7 if you feel like it!
Doesn't that defeat most of the benefit though?
Its a smaller binary size, general representation and amortizes the decoding to the consumer hardware
This question reminds me of one that people apparently have seriously asked: "Why didn't they simply send up the Space Shuttle to rescue the Apollo 13 astronauts?"
Haha older hardware can do multiply-accumulates as well its not an insurmountable compute load :P
enjoyed this
Fantastic read! Not even a hint of LLMisms.
Thank you. I put a lot of care and effort into explaining my own insights the way I felt they needed to be expressed in my own voice. Letting a machine replace those words would have ruined that premise.