DirectX9 - For More Than Just Gamers?

Xev writes "HEXUS.net are showing a review of a new product called 3DEdit. This uses the DirectX 9 3D rendering engine; 3D transitions; DirectX 9 Shader-based filters, in order to give you a powerful home DV editing suite. This proves a lot more value to me as a Video editor than a card which just lets me play the latest games. Perhaps there is more use for these cards even at a consumer level?"

Comments

Typo

[PhraudulentOne]

For more THAN just Gamers.

Re:Typo

[Carthag]

If it were only a typo it wouldn't really be a problem. However, a and e aren't nearly close enough for this to be anything but ignorance. :)

Re:Typo

[rbarreira]

What about the editors? They're the ones who have the obligation to fix those mistakes, I think...

Re:Typo

[rbarreira]

And oh yeah, I only saw your joke after I posted that :D

Re:Typo

[FIGJAM]

How does a crowd of geeks not understand basic grammar?

Like, doesn't every geek recognize the IF... THEN... ELSE logic?

I am going to take this opportunity to say to all the people who make mistakes like this: Whenever you hear about illiteracy problems, you are part of the statistics.

Hemos is meant to be amongst the king of geeks, operating Slashdot and all. Maybe Slashdot should be News For Dumbass Illiterate Nerds instead.

If you cannot spell very well and/or often cannot use correct grammar then you are just fucking ignorant and you need to recognize this fact. You lack basic knowledge and for it you should have "I am ignorant" stamped on your forehead daily until you at least achieve above average communication abilities.

It's about time these people went back to third grade to learn basic English.

Re:Typo

[Carthag]

Way to go, I put an error in my own grammar nazi post.

Re:Typo

[brkello]

What's the big deal of correcting it? If it helps a single person learn the proper use of then vs. than, then it was worth it.

Re:Typo

[SenorChuck]

Who modded these posts? They should all be either -1 Offtopic or -1 Troll. How did they get modded Insightful or Informative? Time to meta-moderate!

Re:Typo

[jdeisenberg]

Why, oh why, does this matter? It's just one wrong letter.

As another post points out, this is a matter of word confusion, not a simple typo. In either event: if someone posted sample code to slashdot with a one-letter difference that created a syntax or logic error, would you be equally unconcerned?

Re:Typo

[rjkimble]

If you truly care about this stuff, you might want to learn how to spell "existence."

Re:Typo

[bigtangringo]

Nah, he probably meant "ur"

Re:Typo

[rjkimble]

Right. I was just pulling your chain! I can't resist commenting on "spelling/grammar Nazi" posts that contain spelling or grammatical errors themselves. Sorry! :)

Re:Typo

[LordPixie]

But that's just me. I feel the same way about other weird rules of the English language, like capitalization. Why do we do it? It doesn't change the meaning of things one bit, and you can very easily hold entire written conversations without capitalizing anything. /shrug

FYI, capitalization makes the job of the reader a lot easier. They help break up sentences. (Along with periods, obviously) Trying to read a paragraph sans any capitals can be quite the headache.

Likewise, capitals on proper nouns helps distinguish them as such. Now, with context, you can usually figure it out in the end. But it's still a crazy hassle. And in the case of legitimate mistakes, discerning the true meaning of a malformed sentence becomes impossible when one is also throwing away 'convenience' rules such as this.

Basically, written language is a means to communicate. While it still works without some of the rules, it's nowhere near as efficient. So show a bit of respect for your readers, and don't make them spend five minutes decoding your work.


--LordPixie

Re:Typo

[D. Bryce G. H.]

When you think about it, the meaning is changed by this mistake. "For more then just gamers" would mean that it was initially for more, but then later on it was just for gamers.

Sure, the mistake is obvious to you and me, but non-native English speakers may not find it so. And even if you personally don't care, the author may not want to give the impression of carelessness that this mistake implies. I'm sure the site as a whole doesn't.

Re:Typo

[Ironsides]

People think phonetically. That's why we keep screwing up words that don't sound the way they are spelled ("pho"netic, conec"tion" as an example) or words that have the same pronunciation but multiple spellings(they're, their, there). As a native english speaker I have this problem. One reason why I like phonetic languages so much better (such as Japanese).

Re:Typo

[emilng]

This is a website for nerds.
The reason why people complain is because if they misuse a word in their code it probably will throw a syntax error.
The same goes for capitalization. This is just our way of throwing a syntax error at Hemos.
Of course some of us have a more obnoxious and abrasive way of doing it than others, just like any IDE you might encounter.

Re:Typo

[PhraudulentOne]

I don't find it a big deal. I just thought I'd snag first post, and since I really don't care about DirectX, I thought I'd whine about the grammar ;)

Re:Typo

[88NoSoup4U88]

It has come to my attention that most of the non-native-english speaking crowd (which i also belong too) seems to be more aware of -not- making those 'then/than' mistakes , because we are taught the difference (at least I was) when first encountering those words in english class :

I always wondered if the difference between those two words is also explicitly being taught in the US.

No troll, just wondering.

Re:Typo

[88NoSoup4U88]

Y3s, I d0n't sea why peeple are making such problems whan encountering thoose sentences.

Re:Typo

[uglyduckling]

I tend to pronounce "they're" slightly differently to "their" and "there" precisely because I'm aware that "they're" is a contraction of "they are".
What's worrying isn't that people 'think phonetically' but that the majority appear blissfully unaware that these homophones are radically different words. I can understand confusion over "practice" and "practise" because it's subtle (and, IIRC, Americal English doesn't even make the distinction; in British English, -ice is the noun: "a band practice" and -ise is the verb: "the band were practising") but to confuse "their", "there" and "they're", and especially the last, is to be completely unaware of what one is actually saying.

Re:Typo

[RangerElf]

It's a semantic error: the sentence, although syntactically correct, had it's meaning changed by the typo.

Any programmer worth his/her salt knows that semantic bugs are much more difficult to diagnose and correct than syntactic ones.

-gus

Re:Typo

[Profane MuthaFucka]

Ummm, it's "YANKING" your chain. My particular fetish happens to be idioms, and since this seems the thread for it, I thought I'd contribute.

OK, now who will be next? Perhaps it's the guy who hates it when you mix a greek root and a latin suffix on the same word.

Re:Typo

[jsebrech]

but to confuse "their", "there" and "they're", and especially the last, is to be completely unaware of what one is actually saying

No, because when you're typing a stream of thought you're thinking in audio and relying on your brain to subconsciously convert that to type. At least that's how most people do it. Obviously people who were born deaf don't think in audio. I agree it's a consequence of not paying much attention to what you're typing (and so appears more often when people are lazy/tired), but more often that not the speaker actually does know the difference between the terms perfectly, but his brain is messing up the conversion to typed text.

Re:Typo

[networkBoy]

See, I guess I don't get it. I think their and there as two different things, so when I type them out they come out correct. It must be the difference in how people are taught to speak and/or write that makes the difference. I had a large ammount of supplimental education in language outside of the public education I was given.
-nB

Re:Typo

[ultranova]

Shouldn't feed the trolls, but...

How does a crowd of geeks not understand basic grammar?

Why should they ? Why should understanding computers or technology in general give one awesome communication abilities - especially since the time spent on playing around with computers is away from playing around with other people ?

Like, doesn't every geek recognize the IF... THEN... ELSE logic?

Propably, along with most other humans on the planet. What has that got to do with english grammar ?

If you cannot spell very well and/or often cannot use correct grammar then you are just fucking ignorant and you need to recognize this fact.

AFAIK deficiencies in written communication are not directly linked to lack of knowledge about sexual matters. However, deficiencies in spoken communication might be.

You lack basic knowledge and for it you should have "I am ignorant" stamped on your forehead daily until you at least achieve above average communication abilities.

Of course you realize that it is impossible for everyone to have above average skills, due to the definition of the term "average", right ? And you also realize that knowing the definition of "average" is also basic knowledge ?

It's about time these people went back to third grade to learn basic English.

Why ? Why should others learn things just because you find their lack of knowledge disturbing ?

Re:Typo

[uglyduckling]

That should be American English.

Waiting for the chain-yanking...

Re:Typo

[hunterx11]

Japanese is phonetic? Then why do they spell tsunami tu-na-mi? Or futon hu-to-n? Seriously though, nobody wants phonetic writing, or else people with different accents would spell things differently. Phonemic writing wouldn't be too bad, although that doesn't rule out words like "phonetic" as long as they're consistent.

P.S.: Please Slashdot, add UTF-8 support. I'm not asking for quintuple bucky bits, just a reasonable, mostly backwards-compatible standard that extends functionality significantly.

Re:Typo

[hunterx11]

Why should understanding computers or technology in general give one awesome communication abilities [...] ?

How else are they going to program COBOL or AppleScript?

Re:Typo

[KDR_11k]

Both, actually, depending on context. What does "fuck" mean?

Re:Typo

[dirty]

Because they don't. Both of those spellings are romanizations of the Japanese alphabets (yes plural, there are two phonetic alphabets). A lot of books will also now spell Tu as Tsu. Futon is spelled Huton when romanized because in Japanese there is no F or H sound, the sound is a mix of the two same with L and R and the cause of many jokes about the way native Japanese speakers pronounce English words. Personally I think Japanese is great for spelling. I know very little, but I can spell a great deal if I can hear the word clearly.

Re:Typo

[drinkypoo]

Assuming English is your primary language you really have no excuse, but more importantly the people with the money, power, and influence are mostly pretty smart (whatever else you might say about them) and unless you are otherwise exceptional (athlete, supermodel, stripper) you will never gain their attention except as relates to disdain. In other words, the people most people tend to respect will think they're a fucking moron and will probably go out of their way to avoid them, consciously or sub-. Perfection is not required, but speaking the same language as the other, more intelligent people, goes a long way. If you have a learning disability then you have an excuse, otherwise...

Note that this is written to the general "you".

Blogging/commenting with ridiculously poor spelling and/or grammar is the internet equivalent of giving a commencement speech with your pants down and wearing clown shoes. Some people will still take you seriously. Most of them are certifiably insane or have an IQ that sound more like the number of empty dinner plates at a supermodel convention.

Re:Typo

[FIGJAM]

Re: Why should they?
Spoken and written language is the basic form of human communication. Are any of your excuses justifiable? Apparently, you strive for mediocrity.

Re: IF... THEN... ELSE logic
Maybe if I compared it with "IF... THAN... ELSE" it would make it easier for you to understand.

Re: "sexual matters"
Have you considered the word "fucking" can be used as an intensive? Take your mind out of the gutter, dickhead.

Re: "impossible for everyone to have above average skills"
Geeks are meant to have above average skills. Can you please highlight where I had referred to all average people needing above average skills? Clearly, you did not recognize the fact I was contextually referring to geeks.

Re: Why should others learn things...
First, you should learn there is no space between a question mark and the last word of a sentence. Secondly, imagine a world where language matters little. By your logic, we should all be grunting and living in caves. Maybe you mean we should all know how to talk but not know how to write. What a glorious vision you have.

Re:Typo

[C0rinthian]

It is my personal belief that the internet will be the death of the english language as we know it.

One look at the average AIM user's communication skills, or a quick browse through most internet forums will make this obvious. Capitalization, punctuation, even consistent spelling is going out the window. I think the worst I have seen is that 'lol' and smilies are the only punctuation some people use. (Except for the exclamation point. People use that excessively)

i mean seriously people just write these huge blocks of text stream of conciousness and never try to make sense of their thoughts you know lol its totally impossible for anyone to read but they dont care and call you a grammar nazi if you mention how incomprehensable their attempts at communication are rofl so they just keep doing it and wonder why noone takes them seriously and why the hell does anyone abbreviate 'any' to 'ne' like that one extra letter is so much damn work!!!!1!!1

It frightens me when I get corporate emails that have such words as "ne" and "ur" in them. I'm not kidding. This stuff deeply worries me.

Re:Typo

[ultranova]

Re: Why should they?
Spoken and written language is the basic form of human communication. Are any of your excuses justifiable? Apparently, you strive for mediocrity.

What excuses would those be ? You stated that geeks should have above average communication skills, I asked reasons for why this would be so. Never did I offer any excuses, nor are any neccessary, since, AFAIK, no one needs your approval for their communication skills or lack of them.

Re: IF... THEN... ELSE logic
Maybe if I compared it with "IF... THAN... ELSE" it would make it easier for you to understand.

Human languages, unlike computer ones, don't have a rigid syntax. This is because human beings have rather powerfull error correction algorithms in their disposal, so simple spelling errors don't usually form significant problems in understanding the meaning of the writing in question.

On the other hand, computers do not have such algorithms (known sometimes as "common sense"), and thus need every dot at exactly the right place; a computer trying to guess what its operator meant will likely guess wrong.

So no, comparison between these two different languages, made for completely different purposes, does not make things any clearer. Unless, of course, you are trying to imply that you are no better than a computer at parsing text ?

Re: "sexual matters"
Have you considered the word "fucking" can be used as an intensive?

Certainly. Have you considered that literalism can be used to annoy arrogant creeps like you ?

Take your mind out of the gutter, dickhead.

Um, I'm not the one using expletives derived from sexual acts and organs here.

Re: "impossible for everyone to have above average skills"
Geeks are meant to have above average skills.

From http://dictionary.reference.com/search?q=geek:

1. a. A person regarded as foolish, inept, or clumsy.
   b. A person who is single-minded or accomplished in scientific or technical pursuits but is felt to be socially inept.
2. A carnival performer whose show consists of bizarre acts, such as biting the head off a live chicken.

Now, unless you meant the carnival performer with "geek", I'd say my point (which was that geeks don't need to have above average communication skills to be geeks) is proven, since communication is propably the most important social skill.

Can you please highlight where I had referred to all average people needing above average skills?

Certainly. You established the context of referring to all the people when you said (emphasis mine):

"I am going to take this opportunity to say to all the people who make mistakes like this "...

After this, you continued with:

"If you cannot spell very well and/or often cannot use correct grammar then you are just fucking ignorant and you need to recognize this fact. You lack basic knowledge and for it you should have "I am ignorant" stamped on your forehead daily until you at least achieve above average communication abilities."

Since you had specified "all the people who make mistakes like this" (which, since all people make mistakes sometimes, equals "all people") as context earlier, your second sentence referred to the same group. The end result of this was that you said that all the people need above average communication abilities, to avoid a stamp saying "I am ignorant" on their forehead, restamped daily.

Between these two quotes, you also said:

"Hemos is meant to be amongst the king of geeks, operating Slashdot and all. Maybe Slashdot should be News For Dumbass Illiterate Nerds instead."

However, there doesn't seem to be any obvious

Re:Typo

[scottj]

I am a native speaker of English. English is the only language that I speak. I was taught the difference between then and than. I think most people here in the US were taught that difference. I also think that at least 50% of what most people in the US were taught as children was either quickly forgotten or ignored.

And?

[Darren Winsper]

Using DirectX to create a horribly non-standard and ugly interface? Meh, it's been done before.

Re:And?

[essreenim]

I've an idea:

DirectX9 used to render: a new amazing 1d universe!!

Re:And?

[shaka999]

Guess its all in how you define "standard."

Looking at dictionary.com the most relevant definition

"Something, such as a practice or a product, that is widely recognized or employed, especially because of its excellence"

Now the excellence part is in doubt but I dont' think any sane person will argue that windows DirectX 9 isn't "widely employed." Because most people have access to this platform it is very standard and thus useful to the majority of people.

Oh, or were you just ranting because you can't use it on your Mac or Linux machine?

Re:And?

[T-Ranger]

He means that it isnt using MFC controls, frames.... Standard Windows widgets.

Re:And?

[Max Romantschuk]

Using DirectX to create a horribly non-standard and ugly interface? Meh, it's been done before.

RTFA. DirectX is also used to render video effects in real time. That's innovation, despite the fact that the interface may well be sub-optimal.

Re:And?

[Durandal64]

Apple's been doing the same thing with OpenGL for quite a while now.

Re:And?

[Darren Winsper]

"RTFA."
I did. I wasn't impressed.

"DirectX is also used to render video effects in real time."
And? People have been using videos as textures for years. People have been writing pixel shaders that operate on textures for years. It's rather obvious that you could combine the two.

Perhaps you will come to this conclusion?

[delta_avi_delta]

Perhaps there is more use for these cards even at a consumer level?"

Is it just me, or has almost every second story today had some kind of spurious leading comment tagged on to the end?

Give me facts dammit, I can make my own opinions from there!

Re:Perhaps you will come to this conclusion?

[EpsCylonB]

Slashdot isn't a news site, its more of an aggregator of tech related stories, as such limited analysis by the submitter is appropriate.

Re:Perhaps you will come to this conclusion?

[slapout]

Didn't you know? The slashdot editors are secretly working for a hard drive maker. The more comments a story gets, the more hds their parent company has to buy!

Yeah, maybe

[DOS-5]

Most 3D APIs are good for doing this sort of thing. I think OpenGL would have been a better choice for it though. Although OpenGL is a little behind Direct3D in terms of standard features I still find that it's not only easier to use but generally much faster. I just hate being forced to put absolutely everything into a stupid primitive list.

Oh and I've found that Direct3D has major issues with modifying and accessing texture data directly, which would be necessary for something like this.

Re:Yeah, maybe

[bersl2]

Although OpenGL is a little behind Direct3D in terms of standard features

That's news to me. You got anything to back this assertion of yours?

Re:Yeah, maybe

[DOS-5]

Until the newer version of OpenGL is available you're going to find most of the extra functionality through OpenGL extensions rather than the API itself, and any extension is not always guaranteed to be there. At least if they had some way to check for capability through the API (like Direct3D does) it would have been made easier on developers.

Re:Yeah, maybe

[pmjordan]

I see a contradition here, emphasis added:
...I still find that it's not only easier to use but generally much faster.
I just hate being forced to put absolutely everything into a stupid primitive list.

Care to explain how OpenGL immediate mode is faster than Direct3D Vertex Buffers?

Disclaimer note: I've used both, and am currently using OpenGL due to its open nature.

Light-hearted flamebait: In addition, if you think OpenGL is easier to use than Direct3D you clearly haven't understood OOP or C++.

~phil

Re:Yeah, maybe

[Anonymous Coward]

The capacity checking in OpenGL is .. the API which checks for extensions being present. Surely.

Re:Yeah, maybe

[Mr.Zong]

Managed Code in DirectX 9 is like chocalate pie. Or warm bread. It makes my tummy all warm and everything has a nice fuzzy glow.

The DirectX sdk is my favorite free MS product. I also belive its one of the rare occasions that somebody got OOP to do what it was meant to do, and actually be reusable.

Re:Yeah, maybe

[Lumpy]

but it's worthless for video editing.

I had a card that was great for this, it was called an "ice" card. and coupled with Adobe AfterEffects it would perform effects processing faster than the dual processor P-III could because it used a processor that was designed for video effects.

99% of video editing on a PC is not intensive to the point that the video card is able to help. and a 3d video card will certianly not be able to help except for some of the more cheezy effects that scream "amateur home video!"

Adobe Premiere is the bottom rung of useable video editing. and this does not even touch that nor gives any real advantages to the video editing process.

Adding eye candy to the creative process does not make things easier for the operator.

When they can make my FX5700 do the work of fade between two video streams and render back to a file, (something that a video card is NOT designed to do)then I'll be impressed.

Re:Yeah, maybe

[jinzumkei]

D3D is harder to use?

Did you miss school the day they taught OOP?

Re:Yeah, maybe

[mikael]

At least if they had some way to check for capability through the API (like Direct3D does) it would have been made easier on developers.

Have you tried 'glGetExtensionString( GL_EXTENSIONS)'?
This will return a string listing the extensions supported by the current context.
You can also use GL_RENDERER, GL_VERSION and GL_VENDOR to
check that hardware acceleration is supported, and to determine the driver
version and vendor.

Re:Yeah, maybe

[pmjordan]

Meh, I'll bite. Direct3D uses OOP to the extent that it's actually useful. The various graphics resources (geometry buffers, textures, shaders, etc.) are logically represented as objects. The various different classes of textures are implemented as derived classes that inherit from a 'base texture' class. I suppose instead of calling Direct3DDevice::SetTexture(int, Direct3DBaseTexture*) they could have made SetTexture() a polymorph member function of textures, but either way works practically equally well, and the texture stage index makes more sense in the context of a device than a texture directly.

Contrast this with doing the same thing in OpenGL:
1. (if necessary) switch to the correct OpenGL context.
2. (if necessary) switch to the correct texture stage.
3. Bind the texture.

To me, that's an obvious win for the OOP (Direct3D) version, but there you go. OK, so (1) will only be necessary in very special circumstances, but (2) is practically always necessary, avoiding it tends to be more work than not.

The point of a production library is not to demonstrate design patterns, but to apply the most appropriate techniques to whatever it is abstracting. If you consider a library's API incomplete or inferior just because it doesn't utilise polymorph multiple inheritance from virtual template base classes, you might want to consider a career as a computer science professor, a few of them will actually agree with you.

I'm also not really sure what C++ features you're missing. OK, so instead of exception handling they use return values, which I personally consider more appropriate in this case. Feel free to disagree on that point.

You still failed to miss the point I was trying to make: OpenGL's could benefit a lot from a better API. As it is, all the newer features are added-on hacks that add obfuscation by introducing statefulness at the API(!) level.

by Anonymous Coward

Good work stuffing your foot in your mouth.

*chuckles* Ah, kids these days.

~phil

Re:Yeah, maybe

[DarkMantle]

Well... My question is, If D3D is go great, why is MS not supporting it in longhorn? Instead they are using the OpenGL based Avalon architecture. Hmm... seems even M$ saw the error in there ways this time

Re:Yeah, maybe

[Raistlin99]

That link says nothing about OpenGL. In fact if you look around you find it say that Avalon is based on DirectX. Nice try but I read links like that.

Re:Yeah, maybe

[Raistlin99]

Forgot the link http://msdn.microsoft.com/Longhorn/understanding/p illars/avalon/avnov04ctp/default.aspx

Re:Yeah, maybe

[Mac Degger]

"of fade between two video streams and render back to a file,"

Oh, kind of like how you do a textureblend between two or more textures using ramping opacity and render to any surface?

Had you read the FA, you would have noticed that, hey! Lo and behold: 3DEdit does exactly what you want: with a 5700 (absolute minimum req), it does the work of fading "between two video streams and render back to a file" just shy of real-time.

So: be impressed.

Re:Yeah, maybe

[TrancePhreak]

You don't seem to know much about what you're talking about. For a video, you would definitely be using the dynamic pool which would place the texture in AGP memory. Fast modify and fast display. Such things have been the case for years, so obviously you are thinking of D3D3 or something.

Re:Yeah, maybe

[TrancePhreak]

Please re-read my post. I didn't say it was in video. It's not in software either. It's in AGP. You know, the stuff used for dynamic data that the video card can dma into, and you can access quickly with the CPU.

Re:Yeah, maybe

[DarkMantle]

Odd.... since ATI is upset that there current line of cards (including the x series) won't be compatable, but nVidias OGL based cards will be.... don't make sense unless it's OGL based. (can't find links to hardware sites where I saw this before)

"Than"

[HarveyBirdman]

Geezus... A typo is fine in the body of an article, but at least proof your headlines, people.

Re:"Than"

[punkass]

Touche.

In the same boat

[KirkH]

We're actually considering going this route with an app here at work. It's a GUI-intensive app that spends most of its time drawing to the screen using custom MFC controls. It's fast enough most of the time but begins bogging down when we try to push through too much data.

Anyone have any experience going the DirectX route? Would it possibly be faster than what we're doing today? I assumed from my experience with the interfaces on games (Unreal Tournament, etc) that DX would be slower.

Re:In the same boat

[chris09876]

One of the issues (especially if you're planning on deploying the app) is that your customers will need graphics cards. I know it sounds obvious, but a company I worked for about 2 years ago did the same thing. Most of their customer machines didn't have current graphics cards, and they were unwilling to buy them (which is strange, seeing as the software was selling for like $10,000). In the end, we ended up just giving away an nvidia card with every license. It worked, but that's an issue you should probably be aware of. We weren't, and it caught us off guard.

Re:In the same boat

[argent]

Use OpenGL, it's a better API, works on more video cards, and works on more than just Windows boxes.

Re:In the same boat

[KirkH]

It's an in-house app developed for company use only. We don't deploy to off-site customers, so we have some control over what hardware it's run on. Current minimum gfx card is a GeForce4MX. Others in the mix are Radeon 9700s and GeForce 5900s.

What was your experience with your apps performance compared to regular 2D?

Re:In the same boat

[arkanes]

Fix your crappy rendering algorithms :p Seriously, algoithmic problems are probably your biggest enemy. Especially if you're basing everything off of MFC, which doesn't lend itself to optimized drawning. I can't possible emphasize how much you shouldn't re-produce the entire UI in DirectX.

Have you actually benchmarked and you're sure it's drawing thats the problem? The only time I've ever seen this has been in very complicated 2d visualization and graphing applications (stuff like MRE renders). Every other case where the "GUI was slow" was really a combination of poor drawing code (one application I saw re-painted 5 times for every data update) and poor algorithm choice/implementation. Switching to DX might highlight some of those problems for you.

Re:In the same boat

[KirkH]

Have you actually benchmarked and you're sure it's drawing thats the problem?

Yes and yes. You would not believe the amount of real-time data we have to display...while polling for more data on six sockets...while remaining responsive to extremely active users. The one call that takes the most time is DrawText(). Most of the time the app is fine, it's just when the data really floods in that we get overwhelmed. But that also happens to be the most critical time for our users.

We have profiled and optimized. There are a few more non-trivial optimizations I've thought of, but they'll take some time to implement.

I can't possible emphasize how much you shouldn't re-produce the entire UI in DirectX.

This was what my gut was telling me, but I really didn't have enough experience with DX to back it up.

Re:In the same boat

[johannesg]

The one call that takes the most time is DrawText(). Most of the time the app is fine, it's just when the data really floods in that we get overwhelmed.

You should draw less, then. On a reasonably modern machine DrawText() can fill the screen in a frame, so if it is slow you are either:

- Drawing stuff that is never actually displayed because it will be gone before the frame is even refreshed.

- Drawing stuff that isn't even on the screen in the first place (either off the side, or hiding underneath other windows).

- Drawing multiple items whenever a single item changes, which is obviously highly inefficient.

By the way, I obviously have no idea what your code looks like but I found that the conversion for doubles in sprintf() is actually the slowest part of my code, instead of the drawing of the converted string. Without that conversion I can convert and write out 180,000 values per second, with the conversion no more than 50,000.

Re:In the same boat

[arkanes]

Have to agree here. DrawText() is a very slow call because it tries to produce high quality text. If you're willing to dispense with stuff like kerning, anti-aliasing and, Unicode support, you can roll your own text output in just a few lines of code that's orders of magnitude faster. However, I strongly suspect that your problem is too many calls to DrawText() rather than DrawText() being too slow.

Sys requirements...

[gandell]

"Nvidia GeForce FX 5700 is the bare minimum."

I guess I was expecting a little more for such high video card requirements. I just don't see the benefits of this as compared to other video editing software. Anyone care to enlighten me on what this can do that others cannot?

Re:Sys requirements...

[0123456]

"Anyone care to enlighten me on what this can do that others cannot?"

You can add fancy 3D effects to your home movies, just like a real wedding video editor :) !

Re:Sys requirements...

[fudgefactor7]

Near as I can tell it's a two-fold requirement that makes DVDoctor need the 5700... (1) CineFX Engine version 2.0 and (2) Shader Model 2.0b or better, both those are only available in the 5700 or newer Vnidia cards (the 5600, for instance, has CineFX 1.0).

Re:Sys requirements...

[SenseiLeNoir]

Currently, Transitions use a lot of software rendering techniques for generating effects such as transparency, 3D flyovers, and composition.

Traditionally, using software rendering, a simple 10 minuite clip can take 1 hour to render. Just over the weekend, I created a 8 minuite "moving slideshow" video clip from Still photos, and titles, which consited of photos gently moving in and out, and cross fading, and titles being added to it. A very simple composition task, yet it took my Athlon 2500+ over 40 minuites to render frame by frame into High Quality MPEG2 for DVD using software rendering.

Few years ago, it was suggested that maybe a 3D card can be used to assist in that, so the blunt of the rendering was done on a 3D card, and then frame by frame captured from the frame buffer to create a final AVI/MPEG of the composition.

The presumption was that frames of the video or Stills can be used as Textures, and the power of the Graphics card to render it all

Also it can be used for Real-Time composition of effects, as you can hook up a video recorder to the output, and directly record onto tape.

This technology was used extensivly in the Matrox RT2000 and beyond. The RT2000 was a professional video editing suite which consists of a Modified Matrox G400 graphics card (called the G400 flex) and a RT2000 video in/out card, which did realtime DV/MPEG encoding/decoding, and had firewire/analogue connections.

The RT system used the Matrox G400 Flex to perform the realtime compositing and rendering, and is powerfull enough to do the same effects in realtime, then send back to the RT card to directly send to DV tape, or MPEG2 file.

Then in 2000, ATi shown a proof of concept software using a normal Radeon card to render two video sequences onto a spinning cube in realtime, which was really stunning to look at.

So i assume this is further development is the realisation of this proof of concept.

As for WHY all this is nessasary, well for professional video editors, it gives the ability to have instant high quality previews, and fast rendering, which saves so much time, hence increases productivity.

Re:Sys requirements...

[homer_ca]

"it uses Dorext X for the effect PREVIEWS only. it will not use the video card to speed up rendering time."

That makes sense since AGP is essentially a one way interface. It's fast for sending data to the video card, but slow as hell if you want to read a rendered frame off the video RAM. Maybe PCI express will change that.

Re:Sys requirements...

[Hamled]

I'm quite a bit surprised that this hasn't been done already. Academics and scientists are already using modern GPUs to speed up scientific applications and other non-visual calculations by enormous amounts.
If they can jump to the idea of doing non-visual work on the GPU, it's a surprise they didn't think of rendering video with it...

Re:Sys requirements...

[SenseiLeNoir]

But it IS being done already.. the RT2000 i mentioned in my post earlier.....

CoreImage

[cjwl]

Check out Apple's upcoming CoreImage system if you're interested in uses of a video card for things other than video games:

http://www.apple.com/macosx/tiger/coreimage.html/

Re:CoreImage

[shaka999]

Whew, I was starting to doubt the power of slashdot. I had read 5 posts without a mac reference.

Re:CoreImage

[flabbergast]

If you really want to get away from video games you should also check out GPGPU. Its a site dedicated to general processing on GPUs. There is also a lot of research done in this area, such as a DB run on a graphics card as well as a lot computational geometry problems. Also, Stanford has developed a more straightforward C-like language to do GPGPU. Pretty cool stuff.

Re:CoreImage

[Blakey Rat]

Yeah, but in this case it's actually relevant. The post is about using video cards to do something more than games, and Apple's been doing that for, what, 4 years now.

Re:CoreImage

[NardofDoom]

Where do you think Microsoft got the source code?

Re:CoreImage

[Tim Browse]

Meh. Apple and Microsoft have been using video cards for 3D that isn't games related for a lot longer than 4 years.

Wake up!

[NYhXc]

WOW, that's the ugliest interface I ever saw at a video editing program!

Re:Wake up!

[Nogami_Saeko]

I think the final page of the review said it best "Tenomichi's 3D Edit is a curious beast, and a very British kind of editing application."

Lots of style, not so much substance. I'll stick with Avid.

N.

Re:Wake up!

[SenseiLeNoir]

Being British, I didnt understand how the link was made that:

"very British kind of editing application" =
"Lots of style and not so much substance"

OpenGL

[Glock27]

OpenGL stands for "Open Graphics Library". Microsoft did Direct3D as a way to lock 3D content into the Microsoft platform. Only id Software's stance on OpenGL for gaming saved the day. At the time, even Microsoft admitted that OpenGL was more appropriate for "professional" 3D apps like CAD. Now I'm sure they'd like to lure developers into using Direct3D for professional apps, just as this developer has done.

Developers should use OpenGL in preference to Direct3D if they want cross-platform compatibility, or simply to use a better API. One way to do this that provides a lot of flexibility is to choose a high-level scene graph library that uses OpenGL or Direct3D at a low level.

OpenGL apps run on Windows, MacOS and Linux. OpenGL has always been "For More Than Just Gamers".

Re:OpenGL

[Anonymous Coward]

The benefit to DirectX is that it does so much for the developer, it takes care of sound/input/networking. All of which aren't fantastically easy to recreate, unfortunately.

When a DirectX game gets ported to OS X or any other platform you'll often find that the multiplayer is limited to the platform you're using.

Perhaps someone can suggest some other libraries/frameworks for input/sound/networking. There's GLUT for input, but it's pretty simple so it might not work for everyone.

Re:OpenGL

[91degrees]

For the rest of DirectX, there's DirectX. The control and sound parts of a game are such a small portion that they should be easy to replace for other systems.

Re:OpenGL

[Glock27]

The benefit to DirectX is that it does so much for the developer, it takes care of sound/input/networking. All of which aren't fantastically easy to recreate, unfortunately.

They also aren't fantastically hard to recreate. There are plenty of third-party cross platform sound, input and networking libraries.

Regardless, in this case we're discussing a graphical application.

When a DirectX game gets ported to OS X or any other platform you'll often find that the multiplayer is limited to the platform you're using.

Or not. World of Warcraft comes to mind, for instance.

Perhaps someone can suggest some other libraries/frameworks for input/sound/networking. There's GLUT for input, but it's pretty simple so it might not work for everyone.

OpenML and OpenAL are two possibilities. Another thought might be to use a modern programming environment like Java, and use its abstracted sound, input and networking along with OpenGL. You could always write your own wrapper library in your language of choice as well. You could also use a third-party cross platform game engine like the Torque engine that supports OpenGL.

Here's a good link for those that'd like to learn more about OpenGL in general.

Re:OpenGL

[robocrop]

Filtering through the "open source" - and therefore anything with the word "open" in it - fanatacism:

Microsoft did Direct3D as a way to lock 3D content into the Microsoft platform. Only id Software's stance on OpenGL for gaming saved the day.

First, while I certainly won't defend MS's business practices, it's more than a bit short-sighted to say MS only created D3D to "lock" developers onto Windows. MS wanted to grow Windows as a gaming platform (beyond crap like WinG) so they made their own accelerated API. One could certainly argue that the existence of D3D has pushed OpenGL to stay modern and competitive, so has been twofold beneficial.

Second, where does this nonsense about id saving OpenGL come from? OpenGL has always had plenty of users - as you mention it is consistently used in CAD tools. There is no valid reason, other than fanboyism, to believe that OpenGL would have died if Carmack hadn't used it in his games. And there are other people who have used OpenGL to make their games work on multiple platforms (which, by the way, is usually a very hard sell to your publisher).

Developers should use OpenGL in preference to Direct3D if they want cross-platform compatibility, or simply to use a better API

What a smooth, effortless transition from fact to propaganda.

As a person who has written numerous game engines, I can attest to the fact that OpenGL is - for me - not a better API. It is convoluted and over-complicated by the very fact that it is an "open" standard, decided upon by a committee of people who wouldn't know good design if it bit them in the rear.

Just the fact that most of the features easily found in D3D cannot be accessed in OpenGL, years after their invention, without the use of custom extensions is enough to throw me off the API.

If you want to use OpenGL that's your right. But it is not an objectively better API any more than Mac is objectively better than PC, Linux objectively better than Windows, or any of the other /. propaganda that we consistently read.

Re:OpenGL

[gstoddart]

Developers should use OpenGL in preference to Direct3D if they want cross-platform compatibility, or simply to use a better API. One way to do this that provides a lot of flexibility is to choose a high-level scene graph library that uses OpenGL or Direct3D at a low level.

Well, sadly, needing to support a Microsoft platform often makes people decide to go a primarily Windows route.

We'd all love to see our companies not lock into Microsoft's non-standards, but the unfortunate reality, is management and reality frequently make it a non-choice.

In many cases, it's just easier to do it on a Microsoft API 'cuz that's the skills you have in house or management decrees it must be so.

Re:OpenGL

[Darren Winsper]

My point is that when comparing DirectX with OpenGL, you don't get the complete picture. I'm not arguing against using DirectX where appropriate, it's just that complaining OpenGL doesn't handle sound is a bit silly.

Re:OpenGL

[GameMaster]

As someone who has actually programmed with both APIs (including multiple earlier versions of DirectX) I have to take exception to the blanket statement "or simply to use a better API". While this may have been true with previous versions of DirectX (especially pre-DirectX 7) it really isn't true anymore. Microsoft has learned, through painful trial and error, how to create an easy to use graphics API. At this point, which one is better has really become a matter of personal taste more than anything else. In fact, on most graphics programming forums asking which one is better or putting up threads that debate the issue is strictly forbidden because it doesn't serve a practical purpose anymore. The only thing it can lead to is a flame war.

As for cross-platform compatibility, the simple fact is that OpenGL has DirectX beaten in that regard. Things like WineX have helped to neutralizes this to an extent, but for the most part, OpenGL really does beat DirectX here. There is a corollary to this thought. For much of the purpose that DirectX was developed for in the first place, namely pro-level game development, cross-platform compatibility really doesn't matter. Like any business, what commercial game companies care about is making as much profit as possible from their product. As far as the game industry is concerned, Linux has been proven to be a non-existent market. To use your own example of Id software, very shortly after the game's release (and to this day in some places that still have it in stock) you could pick up a copy of the tin box "Linux" version of Quake III: Arena for almost nothing because very few people were interested in buying it (As a side note, supposedly you can pick up this copy and download the windows executables needed to run the game on a windows box. Though, it's been so long since the game came out that even the windows boxed version of the game might be that cheap now). In the end, so few people bought the Linux version of Q3:A that they didn't even make back the cost of producing it.

At some point in the future this might change and Linux might become a respectable game market but by that time any game made now or in the near future will have been off store shelves for a while so present day game company don't really care. A market that Linux can claim growing relevancy is in professional tools, but then this was never the stated purpose of DirectX and any pretenses that Microsoft might make to improving the use of DirectX in the world of pro-level development apps is just them taking advantage wherever they can (Something any good business would do). How realistic those pretenses are is debatable.

-GameMaster

Re:OpenGL

[jinzumkei]

wtf "id Sotware's stance on OpenGL for gaming saved the day" for who? OpenGL did not NEED saving. There are plenty of OpenGL developers who used openGL. Graphics isn't all about gaming, son. Just about the entire scientific visualization community uses openGL.

Re:OpenGL

[Anonymous Coward]

Good luck finding a modern video card with OpenGL drivers in a world where gaming didn't make it profitable to produce an OpenGL implementation.

Re:OpenGL

[Enrico Pulatzo]

I think the parent was simply speaking from the (quite common) point of view that computers are for little more than video games. Hence, any software package, platform, or tool is immediately viewed in comparison to gaming, and often judged in that respect.

Re:OpenGL

[archen]

Second, where does this nonsense about id saving OpenGL come from?

Not to disagree with you, but I wonder if graphics card companies (ATI and Nvidia anyway) would even bother supporting OpenGL if there wasn't even a single game that required it? I doubt OpenGL would have died, but I wouldn't be surprised if OpenGL would have become a 'novelty' thing to suport more for nitch video cards. And before anyone pipes in with "but Apple uses OpenGL", I'd point out that ID had their foot in the door for quite a while before Apple moved to Unix. I'm not sure how many other game companies use OpenGL (because I haven't seen any myself), but I doubt they would have been enough by themselves.

Re:OpenGL

[Funk_dat69]

OpenGL is - for me - not a better API. It is convoluted and over-complicated by the very fact that it is an "open" standard, decided upon by a committee of people who wouldn't know good design if it bit them in the rear.

What a smooth, effortless transition from opinion to propaganda.

most of the features easily found in D3D cannot be accessed in OpenGL, years after their invention, without the use of custom extensions is enough to throw me off the API.

I can see why people use D3D for a game on a windows machine. The new features appear sooner so you can use them sooner. That's one of the benefits of D3D. But, of course, that's not the whole story. For a game, standards and reliability don't matter.

However, many companies prefer something mature, open, and standard before undertaking the huge task of moving their massive applications onto the new features. It's very important that the ARB get it right and this takes time.

Take the new Framebuffer_object extension for example, it took a lot of time and compromise to get it completed, all the while D3D had similar ability to access textures/buffers, but the final FBO design, IMO, is top notch all around.

Also, rather resent the comment about GL developers not knowing good design. Tell that to Mark Kilgard, Allen Akin, or Michael Gold. I'm sure they could explain any problems that you might have understanding the GL Spec.

Re:OpenGL

[Anonymous Coward]

The popular and affordable consumer-level graphics cards 'supported' OpenGL to the extent that they ran the functions Quake used well. Maya has gone so far as to quit using particular features ("menu plane") because cards didn't bother to support them because Id didn't use them.

Why are you so quick to discount their impact? OpenGL would not be readily available and affordable if it wasn't for id software.

Re:OpenGL

[MnkyKnifeFghtr]

Wow, you must have spent a long time loading up that reply.

OpenGL lags far, far behind Direct3D with regard to simple things like multiple render targets, and pbuffers are still a WGL extenstion. The OpenGL 2.0 spec was over due by more than a year. They finally got on board with vertex/index buffer objects last year, and guess who's interface it was modeled after. (DirectX)

They want to "lure" developers into using Direct3D? Developers work with Direct3D when they want an API that is actually decent to work with, has a good CAPS interface, actually supports all of the features of modern cards, and is constantly being upgraded. If the ARB moved any slower we'd still be doing software T&L.

Seriously, put down the "open-pipe".

Re:OpenGL

[jinzumkei]

You poor soul, Never heard of SGI huh?

Re:OpenGL

[Schnapple]

Only id Software's stance on OpenGL for gaming saved the day

There are only two reasons OpenGL is used in gaming at all:

1. id Software's games and any game they licensed the engine for. This does include quite a few games, but far from the majority. Also, Carmack has stated that frustration with OpenGL 2.0 was the closest to going to Direct3D he's ever been.

2. Games which are concerned with cross-platform compatibility. Consoles are a different kettle of fish, but on the "computer" side of the fence, there's not many games which concern themselves with this. If your game is not going to be released on a console, you can still target 95% of the market and stay on Windows only.

You speak of Direct3D as if it was an evil scheme by Microsoft to destroy non-Windows gaming, but why would Microsoft put out an API on platforms they're not trying to sell? Could Microsoft botch OGL 2.0 to the point where even hardened developers despise it?

Re:OpenGL

[IamTheRealMike]

First, while I certainly won't defend MS's business practices, it's more than a bit short-sighted to say MS only created D3D to "lock" developers onto Windows. MS wanted to grow Windows as a gaming platform (beyond crap like WinG) so they made their own accelerated API.

Why did that mean replacing OpenGL, which at the time game developers like Carmack were perfectly happy with. In fact Carmack and others roasted the first versions of Direct3D, and petitioned Microsoft to release the OpenGL MCD.

GameDev magazine reviewed the two APIs in the early days and concluded Microsoft should just discontinue Direct3D altogether. OpenGL was clearly and unambiguously the better technology and it's widely regarded that it wasn't until DirectX 8 that it caught up. Yes, after Microsoft squashed them out of the market it stagnated: what a surprise. Just like Netscape Navigator stagnated after Microsoft killed the revenue stream from it.

One could certainly argue that the existence of D3D has pushed OpenGL to stay modern and competitive, so has been twofold beneficial.

Microsoft dropped OpenGL support entirely as soon as Direct3D started development. It didn't ship with Windows 95 at all originally. They went out of their way to kill OpenGL, so quite how that was "beneficial" to it by providing competition I do not know. Competition would have been if both OpenGL and Direct3D were fully supported on Windows, but they were not. It was never a fair fight, so to reach this conclusion is ridiculous.

Re:OpenGL

[robocrop]

Why did that mean replacing OpenGL, which at the time game developers like Carmack were perfectly happy with

As I said, I won't attempt to defend MS's business practices. But so you know, MS didn't do anything explicitly to kill OpenGL. They simply stopped directly supporting it to focus on their own API. They developed an alternative. This itself is not a destructive action, and is in fact the very heart of a competitive marketplace.

GameDev magazine reviewed the two APIs in the early days and concluded Microsoft should just discontinue Direct3D altogether

I believe at the time GameDev also ran articles on creating "next generation voxel engines". In other news, Bill Gates claimed nobody would need more than 640K of memory; IBM thought PCs were "a fad", etc. Oh, and recently we were told that Firefox would never be a competitor in the browser market. Don't believe everything you hear. D3D went through some extremely nasty iterations, but has emerged as a very stable and useful API (as has most of the rest of DirectX).

OpenGL was clearly and unambiguously the better technology and it's widely regarded that it wasn't until DirectX 8 that it caught up.

What a hyperbolic statement. OpenGL was better than a new, virtually untested technology? Shock.

Yes, after Microsoft squashed them out of the market it stagnated: what a surprise.

A demonstrably false statement: OpenGL is alive and thriving today, so nobody "squashed them out of the market". And OpenGL has mostly stagnated because of the nature of open standards and the weight of having to support so many disparate fields. Having used both APIs I, personally, find D3D to be easier to manage. But the choice is still there, so there's no need to exaggerate.

Microsoft dropped OpenGL support entirely as soon as Direct3D started development. It didn't ship with Windows 95 at all originally. They went out of their way to kill OpenGL, so quite how that was "beneficial" to it by providing competition I do not know

More hyperbole and propaganda. MS simply stopped actively supporting it. One can easily download the OpenGL libraries (just as one normally has to do for other operating systems), and GL apps still compile in VC.

Competition would have been if both OpenGL and Direct3D were fully supported on Windows, but they were not. It was never a fair fight, so to reach this conclusion is ridiculous

Really? So, by your argument, OpenGL is not competitive because D3D "isn't supported" on SGI workstations.

Odd how people like you are for competition, unless it's competition by someone you don't like. D3D is a very useful API, and has been beneficial for the game development industry.

Re:OpenGL

[Funk_dat69]

MS didn't do anything explicitly to kill OpenGL.

Right...try looking up Fahrenheit. MS had no intention of persuing it. It was, in essence, a thinly veiled attack to kill OGL. Almost succeeded too.

And OpenGL has mostly stagnated because of the nature of open standards and the weight of having to support so many disparate fields.

Stop spreading this crap, seriously. Yes, the ARB moves slowly, but that does not equate to stagnation. There is a whole world outisde of windows gaming, y'know.

One can easily download the OpenGL libraries (just as one normally has to do for other operating systems), and GL apps still compile in VC.

True, although it's essentially a big driver hack getting all the new GL stuff supported. Nothing unethical there by MS, though, for once - business is business.

Re:OpenGL

[Glock27]

What a smooth, effortless transition from fact to propaganda.

Gee, I guess I should have added an "IMNSHO".

As a person who has written numerous game engines, I can attest to the fact that OpenGL is - for me - not a better API. It is convoluted and over-complicated by the very fact that it is an "open" standard, decided upon by a committee of people who wouldn't know good design if it bit them in the rear.

I certainly disagree with this statement. I suppose you could make an argument that a C API vs. COM is a matter of taste. I think the simplicity of the C API (which is also slightly more efficient) is elegant.

Just the fact that most of the features easily found in D3D cannot be accessed in OpenGL, years after their invention, without the use of custom extensions is enough to throw me off the API.

The counterpoint to this is that hardware companies using OpenGL can expose new functionality without ARB approval. That has many beneficial effects, including the ability of game developers to use said new features in production games before the ARB gets around to deciding on the "best" official implementation.

DirectX developers, on the other hand, get to wait for Microsoft to release a new DirectX version. Guess which I find preferable?

If you want to use OpenGL that's your right. But it is not an objectively better API any more than Mac is objectively better than PC, Linux objectively better than Windows, or any of the other /. propaganda that we consistently read.

I'd say there are excellent cases to be made that "Mac is better than PC" and "Linux is better than Windows". However, the real problem is that "better" is a vague term. I should have been more specific. Let's try again:

OpenGL is a more elegant, leaner, and better designed API than Direct3D. Its cross-platform availability is simply icing on the cake. OpenGL will ultimately bury Direct3D. All IMNSHO, of course.

Happy now? :-)

Re:OpenGL

[Glock27]

World of Warcraft doesn't use DirectX for it's networking, nor does Unreal Tournament 2004. They use DirectX for pretty much everything else on Windows and OpenGL and whatnot on OS X or Linux.

What I was referring to was games like Dungeon Siege, Age of Mythology, Rise of Nations, which use DirectX for pretty much everything. Not surprising as they are Microsoft Games, but the OS X ports have no OS X -> PC networking, it's PC -> PC, OS X -> OS X.

That sounds like excellent justification for not using DirectX for networking. Thanks for the information.

Re:OpenGL

[drinkypoo]

Microsoft doesn't even do the opengl drivers, except the software renderer. They come with the video card driver.

Has it really gotten to the point

[the_skywise]

That it's faster to render the 3D controls than to do a 2D paint of them? Even the article points out that it restricts the user to specific hardware configurations and shuts out traditionally popular cards for video editing. All for a "purty" interface.

The underlying workings of 3D Edit mean that it is picky what platform it runs on. There's a long list of compatible graphics cards on Tenomichi's website at http://www.tenomichi.com/Compatible.htm. Essentially, a DirectX 9 adapter is required, which currently doesn't include any of Matrox's graphics cards.

Re:Has it really gotten to the point

[Jimmy The Leper]

Well, if your 2d images of controls are in system memory, then yes, it is MUCH faster to just draw a textured quad. If both were kept in video memory, then I suspect the difference is speeds between a straight blit and drawing a textured quad would probably be negligable (compared to a blit from system memory).

I don't really know much about graphics hardware performance except for that unless you're using many fancy pixel shaders the biggest bottleneck on a graphics card is the AGP bus.

Re:Has it really gotten to the point

[the_skywise]

But for 2D interfaces you don't blit the entire screen. Only the portions that change (and the mouse drawing/redrawing is handled in hardware). That's why there's all this built-in functionality in the Windows GDK to handle clipping and redraw areas and such.

For a 3D interface, you're pretty much guaranteed of having to redraw the entire screen. Now that's doable with a 3D card and you have the same performance. But as I pointed out, it knocks off certain other users that have "non-approved" cards or older hardware (which is almost always non-approved after a year or so)

I'm not against 3D rendered interfaces (they're cool for games and fancy displays) but I question the need for such a beast on a video editing tool.

It also breaks UI "law". The interface is no longer consistent with Windows so it requires some effort on the part of the user to learn new interactions. (Though in this techno-savvy society I don't think that's as big of a concern).

Or to put it more simply: It's a whiz bang feature that adds a nice coat of gloss to a product but cuts off a percentage or potential customers by the sheer hardware requirements and/or custom interface.

Re:Has it really gotten to the point

[Jimmy The Leper]

You could have properly clipped drawing in OpenGL, but it would probably be more trouble than it's worth. Too many calls to glViewport or glScissorTest would probably slow things down a bit.

I guess the point would be to use this sort of thing on a platform where hardware can be controlled (Macs) or use it in simple applications entirely for it's "whiz bang" appeal (a video player, game, or other similar apps where GUI rules arn't followed anyways).

I was just pointing out that you could probably get a more 'snappy' GUI by doing things entirely in hardware these days, even with lighting, fog, and rotation. Not that I would want something like that in an OS (although OS X pulls it off nicely), it would just be a distaction after the initial awe wore off.

AA required?

[Geccoman]

From the article: "It's even necessary to turn on anti-aliasing in the graphics card drivers to smooth the on-screen elements of the interface."

I can see this needing a little more horsepower to really run great. I love the effect of AA, but my meager P4 2.4 can't always take the extra processing required. I wish they had tested the program on a lesser machine than a Dual Xeon. =0

Re:AA required?

[maxbang]

Oh god I hope AA isn't required. I'm incapable of editing video without my bitter lager.

Re:AA required?

[SoTuA]

I love the effect of AA, but my meager P4 2.4 can't always take the extra processing required. I wish they had tested the program on a lesser machine than a Dual Xeon. =0

Except that the AA is taxing on your videocard, and it won't matter if you have a dual xeon if your card is a Geforce2 MX200.

Re:AA required?

[cthrall]

If it's necessary to turn on anti-aliasing on your video card, your CPU speed might not matter as much as what video card you have.

Ugly UI, Functional UI

[wcitechnologies]

I think operating systems designers could take a lesson from these guys (the windows have depth!), but I wouldn't suggest the same thing for window style designers (the windows are ugly!).

Re:Ugly UI, Functional UI

[amalcon]

I have a friend who works for Sun Microsystems, and he tells me they're working on a 3D version of JDS (don't get me wrong, a desktop system is the LAST thing I'd want to see Java used for, but that's not the point). Among other things, you can rotate a window, say, 80 degrees and stick it in a corner. It's then still recognizable, but it doesn't take up much space. You can also turn things around and write notes on the back. This is a REAL use of 3D in a graphical shell. Now, if only somebody would take their idea and implement it in C++ (or some other language with a goal other than portability).

Re:Ugly UI, Functional UI

[afd8856]

Looking Glass. Nothing new. Come here more often :)

Re:Ugly UI, Functional UI

[y2dt]

Java Desktop Ssytem (JDS) actually has nothign to do with Java. It's just a rebranded Suse linux running Gnome, integrated with evoluition and everything. It's actually pretty cool.

the name is pretty decieving.

Re:Ugly UI, Functional UI

[radish]

The JDS (Java Desktop System) is not written in Java (isn't it just KDE or something?). The name has been borrowed for the purposes of branding. A dumb idea IMHO, because it leads to people who still have the misconception that Java is slow to believe that the product in question is also slow.

Re:Ugly UI, Functional UI

[dioscaido]

Now, if only somebody would take their idea and implement it in C++ (or some other language with a goal other than portability).

Avalon?

http://msdn.microsoft.com/Longhorn/understanding/p illars/avalon/default.aspx

Re:Ugly UI, Functional UI

[GigsVT]

Misconception?

I am glad to say I finally found a Java app I like, Drawboard. It's a shared whiteboard thingy, and it's really cool.

It's still slow as molassas. When it's running, my whole computer is noticably slowed down. I don't know if it's just some Java reality distortion field that makes you guys think Java isn't slow as hell, or what.

Re:Ugly UI, Functional UI

[bagel2ooo]

I'm not sure, but I feel this (http://www.sun.com/software/looking_glass/) may be what you are talking about. I actually got a compile working (despite my hatred of Java.) It does appear to be kind of sluggish but the idea of it is quite nice. I wasn't able to get it to load anything but from the demos of seen (taken with a grain of salt) it is an idea whose time came a couple of years ago but was never implemented. To anyone who has a spare *n?x machine and a bit of curiosity; I suggest you throw it on there and browse around. I haven't taken a look for 2 months so it's likely polished enough to have a lot of the kinks I didn't like worked out. Just bare in mind it is pre-alpha iirc.

Re:Ugly UI, Functional UI

[illumin8]

I have a friend who works for Sun Microsystems, and he tells me they're working on a 3D version of JDS (don't get me wrong, a desktop system is the LAST thing I'd want to see Java used for, but that's not the point). Among other things, you can rotate a window, say, 80 degrees and stick it in a corner. It's then still recognizable, but it doesn't take up much space. You can also turn things around and write notes on the back. This is a REAL use of 3D in a graphical shell. Now, if only somebody would take their idea and implement it in C++ (or some other language with a goal other than portability).

You're talking about Project Looking Glass, which is still in alpha, but will eventually bring a true 3d interface to the Linux desktop. It truly looks like a revolutionary interface, and you can see a video demo of Satan himself (Jonathan Schwartz) ;-P demoing it here.

Some developers are already beginning to contribute to the project, which is open sourced. You can find more details and even download a developer preview of the release at this website.

I downloaded the developer preview and briefly got it up and running on my system. I'm running Suse 9.2, and it requires an ATI or Nvidia 3d card with DRI support enabled in your X config.

Re:Ugly UI, Functional UI

[radish]

Seeing as Drawboard is (as far as I can tell) an applet, it probably uses AWT for it's graphics output. AWT is slow. Swing, which is built on AWT, is also slow. But Java != AWT; just because your applet is slow that does not mean all Java apps are slow. Guess what - I could write an app in C which would also slow your machine right down. Extrapolating from one slow applet that all Java applications (notice the difference) are slow is quite a leap.

I write server side Java, there is no GUI layer. Performance is as good as any native language would be. In fact, as has been demonstrated a number of times, Java code running on a HotSpot JIT compiler has the potential to be faster than native code.

Re:Ugly UI, Functional UI

[Joe Tie.]

Now, if only somebody would take their idea and implement it in C++ (or some other language with a goal other than portability).

I don't know if there's really much need to. I gave Looking Glass a shot on my rather old computer, 850mhz and 384mb ram with Java 1.5, and it performed remarkably well.

Re:Ugly UI, Functional UI

[jsebrech]

Sticking a tilted window in a corner isn't particularly useful, because you'd have to tilt it yourself, place it yourself, retrieve it yourself, and then tilt it back to normal. Too much work, too little benefit.

Flipping over a window to get more information is basically a tabbed dialog interface, which with minimal screen real estate wasted you can do right now (admittedly without the cool flipping effects).

Honestly, if 3D interfaces on 2D screens actually let you do more work, they would be all over the place. 3D interfaces are only "faster" on true 3D screens with true 3D manipulating devices (like gloves and so on).

Though I admit 3D interfaces look like they could kick your ass without even trying.

Re:Ugly UI, Functional UI

[Joe Tie.]

The JDS (Java Desktop System) is not written in Java

The JDS is not, but the 3D Desktop written for it is written in Java.

Re:Ugly UI, Functional UI

[cecom]

"Performance is as good as any native language would be. In fact, as has been demonstrated a number of times, Java code running on a HotSpot JIT compiler has the potential to be faster than native code."

A popular myth, which unfortunately doesn't make it true.
There are fundamental performance limitations in the Java language _by design_. The biggest one of them is that all Java code must rely on excessive "pointer chasing" - all data is accessible only through a pointer. This is the ultimate performance killer for modern CPU achitectures and it gets worse as CPU speeds and memory speeds diverge further.
A simplistic explanation for this is that the CPU cannot operate on the data if it has to first fetch the reference which points to the data. This kills any ILP (instructionlevel parallelism).

Java hits the "memory wall" (which many consider to be the essential problem of high-performance computing) sooner than other languages.

This is not to say that slow pointer-chasing code cannot be written in other languages, or that Java is a bad language. Java's design is a deliberate tradeoff between ellegance and performance.

Re:Ugly UI, Functional UI

[cecom]

The problem with 3D desktop is that it doesn't scale well to more than one screen. Even Microsoft has problems with that in Avalon. AFAIK, for multiple-screen 3D support in Avalon they require exactly the same graphics card to be used for the second screen and they essentially duplicate the 3D scene data. Bus-bandwidth killer.

Re:Ugly UI, Functional UI

[GigsVT]

So you are saying, when you take a language designed for portability, and run it in a situation where portability no longer matters (server side app), it works well? So what?

That's like saying "this computer really sucks as a computer, but it works great as a doorstop".

I hoped you could tell from my post, this isn't exactly the first Java app(let) I have ever tried. OpenOffice is slow as shit too.

Re:Ugly UI, Functional UI

[radish]

Portability is one of Javas design goals. There are many others. It is one of the few commonly used languages which can really be called OO (C++ is marginal, Smalltalk is not as commonly used, C# is a Java ripoff), it is a safe language to work in (no chance of a buffer overflow here), it's a rapid dev language due to (amongst other things) the huge standard class libraries - I could go on. Portability is nice, and nowhere near as useless as you seem to think on the server side. For example, we have 30 or 40 developers - many of those develop and test in their Wintel desktops, even though our target platform is Solaris.

Well, maybe not today, but...

[Faust7]

Is it just me, or has almost every second story today had some kind of spurious leading comment tagged on to the end?

"Perhaps this is the end of Microsoft?"
"Perhaps this is Apple's rebirth?"
"Perhaps Sun is growing up?"
"Perhaps Firefox really is taking over?"
"Perhaps Linux really is taking over?"
"Perhaps games are sacrificing gameplay for graphics?"
"Perhaps RIAA/MPAA execs really do eat babies?"
"Perhaps AMD's stuff is better than Intel's?"
"Perhaps Bush really is an autistic monkey?"

Re:Well, maybe not today, but...

[mattyrobinson69]

You misspelt "" as "perhaps"

consumer level?!

[Anita Coney]

Did the poster even read the review? The machine 3D Edit was tested on had dual Xeon CPUs running at 3.06GHz with 1MByte L3 cache, water cooling, 2 gigs of RAM, 15,000rpm SCSI hard drives, and a Radeon X800 XT.

Exactly how many CONSUMERS have THAT system?!

Re:consumer level?!

[LiquidCoooled]

I used to.

Then I found out it still wouldn't run Longhorn, so I sold it.

All I got left is this damned t-shirt.

Re:consumer level?!

[British]

The question is: how many Video Toasters is that equivical to?

man, wish I had one of those video toasters back in the day.

Re:consumer level?!

[Anita Coney]

A friend of mine had one back in the 90s. It was awesome even by today's standards, and out of this world when compared to what was going on back then.

Re:consumer level?!

[jackbird]

If you pay your rent doing digital content creation, you buy the top of the line (or 1 generation back, if you're cost conscious). New gear is always reviewed on high-spec machines.

Re:consumer level?!

[Anita Coney]

I totally agree, I'm just not certain what professional "digital content creation" has to do with consumer level machines.

Re:consumer level?!

[jackbird]

The line's getting blurrier by the month. Granted, I hadn't looked at that screenshot before posting my comment - it looks like someone's first skinning project as applied to Media 100's horrible NLE software.

Re:consumer level?!

[Bazman]

It also had a Matrox RT100 real-time editing card, which can do all the things they were doing with the wobbly 3d interface, but using the lovely nice sweet Adobe Premiere interface.

Of course, it does cost 700UKP but comes with lots of software - Premiere, Encore, Photoshop, After Effects.

I'm tempted...

Re:consumer level?!

[Kalgash]

Don't forget the pile of cash.

Re:consumer level?!

[drinkypoo]

Video toasters do both analog and digital things. As such they can do some things on the fly that are hard or expensive to do digitally, but you can't do as much with them as long as you can accept that it will take a while to render through the software method.

Flanders, star wipe!

[PoderOmega]

I can see it now: millions of home videos will have a star wipe between every shot!

Video Shader

[sklib]

A friend of mine who used to work at ATI made a video shader demo that shows some neat video effects you can do in just a pixel shader -- i.e. render 1 rectangle that fills the screen with the video as your texture, and do all the "fun" stuff in a pixel shader. The ATI developer page that links to the binary is here.

If you look at the requirements for that demo, it wants a radeon 9500, which means that cards have bene powerful enough to do these things for years. I wouldn't be surprized if apple's video editing tools used the video card to composite scenes off-screen. Probably the same thing for newer versions of Premiere.

Re:A text-based medium requires literacy

[iamthemoog]

I'll just loose my elitism ...embarassed by this too?

I've used it as 3D Map Renderer for Numerical Data

[freejamesbrown]

It's as good as anything, imho. The learning curve can be a pain depending on backgrounds and what technologies you are coding in. DX9 has a lot of .NET stuff that's sorta solid, yet still has a little bit of a beta quality to the API if you ask me. I got my job done and people were happy. (This is of course, as of a year and a half ago... so things like documentation have probably gotten way better.)

If you are in MFC land, DirectX isn't a bad choice. Of course, I'll always have a soft spot for OpenGL, but platform situations are often out of our control.

m.

business apps should use this untapped power

[Anonymous Coward]

I've always wondered about the untapped potential of video cards.

My father is an architect, and while he has the highest rated hardware his applications take no advantage of the 3d acceleration technologies of the video card - it seems like such a waste to me. All this effecient rendering power in these little affordable cards, and no serious business apps taking direct advantage of that.

Re:business apps should use this untapped power

[Gruneun]

I've always wondered about the untapped potential of video cards.

Take a walk through the Internet. How many animated gifs of flaming cauldrons have you seen? How many of those sites had anything to do with fire or metal cookware? They do nothing for the page, but add to the download time.

Is there a need for a lens flare being rendered as I scroll through a spreadsheet? I'm not knocking 3D acceleration, but most technology is invented to solve a problem and I don't see how any potential problems with Word or Excel are solved by adding DirectX. In the case of the DV editing software in the article, the wildly different interface actually detracts from the usability.

Matrox RT2000/RT2500

[1984]

The (now venerable) Matrox RT2000/2500 made use of the 3D features of the graphics card for video processing. There was still an additional board doing a lot of the work, but it was the same basic notion.

It was very impressive playing with real-time 3D transitions, flips, (one) alpha channel and so on at DV res one a standard PC. IIRC Final Cut HD depends similarly on the graphics board to be able to edit HD content on a Mac without additional hardware.

DirectX Website

[Anonymous Coward]

looks like microsoft arnt so enthusiastic and have pulled the directx page: http://directx.microsoft.com/ ;]

Re:DirectX Website

[ckelly5]

umm

http://www.microsoft.com/windows/directx/default.a spx

try that ;)

Um...

[Mystic0]

DirectX is a WINDOWS product. As long as Microsoft controls it, it is useless unless you don't care about your product being cross-platform.

Any sort of fancy API you can write for DirectX can be written for OpenGL, although it may be a bit harder.

Re:Um...

[Junta]

Strictly speaking, that's not true. Direct3D is analogous to OpenGL, DirectX is more analagous to SDL, in other words including multimedia functionality beyond 3D graphics extending to sound, 2d video acceleration, input devices, etc..

Still, the point is valid, SDL/OpenGL is a cross-platform approach that leaves a product's options open.

Re:Um...

[TrancePhreak]

Does SDL have video playback functionality similar to DirectShow? It seems that something like that would be usefull for the purpose given in this program.

Re:Um...

[Junta]

Strictly speaking, no. SDL at best provides fast/convenient methods for getting data to a surface, but no support for non-GL scaling, and no architecture for video codecs (ala gstreamer's intent). So SDL isn't even a good comparison as their scope is limited relative to DirectX.

As far as cross-platform video architectures, don't know of any.

Re:A text-based medium requires literacy

[gobbo]

I would be, if it were unintentional.

not new

[MrP-(at work)]

Nvidias volume control software for nforce based motherboards with onboard sound use directx

Directx has been used for things other than games for years

Hell my tv tuner card ive been using since 1998 uses directx

Comment removed

[account_deleted]

Comment removed based on user account deletion

If this turns you on, get a Macintosh...

[argent]

The whole GUI is OpenGL-based and GPU accelerated... not just one application on a few video cards...

Re:oh boy

[mattyrobinson69]

I didn't say i belived them all, infact 4 is possible but i dont believe it.

Would you trust somebody who used theft and deciet throughout their career to not have ulterior motives?

Im just undecided on which of the first three (although im leaning toward beliving its because of the tax breaks)

Unique UI != Good UI

[Jugalator]

I thought the review showed this...

"To say that 3D Edit has a unique interface is a rather large understatement. The fact that the GUI is entirely rendered in 3D means that it diverges entirely from Windows standard features. There is no file menu along the top, and no button bar beneath. Instead, everything is operated using proprietary buttons and dialogues. With no context-sensitive help or tool tips, this takes some getting used to."

Eh, ouch.

"Almost all Windows apps put a handy Look in: drop down at the top of the File>Open dialogue, containing common destinations such as My Documents and drive letters. But with 3D Edit, finding files stored outside the default locations is a laborious task of cycling up and down directory tree structures in the Browser Tool. Theres no recent-files list either, and thats a further pain. The file dialogue is fundamental to using 3D Edit."

"The 3D Edit interface is full of such quirks. The bottom line is that a close read of the HTML-based manual is a necessity for anyone hoping to make any sense of this software."

Well...

Thanks, but do you offer this editor with a normal UI? No?

Yes, for more than just gamers.

[MarkGriz]

If history is any guide, it's great for DoS and remote exploits too

Re:Yes, for more than just gamers.

[TrancePhreak]

Doesn't that page say all vulnerabilities are patch? :P

Re:Yes, for more than just gamers.

[MarkGriz]

Doesn't that page say all vulnerabilities are patch? :P

It merely says *known* vulnerabilities are patched. Besides, it was a joke. Laugh.

A grammar lesson

[GatesGhost]

LOL. its annoying to see people (presumably well educated) cant tell the difference between "than" and "then"; the former implies some sort of comparison, the latter is used to denote some sort of consequence or a temporal change ("he is dumber 'than' her"; "if i do this, 'then' i am dumb"; "i will do one thing, 'then' i will do something else")

Re:A grammar lesson

[Freexe]

Maybe you can explain the difference between affect and effect while you're teaching us.

I think in this case though, it was probably a mental typo.

Re:A grammar lesson

[GatesGhost]

the meaning is pretty much the same. the difference is that affect is a verb while effect is used as a noun. ("the affect is the effect") and yes, i am a smartass.

Right, but...

[the_skywise]

You're still requiring a GPU.

(I know, I know... it's like saying "but you're still requiring a math-coprocessor" just before the Pentium came out... I'm just sayin'...)

Anybody noticed...

[fozzmeister]

... that picassa is super smooth when moving things around, it feels like its using d3d stuff etc. even my dad who's the biggest technophobe around commented on how smooth it was. This is on a Matrox Mystique however!

I use DX9 and HLSL to turn a machine...

[Assmasher]

...only capable of performing machine vision on 2-4 video streams into a machine that can run multigaussian motion detection on 20+ video streams via pixel shading. Image processing can really benefit from the floating point monsters that are today's consumer level game cards.

BTW, the shader for multigaussian motion detection is really large and yet runs at THOUSANDS of video frames per second including the time to upload a dynamic texture (video frame) to the card every frame render and pulling it back down to get the motion map. ;)

Without the texture upload and download(a useless test but indicates the relative power of these 'game' cards) the shader runs in the TENS OF THOUSANDS of frames per second.

Re:I use DX9 and HLSL to turn a machine...

[Assmasher]

Actually, this IS using AGP 8x and using a rendertarget.

For example, with a single video input source feed at 320x240 (a standard) I upload each ::Render() and lock & read the resultant render target and I get 3500+ fps (the video is simple textured onto a quad made from 2 tris.) This depreciates quickly when adding video streams.

You missed the point....

[WARM3CH]

...All for a "purty" interface...

The point about this program is not about using DX 9.0c accelerator cards to draw the GUI, but to accelerate the effects engine of the program using pixel shader 3.0 insturctions. A modern graphics card can run filters and nonlinear effects MUCH faster than the faster CPU. Now, with nVidia cards you have 32 bits floating point numbers for each of the RGBA channels of a pixel which is quite enough for many signal processing algorithms while ATI's 24 bits floting point numbers are a bit limiting.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Motion

[ibentmywookie]

Not to mention Motion, which uses OpenGL to apply effects to video in real time.

Re:Motion

[damerow]

And Jahshaka has been working on it for a couple of years.

GPU's are the new CPU

[katorga]

Reality is that new GPU's from Nvidia and ATI are larger and more powerful than existing AMD and Intel CPUs (albiet at the cost of being less general purpose). for instance, the latest ATI chip has a cpu, fpu, mmu, and 12 vector pipelines. The pipelines are fully programable.

Thats a lot of power going to waste when not playing games. Longhorn will require and use it for the GUI. OSX Aqua uses it for the GUI, and now this app uses it for 3D processing and for the applicaiton interface.

IMO, its time to reverse designing a system. The motherboard needs to be the video card and the CPU needs to be on an expansion card with a HT pipe into video memory.

Already done

[IoN_PuLse]

Ever heard of Blender or Jahshaka?

Other uses

[adeydas]

There are tons of uses of Direct X other than gaming in the animation industry. Softwares like Maya uses Direct X too for rendering its animations.

Motion

[Have Blue]

Apple has been doing this for months.

What about business intelligence?

[3n1gm4]

Seems like the use of DirectX for data mining and correlation analyses would be useful

Re:A text-based medium requires literacy

[DevolvingSpud]

> Or maybe I'll just loose my elitism, and start speling like everyone else, and call it the slashdot affect.

Their you go, thats the spirit! You're elitism just seperates you from teh group.

Comment removed

[account_deleted]

Comment removed based on user account deletion

OffTopic: Galileo

[Sique]

Galileo: "The Earth revolves around the Sun!"
Score: -1 100% Flamebait

The correct score (according to the world of ~1630) was "-1: Just a theory, not a fact" ;)

Actually Galileo Galilei (and other scientists like Johannes Kepler) were allowed to explore other concepts for the cosmos than the ptolemaian, provided they called it "a mathematical/astronomical theory" and were not talking about "proven facts".
The catholic church in the 16th and 17th century was very interested in the results of the new research to improve the calendar (resulting in the reformed Gregorian calendar introduced in 1582) and the navigation, to help exploring the newly discovered worlds and baptise the people there (ok... actually to enslave the population).

Galileo Galilei got into trouble because he refrained from the "theory" and was starting to talk about "facts", which resulted in a scholar dispute between him and theologicans at the Sorbonne University and the Vatican. He never was incarcered into the dungeons of the Vatican. During his trial he stayed in the Villa Medici, because he was the Florentian Medici's court teacher at the time. It is still not clear why he was withdrawing all works he ever published at the trial, when the only thing he was accused of was calling a "theory" "fact". Being put to house arrest for the last 10 years of his life at his own house near Florence (later he was allowed to move into the city to live near his doctor) and being paid a monthly sum by the Vatican to cover his life expenses, because he had to give up teaching is a quite nice outcome of a trial where you got convicted of heresy. Other people got burned at a stake for less.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

+1 Insightful

[ggvaidya]

Yeah, what's with that?

Corrected Link

[TheWama]

http://www.apple.com/macosx/tiger/coreimage.html

Same content, improved presentation.

okay, somebody woke up (again)

[l3v1]

Reading the news, than the article, it seems nobody ever saw any app which would have happened to have OpenGL-based interface [i.e. GUI]. Okay, I help, think Blender.

It is good there are people who are open minded enough to see through the cloud and recognize usabiltities and applicabilities of provided tools. But come on, don't hype it for this reason.

You can hype the gpgpu idea though, which is a very nice way to go for these kinds of applications (too). [Although using the GPU for other tasks then drawing doom3 is mostly revolutionary for the joesixpacks out there. Lotsa many conference papers dealing with the matter popped out mostly in recent 1-2 years.]

PCI Express, not Direct X

[tinrobot]

The real power for video editing will come as PCI Express takes over.

The reason - unlike AGP, the PCI express bus is truly bidirectional, and allows for a whole lot more data to be sent to the card from the application. If you went to Siggraph, you would have already seen vendors ready to take advantage of this added bandwidth. You'll see HDTV editing, 3D Rendering, and many other apps moved to the GPUs of high end graphics cards using PCI express. As to the interface, well, Direct X is a possibility, but a lot of card vendors have an interest in promoting their own standards.

Add to list

[ggvaidya]

Could this be the most leet OS and language of 2005?

Re:Old news. GPGPU.

[jsebrech]

Two reasons:

- It requires learning a new way of programming.
- It's not available on all systems.

Unless and until this stuff is available as os-level api's where a single function does all the work for you from any programming language, and where it falls back to software-based algorithms if no appropriate gpu is available this stuff isn't going to go anywhere.

That's why apple created coreimage.

You mentioned transparencies and texturing...

[gandell]

In theory, what you're saying sounds like the greatest thing for video editing since Final Cut Pro.

I wonder if anyone has the numbers on 2 similar techniques, using hardware rendering, then software. I would love to see comparisons when this technology fleshes out.

Re:You mentioned transparencies and texturing...

[SenseiLeNoir]

There are plenty of number already in existance.. search on Google for advantages to the RT2000 system. This already uses the 3D card, and yes, you are able to do real time work at full res NTSC/PAL with up to 5 sources i believe. It can even do it with "live" video.

Shameless plug for the M4 videojockey tool

[kop]

A video-jockey mixes video to the music at party's'. It is the perfect excuse for a geek to bring his laptop to a party.
M4 is a free as in beer VJ tool, not to be confused with GNU m4: an implementation of the traditional Unix macro processor.

M4 can allows realtime manipulation of Texts, Fonts, Images and Movies
It has proven itself to be fun and useful during many of
our VJ sessions.
Download it at http://www.captainvideo.nl/m4//

So why is this on topic? M4 uses Opengl and DirectX to do image manipulation, and we should all write to Rob the developer to make him do an linux/macosX port!

and of course

[drfrog]

its crossplatform capabilities bar none

As mentioned in other post...

[M.C. Hampster]

Your post is total bull. From the Longhorn Developer Center Home:

"Avalon" is built on top of DirectX, which enables it to unleash the full power of the graphics hardware present in modern computers, and is engineered to exploit advances in hardware moving forward.

Not the right market

[mkiwi]

The primary market lies in two places for video editing on a high-end system: Avid or Apple. The specs on that machine, while not nearly as good as mine ;), were very high and i do not suspect a professional or consumer would buy a high-end graphics card for a "normal" machine and expect their software to scream. At that point, one may as well go with the Avid or Apple system, two proven industry standards.

Re:Old news. GPGPU.

[renoX]

> - do large-scale matrix math, or other parallelizable calculations?

Because usually scientific computations are made with 64bit floating point (or above: Intel FPU provides 80bit floating points precision), 32 bit won't cut it..

Oh and don't believe the hype: when Nvidia made noise about 128bit computation, it was 4*32 which is normal in graphism (ARGB) but very different from 128 FP math..

SDL ?

[MarkTina]

That would be SDL then wouldn't ?

Re:Old news. GPGPU.

[swb]

IIRC the newest series of nVidia cards support that plus more, enabling (according to the PR, anyway) stuff like faster-than-real-time video transcoding.

moo

[thealphagimp]

You need Direct-X 9.0c for 3D Studio Max 7! Because I'm cooler THAN a gamer.

objective

[SanityInAnarchy]

From people I talk to, OpenGL is faster than Direct3D for 3D stuff. Whether it's easier depends on who you talk to, but there are enough libraries written on top of it to make it work. Not liking it because you don't like programming for the raw API is like saying you don't like POSIX because it's harder than, say, MFC.

But, it's not fair to simply compare OpenGL to DirectX, or even D3D. It would be fair to compare, say, SDL to DirectX, or Ogre to Source engine (what a beautifully Orwellian name, that). Remember -- what you want out of the 3D API is raw speed and the features you need, and I think Doom 3 proves that both work at least as well under GL as D3D. What you want, instead, are more and better tools to do that.

And, you're right, GL wouldn't have died without Doom 3, but it probably wouldn't have had all the nice custom features and support that it does now. Case in point -- both nVidia and ATI released new GL drivers just in time for the release of Doom 3, just as they released new D3D drivers just in time for the release of Half-Life 2.

But anyhow, if I take you at your word, than OpenGL is objectively better than Direct3D in all ways except that most engines have already been written for D3D, and the authors show no interest in porting to GL, meaning that if most people had to choose, they would choose D3D.

wait -- what interfaces?

[SanityInAnarchy]

You barely see the interfaces, the thing that was cool was that they are hovering in front of you. Let me know when GL can do that. I doubt DirectX will get there first. Oh, and you're thinking of Zion, oddly named because it seems closer to Hell than Heaven.

No.

[SanityInAnarchy]

Ever run a webserver? Of course not.

What we need is to redesign the CPUs to catch up, or design things more specific than a CPU but still more generic than, say, Apache.

You just contradicted yourself.

[SanityInAnarchy]

If only one ICQ client was worth using to you, how would you feel if that one client wasn't available because it wasn't "the standard Linux ICQ client"?

And ICQ has been pretty dead for pretty long. When did you last try Linux?

Oh, and, uh, cough, cough, SDL. Simple Dynamic Library. Does everything DirectX does, uses OpenGL as a background.

If you want to write a book for DirectLinux, which teaches people to use SDL, go ahead. But it'd be better to call it something like "SDL for Dummies".

Linux has games. UT, UT2003, UT2004, Doom 1, 2, and 3, Neverwinter Nights, Quake 1, 2, and 3, Rise of the Triad, Duke Nukem 3D... Go read this and come back when you've got enough info to justify such a long rant.

Linux distros aren't all that different to the application developer. I run Gentoo Linux as my distro. About 5 apps were developed specifically for it, and most of them were easily ported from other, more popular distros like Fedora and Debian. It takes about 5 minutes to port an app to another distro, 15 at most.

And I wouldn't want a "Standard Linux". Competition is healthy, and I'd rather be able to choose Gentoo Linux than be forced into, say, "Standard Linux" running on a Win2k kernel. But hey, you're free to create a new one, or go check out the free desktop project. Or maybe you should go "****" yourself.

Re:Enough...

[shadowxtc]

I'm a former and current Windows user. I was a Debian-only user for a while inbetween there. I'm also a major geek, hardcore developer, and perhaps most amazingly, usually side with the "normal people" when it comes to these debates. All of my perspectives agree - I don't think you could have explained the problems of OSS any better.

Re:Old news. GPGPU.

[Tim C]

That would be very, very interesting if true - I've (very) recently started having a look at video transcoding, and an average rate of ~7fps isn't a lot of fun.

I don't suppose you have any links to hand, do you? (I'll google myself, so don't put too much effort into it :-) )

I'm not against it either...

[the_skywise]

I like whiz bang interfaces. But I'm not sure if I want them in "workhorse" software like video editing, Office, etc.

Re:Old news. GPGPU.

[swb]

Shit, I wish I did, other than PR/Press release crap I read on nVidia's web site.

As with anything promising, it's not worth a damn unless someone comes up with codecs written to make it work. Of course it's typical for a hardware company to add some features that nobody ever exploits.

I have long been looking for/wondering why there's not some $129 board I can plug into my system and get at least 2x real time MPEG2 encoding, given the fact that $229 will buy you a settop DVD recorder that will do at least 1x real time MPEG2 encoding. Put two MPEG2 encoders from that box on a board and you should be able to get 2x encoding.

The nVidia GPU solution sounds much, much more elegant as it ought to be much more flexible and possible to get higher multiples of real time, in addition to custom transcode codecs (MPEG2->Xvid, Xvid->WM9, etc) so that you didn't have to do multiple transcodes to get where you needed to be.

What further surprises me is that as popular as video is, the MPEG2 encode times are crushingly slow even on fast machines. This would be a killer app, IMHO, and I'm surprised there's not further development of this.

Totally lame program!

[Snaller]

It won't even start if you haven't got a mousewheel!

Amateur hour.