Re: Opinion Poll: File Size vs Loading Time?


	/-/S'pht-Translator-Active/-/

Re: Opinion Poll: File Size vs Loading Time?

Posted By: treellama Date: 9/22/06 2:14 a.m.
In Response To: Re: Opinion Poll: File Size vs Loading Time? (Forrest of B.org)

: I just wish I could do good quality batch converts to it on my Mac. I still
: tweak the textures around a lot, and having to send them off to someone
: else to convert them to DDS every time I make a change would be a bitch
: for me and the someone else. Not to mention converting them all in the
: first place... though I could just ask you to do that part I suppose. But
: it's the tweaks that really kill it.
To be honest, I was hopeful that Aorta would do a better job compressing them. It still looks good most of the time, but it doesn't nail each and every texture like nvidia's tools do.
I could add a batch convert, so your testing would at least accurately reflect load times and framerates, even if the image quality isn't the super highest. Many games use Aorta's way of compressing their textures.
I think you can see now why I picked the format I did, even if the Mac tools aren't there yet. For high res textures and 3D games, there's really no other choice.
: Until there's a Mac program that will let me batch convert hundreds of files
: at once with good quality to DDS, I think I'll just keep using PNG until
: just before release, then have someone with a PC convert them then.
An uncompressed format like BMP might load faster than PNG. Since you already distribute the alpha releases compressed, no point making Aleph One do the decompression each time it loads textures when disk space is so cheap. PNG and BMP are the same size in a zip file.
: I don't see a difference, but then I don't know what I'm looking for. Here's
: two screenshots...
: What's the fix supposed to fix, anyway?
Geforce cards (all the way up to the FX) have a bug with DXTC1 (which is what I use for wall textures without masks, and landscapes) where they only use 16-bit color for the second pair of colors in each block, instead of 32-bit color. This can cause banding in textures with color gradients.
You don't seem to have any textures like this, and the nvidia tools dither textures to make the banding less noticeable too.
Aleph One's geforce fix adds an opaque alpha channel and tells the video card the textures are DXTC3, which the bug doesn't affect for some reason. DXTC3 textures are twice as big as DXTC1, so if you can't see a difference, leave the fix off.

[ Post a Reply | Message Index | Read Prev Msg | Read Next Msg ]
Pre-2004 Posts

Replies:

[ Post a Reply | Message Index | Read Prev Msg | Read Next Msg ]
Pre-2004 Posts

Your Name:
Your E-Mail Address:
Subject:
Message:
: To be honest, I was hopeful that Aorta would do a better job compressing : them. It still looks good most of the time, but it doesn't nail each and : every texture like nvidia's tools do. : I could add a batch convert, so your testing would at least accurately : reflect load times and framerates, even if the image quality isn't the : super highest. Many games use Aorta's way of compressing their textures. : I think you can see now why I picked the format I did, even if the Mac tools : aren't there yet. For high res textures and 3D games, there's really no : other choice. : An uncompressed format like BMP might load faster than PNG. Since you already : distribute the alpha releases compressed, no point making Aleph One do the : decompression each time it loads textures when disk space is so cheap. PNG : and BMP are the same size in a zip file. : Geforce cards (all the way up to the FX) have a bug with DXTC1 (which is what : I use for wall textures without masks, and landscapes) where they only use : 16-bit color for the second pair of colors in each block, instead of : 32-bit color. This can cause banding in textures with color gradients. : You don't seem to have any textures like this, and the nvidia tools dither : textures to make the banding less noticeable too. : Aleph One's geforce fix adds an opaque alpha channel and tells the video card : the textures are DXTC3, which the bug doesn't affect for some reason. : DXTC3 textures are twice as big as DXTC1, so if you can't see a : difference, leave the fix off.

If you'd like to include a link to another page with your message,
please provide both the URL address and the title of the page:
Optional Link URL:
Optional Link Title:

If necessary, enter your password below:
Password:

Problems? Suggestions? Comments? Email maintainer@bungie.org
Marathon's Story Forum is maintained with WebBBS 5.12.

Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/19/06 2:06 p.m.
Faster load times over file size NM	Blake37	9/19/06 2:11 p.m.
Same here. NM	newreagan64	9/19/06 2:15 p.m.
Agreed NM	kyjel	9/19/06 2:51 p.m.
Re: Agreed NM	D-M.A.	9/20/06 12:48 a.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/19/06 3:16 p.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/19/06 3:52 p.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/19/06 4:14 p.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/19/06 5:53 p.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/20/06 2:54 a.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/20/06 8:32 a.m.
Re: Opinion Poll: File Size vs Loading Time?	RyokoTK	9/20/06 5:36 a.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/20/06 8:34 a.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/20/06 3:06 p.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/21/06 5:48 p.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/21/06 6:26 p.m.
Re: Opinion Poll: File Size vs Loading Time?	Forrest of B.org	9/21/06 10:40 p.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/22/06 2:14 a.m.
Re: Opinion Poll: File Size vs Loading Time?	treellama	9/24/06 5:24 a.m.
Re: Opinion Poll: File Size vs Loading Time?	Blake37	9/19/06 4:28 p.m.



	Your Name:
	Your E-Mail Address:
	Subject:
	Message: : To be honest, I was hopeful that Aorta would do a better job compressing : them. It still looks good most of the time, but it doesn't nail each and : every texture like nvidia's tools do. : I could add a batch convert, so your testing would at least accurately : reflect load times and framerates, even if the image quality isn't the : super highest. Many games use Aorta's way of compressing their textures. : I think you can see now why I picked the format I did, even if the Mac tools : aren't there yet. For high res textures and 3D games, there's really no : other choice. : An uncompressed format like BMP might load faster than PNG. Since you already : distribute the alpha releases compressed, no point making Aleph One do the : decompression each time it loads textures when disk space is so cheap. PNG : and BMP are the same size in a zip file. : Geforce cards (all the way up to the FX) have a bug with DXTC1 (which is what : I use for wall textures without masks, and landscapes) where they only use : 16-bit color for the second pair of colors in each block, instead of : 32-bit color. This can cause banding in textures with color gradients. : You don't seem to have any textures like this, and the nvidia tools dither : textures to make the banding less noticeable too. : Aleph One's geforce fix adds an opaque alpha channel and tells the video card : the textures are DXTC3, which the bug doesn't affect for some reason. : DXTC3 textures are twice as big as DXTC1, so if you can't see a : difference, leave the fix off.
	If you'd like to include a link to another page with your message, please provide both the URL address and the title of the page:
	Optional Link URL:
	Optional Link Title:
	If necessary, enter your password below:
	Password: