Digitizing VGA? 31
mvidal01 writes: "I have been looking for hardware to digitize the VGA output of a PC. After searching for a long time I have found a few solutions but nothing really great. If possible I would like to avoid converting the VGA signal to NTSC or PAL video with a scan converter because the quality tends to suck. So how about it Slashdot readers. Anyone know of hardware that will allow hires VGA to be digitized?"
Sony DVM-CDA1 (Score:2, Informative)
Nope -- Sony DVM-CDA1 (Score:3, Informative)
Though it probably won't get too close to what you're after, HD on DVHS is a step in the right direction. Unfortunately, these recorders usually accept only the pre-digitized bitstream (already conpressed w/MPEG-2) coming off a set-top box... so unless you're ready to drop a truckload of cash on an HD MPEG-2 digitizer, you're pretty much right where you started.
They said there are only two real perverses (Score:2, Funny)
Why? (Score:1)
If you told us why you'd like to digitize a VGA video signal, perhaps myself or someone else here might be able to suggest an easier alternative to doing that.
Re:Why? (Score:2)
Re:Why? (Score:1)
Re:Grammar. (Score:1, Flamebait)
perhaps myself or someone else here might be able to suggest
quoth the replier:
Of course you meant to say, "... perhaps me or someone else here...."
while you are correct that 'myself' is incorrect, me is also not correct...
because 'perhaps me might be able to suggest...' isn't right.
the correct phrase would be 'perhaps someone else here or I might
Re:Grammar. (Score:2, Flamebait)
A little less clear in this instance is who gets top billing. Ordinarily you would list yourself and a second party as "...foo and I...", but in this case the "...someone else here..." isn't someone acting in concert with you, but as a second source only to be called upon in the event of your failure to help satisfactorily, so I believe the "first one and then perhaps the other" sequence makes "...perhaps I or someone else here..." the better choice of wording in this particular case.
Re:Grammar. (Score:1)
Some people are just fucking morons, though. I suppose your error is justifiable.
Yeah. I guess yours is, as well,
Re:Why? (Score:3, Interesting)
Maybe to transport it over a longer distance than analog cables allow? While I agree that maybe it is a bit silly to do a AD conversion after a DA conversion has already taken place it does make me wonder why monitors don't do their own (presumably optimized) DA conversion and receive their signal in a digital form only.
It has always puzzled me that my monitor has to tell my PC what signal it needs (you needed to look up the proper refreshrates yourself before plug and play monitors arrived). Surely, modern gigabit network interfaces are fast enough to transfer the signal digitally? A 1600*1024 screen amounts to about 1.6 million pixels. In 24 bit color this requires about 4.7 MB for one screen (uncompressed). At 60 screen refreshes per second that's 300 MB/s. That is pretty much for a 1 gigabit line, so lets use two lines or do some compression (at this rate most frames will have a lot in common).
What would be the advantages? Simpler video cards in the PC. Better performing displays since the DA conversion can be tuned to the specific display. Better connectivity to e.g. alternative types of displays (beamers, tv's, etc.). Longer distance between PC and display and input devices (might as well tunnel USB signals over it too!), postprocessing of the signal in the monitor (digitally resize, display signals from 2 PCs on one screen and switch between them or show scaled versions,
There's a lot to be said in favor of getting rid of the analog transmission of video signals between PC and display. There's already standards for the digital transmission of audio (e.g. spdif) so why not do the same for video?
Re:Why? (Score:1, Insightful)
some LCDs do support digital input but they cost $$$$.
Re:Why? (Score:2)
Besides, 300 MB/s would be a worst case scenario assuming no compression whatsoever. I'd be surprised if it weren't possible to cut that by at least a factor 2-5 using lossless compression (bringing it within range of gigabit ethernet and other relevant standards (maybe even ieee1394 or a successor of it)
Digital video signal (Score:2)
Re:Why? (Score:1)
For a CRT, doing the D/A conversion inside the monitor is asking for trouble. Do you realize how much current is charged up inside that box ? I think it would be an electronic challenge to perform reliable data conversion within inches of the tube.
Re:Why? (Score:2)
An imac (the old type) has the entire PC in once case together with the CRT so I think that is a non issue. The DA conversion needs to be done anyway so you might as well do it inside the monitor.
Suggestion (Score:1)
It'd be nice to know the application... (Score:4, Interesting)
A better Ask Slashdot tells us what you're trying to accomplish, and why you've dismissed the 'obvious' solutions (such as software, or a card inside the computer, etc).
The only reason I can see where you'd have to do it this way is if you don't have full access to the computer (can't load software, can't open case, trying to circumvent digital copyright protection, etc), but you still want to see what's on the screen.
I've been thinking about this for awhile mainly because I have to monitor hundreds of computers remotely. In some cases I need to control them before the system completes the boot-up process (ie, before software can be loaded), and in other cases I'm using operating systems for which inexpensive software is not available even after boot up. The solution I'm leaning toward is custom hardware (everything runs at 640x480 or less) which compresses and sends the signal over modem or tcp/ip. I haven't implemented it yet, but I enjoy thinking about it. Capturing higher resolution pictures is more involved.
Speaking of which, what do you need to do, exaclty? Do you just want to capture screen shots, or are you interested in spying on the stream in real (or near real) time? Do you need to capture those streams? How much color depth do you need? What's your budget? Does it need to be user friendly, or geek friendly?
-Adam
Re:It'd be nice to know the application... (Score:2, Insightful)
http://www.dakota-us.com/us/overview.cfm?prodID=3
It's not cheap at $715, but doing your own customer hardware is not that cheap either.
This may even be a solution to the original poster's original question -- this card is supposed to send text and graphics over the LAN/modem connection, so perhaps there's some way to capture that data.
Re:It'd be nice to know the application... (Score:1)
I was looking form a VGA digitization process fairly recently as well..
My concept is that I can remote into any of the machines at my house that I want to, but the access is OS-level (and dependent on video frame buffer in Windows). I wanted to be able to drop into a DOS prompt or go into the BIOS or boot another OS.
If I had a machine (Machine A) that was hooked up to a KVM that was connected to every other machine in the house, an application that could display the screen on the remote machine (Machine B)--this viewing would happen through any TV-input-enabled software, I suppose. Then I could see the mouse and keyboard happenings on Machine B. The biggest problem I found in this concept was the design of an application that would take software-level mouse and keyboard commands sent into Machine A over VNC and converting those into outputs back out through the KVM. If all of that could be solved, the last problem is having a fairly inexpensive KVM that would switch to the different remote PCs with the same keyboard 'output' from Machine A.
My employer's solution involves giving the card its own ethernet connection and IP address. The card then sends and receives data on the machine on top of the PCI bus. You then use special software to access these cards over the network. They called them RIB cards. The problem is that they're way too expensive for a hobbyist to install in every machine.
At some point I saw a serial-based controller card that was installed in the client machine, but I've had a tough time finding those again and they're likely too expensive as well.
Some folks here have mentioned solutions that do what I've proposed through hardware. Again, the problem is cost for hobbyists. I've heard about solutions (and seen on in my company's Data Center) that take Machine A's functions and install them on all the machines that you'd like access to in the form of a card (likely some kind of single-board computer). If the problem could just be solved regarding getting the keyboard and mouse signals back into the KVM from the host computer, then ~$300 worth of equipment could perform the same task as $8000 (Dakota). My research took me to alternative input devices for the disabled. I think that's where the grail lies. Just don't take it beyond the Great Seal ;) .
To answer some questions directly that were asked:
Streaming or images? Streaming.
What color depth? 8 bit will do, but 16 bit would be really nice.
What's your budget? $300-500
Geek or User friendly? Geek
did you try this site? (Score:1, Insightful)
Re:did you try this site? (Score:1)
I don't understand (Score:1, Insightful)
That hasn't happened this time, how come?
Commercial Product: Raritan Paragon 'TeleReach' (Score:3, Informative)
This isn't just the 'Cat5 Video extender [raritan.com]' product which has been around for years, this is something entirely new, with frame grabber and video compression, SSL, RADIUS authentication, etc.
The client runs only on MS-Windows 98+, at a minumum of 20Kbps. The server to be remotely controlled can be anything that supports VGA and PS/2 keyboards, including some Sun products.
So far, I have no complaints. We haven't tried all of the advanced features yet.
I know why *I* would like to have something... (Score:1)
Off-The-Shelf Product (Score:2)
How about RGB over twisted pair? (Score:2)
Here's one by Extron. [extron.com]
(VGA, SVGA, XVGA, etc. are all RGB, for those who don't know.)