Page 3 of 3

Posted: Thu Dec 13, 2007 8:03 pm
by mystran
Craze Frog wrote:The problem is that no such way of scaling exists apart from nearest neighbour, which is THE most ugly scaling method ever. Feel free to invent it.
Sampling theorem says that if a signal is bandlimited, it can be perfectly reconstructed. Resampling is the same as reconstructing the continuous time signal, then sampling again (hence resampling). Since the Nyquist frequency in all intermediate phases is higher than the original Nyquist, no information is lost, so it's possible.

In theory you need to use the sinc-function (which is infinite) as the reconstruction filter, but in practice it is enough to use a filter such that any error is less than quantization noise (which is at approximately -48dB for 8-bit signal). Ofcourse such a filter would still be quite long, so it'd take a while to compute, but we honestly aren't talking about rocket science here.

[edit]: Ooops, I was being sloppy. Bitmap is finite, and the process is therefore offline batchjob, so we can rule anything outside the bitmap is zero, then convolve with infinite sinc-function, since it need not be evaluated in practice outside the bitmap, so there isn't necessarily a reason (other than performance) to go with a finite length filter kernel.

Posted: Fri Dec 14, 2007 10:46 am
by Candy
mystran wrote:(which is at approximately -48dB for 8-bit signal).
That would be 24dB (8*3dB).

Posted: Sat Dec 15, 2007 7:33 am
by mystran
Candy wrote:
mystran wrote:(which is at approximately -48dB for 8-bit signal).
That would be 24dB (8*3dB).
No. In digital signal processing one bit is normally taken to be 6dB, but in reality we're talking about something like 6.0206dB which gives 48,1648dB of dynamic range for 8-bits when the quantization noise is assumed uniform.

I would also like to note, that when talking about dynamic range of digital signals, talking about power makes no sense whatsoever.

Posted: Sun Jan 06, 2008 10:38 pm
by AndrewAPrice
I would like to see kernel-level OpenGL support.

Client windows could render onto a texture, an each window would be rendered as two triangles using that texture. Then you could enable z-perspective and watch background objects disappear into the background.

It'd be similar to how Windows Vista uses Direct3D 10 for everything.

Posted: Mon Jan 07, 2008 12:00 am
by Brynet-Inc
MessiahAndrw wrote:I would like to see kernel-level OpenGL support.

Client windows could render onto a texture, an each window would be rendered as two triangles using that texture. Then you could enable z-perspective and watch background objects disappear into the background.

It'd be similar to how Windows Vista uses Direct3D 10 for everything.
While not exactly kernel-level, Unix/Linux can use AIGLX+Xorg.. AFAICT, Everything is rendered using OpenGL.

I haven't used it myself... OpenBSD doesn't support any 3D acceleration as of 4.2, although works is being done in -CURRENT, so it might be working by 4.3. 8)

Posted: Mon Jan 07, 2008 3:54 am
by ucosty
Out of curiosity what made you choose OpenBSD over FreeBSD?

edit: And finally 100 posts :)

Posted: Mon Jan 07, 2008 4:21 am
by Brynet-Inc
ucosty wrote:Out of curiosity what made you choose OpenBSD over FreeBSD?
I guess several different factors, secure by default.. 2 remote holes in 10 years, light base install.. a directory hierarchy that was easy to wrap my head around.

And puffy.. Nuff' said. 8)

Posted: Mon Jan 07, 2008 4:36 am
by blound
Brynet-Inc wrote:
ucosty wrote:Out of curiosity what made you choose OpenBSD over FreeBSD?
I guess several different factors, secure by default.. 2 remote holes in 10 years, light base install.. a directory hierarchy that was easy to wrap my head around.

And puffy.. Nuff' said. 8)
secure? does it have randomized .text sections yet?