which micro to pick

What kind of effects would that be? There are so many video processing procedures covered by the word "effect" that it means virtually nothing in and of itself.

By the time your micro has all the necessary features in terms of processing power, storage capacity, bandwidth and human-machine interface, you'll most probably end up having duplicated most, if not all aspects of a conventional PC. So yes, you had better start the other way round: begin with a PC and remove what you don't need.

Forget about it. Better do planting flowers or collecting poststamps. This is something not for the amateurs. Regardless of what you do, it will be years and years before you will get anything useful.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

What a load of bullshit.

"Amateurs" are capable of just about anything they put their minds to, unlike some bitter, calcified consultants.

My "amateur" friend built a video frame grabber using TTL logic and wire-wrap, because nobody bothered to tell him he couldn't.

To the OP: you need to understand what it is you're asking for. Is there a product you've seen that did what you want? How did they do it? Do you understand the basics of generating video? Can you calculate how much information you need to process for a given frame rate?

You'll get a better reception here when you can describe the scope of what you're trying to do. Don't start with the micro, start with the task. Once you understand how to do something slowly, it's a lot easier to learn how to do it faster.

--Gene

Gene,

The OP said "being able to do this with a micro".

I would guess you have been in the buisness long enough to know what a "micro" is.

So, if you consider a P4 a "micro", then Vladimirs comments stand correct.

Doing this type of project with "TTL logic" like an FPGA, then yes it can be done.

But a "micro" ??

Good luck to you and the OP.

don

high-end

good

for

them

selecting

with

What I like about the amateurs is their naive optimism. Sometimes I have to fix their designs :-) Crap job, it is always better to redo everything from scratch. With all due respect to the useless efforts of your amateur friend, I must say that a piece of wirewrap board can't be a usable frame grabber. Even if it seems to be working :-) Besides, with so many different frame grabbers around, what is new or special about that particular one?

Forget about OP. Don't you see that he will never get anything usefull?

VLV

Work out the bit rate for a 1024/768 pic, and find something that is at least 10 times faster. The average micro may not meet these specs

Martin

I wasn't telling him to attempt it with TTL. I presented an example, whereby my friend had no idea that what he was attempting (not for a product, but for _fun_) was, at that time, extremely difficult.

He did it so he could learn how to do it, and I give the OP credit for that kind of enthusiasm, as opposed to the half dozen or so posters we get here each week who are looking for homework help in a subject that they actually have no love or natural inclination for.

--Gene

Were you looking to design a product to do this I would say that a PC isn't necessarily the most cost-effective way to get absolutely top-notch video.

But for almost everything else, and certainly if you're going to do a one-off, getting a screaming-fast PC and doing your work incrementally (i.e. start with just displaying video, then some really simple effect, then on up from there) is the way to go.

Hi, There are several commercial chipset vendors like ST Microelectronics,Broadcomm and Connexant making the stuff what you would be looking for.From your description it seems you are looking for High Definition video processing.These vendors supply some sample source code and drivers which will help you to learn stuff.But they are totally commercial and you need to order big volumes to get it.Incase you are into student level and looking for acedamic project,this wont help.On the other hand,building a chip with all these facilities is time consuming and you need to understand lots of basics like what other experts have pointed out.May be you can ask for a evaluation board from your university and incase you choose ST they provide the code and board and you can tweak it and get a better understanding.I believe incase you are a fresher in this area this would be your best bet!

Regards, s.subbarayan

What an oddly bitter man you are. I don't think I've ever met a /good/ engineer with this attitude. I guess it might be a momentary aberration due to hangover or similar ailment.

You have no way of deciding what is "usefull"(sic) to another person.

"I never miss a chance to stuff a turkey with a poison of cynicism" (c) Larwe

Oh, come on. Bla-bla-bla, mutual respect, politically correct... Everybody knows that the initial judgement proves to be right in the most of times.

VLV

Yes, they are looking for the big orders. However there are the technical difficulties also. There is a big distance from the primitive reference designs and software examples provided by those vendors to getting something real usable. Those chipsets are intended for the big companies with crowds of developers.

I would not recommend the ST for the beginners. ST is unfamous for the cryptic datasheets and lack of the customer support. Although it is difficult to beat the ST on the volume price.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant