Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules. |
12th December 2006, 20:44 | #43 | Link |
Registered User
Join Date: Mar 2006
Posts: 567
|
Because Intel is not making money only from selling Core2 Duo processors. They are actually producing more hardware and software than just cpus for personal computers.
Last edited by KoD; 12th December 2006 at 20:46. |
12th December 2006, 21:51 | #44 | Link |
Registered User
Join Date: Dec 2005
Location: Qetchua mountains in Peru, and Klingon battlecruiser D'Mar
Posts: 393
|
Certainly, but...
If someone decide to make pressure on Intel software division, who knows what can happen?
__________________
Live long and prosperLive long and prosperLive long and prosper |
13th December 2006, 03:18 | #45 | Link |
Mr. Sandman
Join Date: Sep 2003
Location: Haddonfield, IL
Posts: 11,768
|
It's like asking M$ to make Windows OpenSource...
and however x264 has ASM optimizations that maybe are even faster than intel libs...
__________________
MPEG-4 ASP Custom Matrices: EQM V1(old), EQM AutoGK Sharpmatrix (aka EQM V2), EQM V3HR (updated 01/10/2004), EQM V3LR, EQM V3ULR (updated 04/02/2005), EQM V3UHR (updated 17/12/2004) and EQM V3EHR (updated 05/10/2004) Info about my ASP matrices. MPEG-4 AVC Custom Matrices: EQM AVC-HR Info about my AVC matrices My x264 builds. Mooo!!! Last edited by Sharktooth; 13th December 2006 at 03:21. |
13th December 2006, 04:27 | #47 | Link |
Mr. Sandman
Join Date: Sep 2003
Location: Haddonfield, IL
Posts: 11,768
|
Hand optimizations are always better than third party libs (that work well only on certain scenarios), also qualitywise x264 is as fast if not faster than commercial h.264 encoders... so how can you say it's not optimized?
Maybe there could be further speed optimizations but it's already quite fast.
__________________
MPEG-4 ASP Custom Matrices: EQM V1(old), EQM AutoGK Sharpmatrix (aka EQM V2), EQM V3HR (updated 01/10/2004), EQM V3LR, EQM V3ULR (updated 04/02/2005), EQM V3UHR (updated 17/12/2004) and EQM V3EHR (updated 05/10/2004) Info about my ASP matrices. MPEG-4 AVC Custom Matrices: EQM AVC-HR Info about my AVC matrices My x264 builds. Mooo!!! |
13th December 2006, 04:41 | #48 | Link |
Fighting spam with a fish
Join Date: Sep 2005
Posts: 2,699
|
I thought that Ateme blew it out of the water, maybe not in terms of quality, but speed wise. Maybe I am just hearing voices again.
Also, I ask about hand optimizations because I am interesting in learning to program and speed is a fetish. Ha! |
13th December 2006, 10:17 | #49 | Link |
pencil artist
Join Date: Jan 2006
Location: Slovakia
Posts: 201
|
so, learn about programming. It's fun and it doesn't hurt - allright, debugging can be very frustrating, but the bigger will be your joy after you finnaly manage to make the code behave the way you want.
Large programs are broken into smaller parts: functions. When you want to make your program run faster, you search for the functions that are most computationally intensive and then you find a better (faster) algorithm for it, you rewrite it in assembly, or very often both. Compiler may not generate the best assembly code for your function -it's just a program, not thinking being- in contrast with creator of the code, who is perfectly aware of what the function is supposed to do. Optimalizations (most of the time?) take advantage of SIMD (single instruction - multiple data) instructions: MMX, 3DNow!, SSE1/2/3 aso. They are kind of... parallel processing instructions, they provide huge speedup if properly used. X264 has plenty of optimized functions. It's kind of complicated for compiler to rewrite your function into SIMD instructions. So is for humans... but humans are intelligent, thinking beings, and creative too. You can find out an algorithm, that would work much better in SIMD. Compiler sort of... can't. That's why I believe humans will always (or for a very long time) be better for optimizing. Also, about optimizing for some CPU - under this I understand using instructions that suit the CPU is best at: be it 3DNow for (old?) amds, SSE2 for core2, 64-bit instructions for 64bit processors on 64-bit OS etc. So in my eyes x264 is pretty much optimized for core2. Also I admire the programmers that take part in developing our beloved oss encoders and find them very skilled and inteligent (and I am most thankful for their work); I doubt that Intel's optimized code would be significantly faster - IF it would be faster at all. sorry if this was kind of boring reading (also corrections welcomed). I also think that x264 is very fast taking it's complexicity (that makes me dizzy) into consideration (hey, it got faster by >30% during 2006 according to my tests). Big thankyou for the devs. |
13th December 2006, 13:48 | #50 | Link |
Friendly Coder
Join Date: Jun 2005
Posts: 18
|
Hear, hear, imcold!
Anyone wanting to learn more about real programming and squeezing the most out of those clock cycles could do worse than read Don Knuth's magnum opus: http://www-cs-faculty.stanford.edu/~knuth/taocp.html And I'd like to add my thanks to the X264 developers too. Happy holidays, everyone! |
13th December 2006, 16:56 | #52 | Link |
Registered User
Join Date: Oct 2005
Posts: 106
|
if you wanna learn more check this links:
Software optimization resources :http://www.agner.org/optimize/ Assembly Optimization Tips : http://www.mark.masmcode.com/ |
21st December 2006, 15:10 | #53 | Link |
Turkey Machine
Join Date: Jan 2005
Location: Lowestoft, UK (but visit lots of places with bribes [beer])
Posts: 1,953
|
Got a suggestion. Bear with me.
The progress indicator only updates every (total frames / 1000), thus for a 100000 frame movie, it updates every 100 frames. This is a little slow with some resolutions, so why not have it update the progress every frame and not by percentage? The current percentage could then go down to 2dp or 3dp instead of 1dp to accomodate the longer encodes, and be calculated each frame.
__________________
On Discworld it is clearly recognized that million-to-one chances happen 9 times out of 10. If the hero did not overcome huge odds, what would be the point? Terry Pratchett - The Science Of Discworld |
26th December 2006, 18:30 | #55 | Link |
Registered User
Join Date: Apr 2003
Posts: 92
|
Sorry if this crosses over into feature request territory, but in the advanced dvd authoring forum we've been discussing codecs that are hd-dvd compliant. X264 is not. Some minor header changes need to be made to conform to spec. Is this being worked on? Fantastic work otherwise, certainly my choice for pc backups.
Heres the thread: http://forum.doom9.org/showthread.php?t=119392&page=4 |
27th December 2006, 10:05 | #59 | Link |
Registered User
Join Date: Apr 2003
Posts: 92
|
I guess the people on the forum who have the specs know. I do know JM supports those data structures which were said to be needed by those who have the specs. I thought JM only outputs avi and I have a feeling demuxing would ruin it.
|
27th December 2006, 10:32 | #60 | Link | |
Registered User
Join Date: Jan 2002
Location: France
Posts: 2,856
|
Quote:
__________________
|
|
Tags |
coding, development, x264 dev |
Thread Tools | Search this Thread |
Display Modes | |
|
|