With the advent of MiSTer FPGA gaming, and its ability to output nearly any analog signal over its vga port, I'd love to see a way to drive vector monitors (or oscilloscope screens) for Vectrex and Vector arcade games. I'm sure someone is working on it.
Could you also use the green VGA output to control the Z input (intensity) of the beam? Then you could have proper blanking and multiple brightness levels.
@@CNLohr Ya, they never seemed to have a standard for the Z input. Some were TTL on/off only, others were analog but with varying voltage and polarity, and some scopes had no Z input at all. This is still a cool concept, and could probably be adapted to drive laser scanners too. I've done it with sound cards, but it's annoying having to deal with DC offset and other limits.
@@CNLohr Cool project, surprised it has not found more use in the many years that we had VGA as standard. Many scopes have the Z-input on the back. On this scope it is called "EXT-BLANKING". Using the Green chan to drive it is a good idea. You can use the SYNC-signals to blank and hold the trace as well when you have no video signal available.
@@CNLohr Yeah, the levels may have needed amplification to get it to work. I think I am going to try this out all the same as it is a cool high frequency analogue output, will pull my scope from storage and play a bit. Thanks for documenting this.
That is cool as heck. Maybe use the green to modulate the intensity of the beam? IIRC there is a BNC on the back of the scope for intensity. And you pretty much have to port battlezone.
Your skills + Nikolas Wirth's skills = the ultimate "Steampunk" computer. You should check out what the "Super Professor" has done. He has built a RISC processor which runs on an FPGA. Everything is written in Oberon, the successor to Modula, Pascal (Which he invented). YOUR MIND WILL BE BOGGLED. Take the time to digest what Professor Wirth has done. Oberon needs some attention from the hardware guys like you. Oh yeah, Professor Wirth just celebrated his 80th birthday... ProjectOberon.com Ps. Oberon uses a hsynch vsynch scheme which you might look at.
I think it's generally a bad idea to go into x-y mode without a signal, because focusing the cathode ray on one spot like that is likely to burn out the phosphor.
Only thing I can think you can go to X and Y mode with no signal for long period of time is the display being digital and not being analog, like that oscilloscope, meaning it can display same input for long period of time without burning the screen unless it uses a display that has it's display burn after say... 60 to 120 days.
This is a perfect example of using clever code and simple hardware to create easy, though unintended, practical compatibility. I love this kind of stuff!
I've used oscilloscope in the electrical lab in my college few times and lines that are drawn in oscilloscope are so much smooth. How do they achieve the smoothness ? Is it because the whole screen is fluorescent coated and the electron beam just hits the screen continuously or it's just that the pixel density is high enough to produce smoother line ? I'm curious to know.
It would been cool if you could start think about a new type of video signal. Today films are encoded in 2D flat pictures. From gaming we all know there's a movie format for 3D images, like for instance source engine, where you can record in games, then import the movie into smoother and you can control the camera angle and make 'new movies' based on that recorded file. Say one was able to make a video file format or standard with 3D images, like one could have pre sets for different camera moves (POV, etc). I would think that's the next step. I'd like to have my own "room" in 'cyberspace' where I could sit down and enjoy watching all the old movies and tv-series I like. Invite friends in etc. Surely you would need a 3D projector or sorts, or VR gear. Or make new software for the existing VR solutions.
It's really hard because so much is done on the other layers, so much is still done in "screen space" as far as the effects, etc. are. But, there are advances in these areas, and interesting compromises every day.
So cool ! I am busy with the plans to build the head of a Star Wars At-AT walker : I want to be able to sit in it (yes, it will be large) and play games or watch movies on the big tv screen that is also built in ;-) I have not yet found details about the actual inside of an AT-AT, but I want to make a nice futuristic dashboard and I thought about an oscilloscope (and other things) and have it display something cool. Now that I have seen your moving terrain output to the osciloscoop, I really think this should be in it, as if the AT-AT was walking in this terrain :-) I would love to get some hints or maybe a little example program how to start with programming the terrain :-) !
The video of "High That Brings The Low" by Screwed Loose uses a very similar method for their oscilloscope clips but I think they changed the resolution and allowed for the pattern to repeat if there was still space in the encoded graphic file for more points/info. The result is here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-uXKzHtaq2HU.html
Damnit, I thought this would be driving a conventional CRT as if it were a vector display. Alas :p A similar trick with composite video on the PS2 was used in a demo at Revision 2013. It was clever, though, outputting a signal that makes sense for both the oscilloscope AND the TV: www.pouet.net/prod.php?which=61225
+TazeTSchnitzel Would you mind doing me a big favor and writing a hack a day tip for this one since I only kinda get how he's doing it with composite. I get the top two displays (chroma/luma) but I don't get the bottom. I think it would be good to get htere.
I wonder if this can be adapted to work with laserMAME, so you can play arcade games with an arcade vector monitor like from an Asteroids? Also do you think it would be possible to drive a color vector monitor like from a Tempest?
Woah! I think it's a very nice idea (I did design a CRT vector monitor btw). How much X/Y "resolution" can you get? The very early Atari arcade vector games could do 1024 "positions" on both X and Y. How do you turn off the beam during "moves"? Are you using a third channel for Z control?
This is one of the coolest things I've ever seen!! Why doesn't the youtube algorithm put stuff like this on my front page? It'd increase my watchtime tenfold!!
I'm curious, how did you manage to turn the electron beam on and off between drawn lines? From what I know about vector refresh displays (which isn't much), you need three controls: X axis, Y axis, and the ability to turn on and off the electron beam. If you managed to make this display without said third control, could you share the method by which your display can draw multiple isolated vectors without traces between them? I hope I'm making sense with my question. Thanks for your video too. It is quite inspiring, and I'm bidding my way to an oscilloscope of my own to create a display for my Arduino.
I used xrandr --newmode to create a new video output mode where I could specify the sync times. I was able to reduce the time the horizontal sync took. For vertical sync, the beam just lives in the lower-left corner anyway which is off the screen.
I have a question. If somebody makes a vector based tube display like the vectrex one.But with colors. ir would also work with pixels? or only drawing on the display? because if it draws fast enough and it is not pixel based i can see the future of 2d canvas like graphics...
That's a little complicated to explain. Pixels or vectors are just a representation of the image data. For pictures, pixels (literally transliterated "picture" "elements" it's just a different representation :-/. You can totally do a color vectorscope, though. Just can't drive it with VGA.
That would be tricky without a chip made specifically for it... Because you would need to somehow get sync information to have two signals, instead of one. Things like NTSC store sync information in the colorburst, or you could use QAM, or something like that but would be tricky.
PIPBOY! That would be the perfect application for a pipboy. Tho you'd need a flat distorted mirror side input type of tube screen like in those tiny tube pocket tvs. Every pipboy prop i've seen uses a flat screen and whilst it looks cool it's not life like or better said time line correct. But if you were to use this type of screen. Wish i'd have your level in tech so i could make a pipboy like that. Mah, ranting, awesome video!
Problem is, no CRT displays of that size are (or more accurately were) available in a portable format - a screen that size has a tune that is probably 12 inches long.
Great project man, I had a little search about doing something like this a couple of months ago and found a program to play asteroids on an osciliscope by hooking up the left and right channels of the stereo sound, to the x and y of the oscilloscope. It worked quite good. What do you think the advantage is in doing this via vga? I imagine it can run faster and it could be easier to encode an image than sound, but the image you showed on the screen didn't look like it changed much from pixel to pixel.
I am surprised it worked because of the lack of DC coupling on the outputs. The VGA helps you in two ways: (1) It's repeatable; the DC offset does not disappear. (2) It is very high speed. The dot clock is in the 20 MHz range, instead of the 40 kHz range.
Can you also control the intensity of the beam with the Green channel? I would definitely like to see some modern Vector displays using separate lasers for each color. They would be difficult to make compact, though.
Some oscilloscopes have a z-axis/luma input (usually on the back). You could use the third color if the vga output to control luminance if you have such a scope.
Awesome! So what software do you use to create the red-blue vector rendering? And with the color space being 8 bit per color you get 256 × 256 positions on your screen, right? But you can draw 1024x786 or some 800 thousand points on the screen every 1/60th of a second, right? How might one increase the the resolution? Are there VGA outputs with 12 bit per color channel?
There may be, but you can also increase the resolution by filtering the output and using a very high resolution, i.e. 2222 2323 3333 would look like pixel 2, pixel 2.5 pixel 3 .
+ultort Soundcards shape the signal wonky, usually don't give you flat response down to 0Hz and indeed don't give enough. Bandwidth is entirely dependent on what you want to do with it, though.
Great idea! Couldn't you use the green channel for intensity modulation? If your scope has a Z input that should be possible with a single transistor for amplification.
Yeah, opamps aren't fun in this frequency range. I thought it could be easier with (RF) transistors but I've yet to mess around with video signals. Would've been cool if you could get rid of that bright dot in the corner.
Pretty cool. How many points can it draw per second? Also, can it print in a smooth analog fashion, or only lines? What kind of input do you give to it? Just points or can you give it a function?
+Krusty Kraken It can totally do smooth lines, etc. I don't know any really good metric... but, I liked the style of lines plus points. I don't really know how to quantify "number of lines" :-/
CNLohr Oh, ok. I guess it moves at a certain speed and you can say that it can travel a certain distance per second, or frame. So, essentially it can move in a smooth fashion, but it's limited by a digital output which operates at a cycle rate. I guess at this point I'm wondering if any output on a computer is capable of changing in an analog fashion over a period of time, rather than abruptly changing like a digital output. It's kind of weird for me. Very cool, though! I wonder how much something like this costs.
Cost is probably about $2, $1.90 for the connector, $0.10 for the resistors... I think just about any computer can do this, but you will need Linux so you can specify your own dot timings on the output signal.
How about this; Take those x and y signals, and instead of feeding an oscilloscope with it, feed it to a couple of galvanometers with a piece of miror glued to it (quick and dirty galvo's: basically a motor with an elastic on the axle to limit movement so that it stalls rotation, but allows for movement of the mirror in the stretch boundary of the elastic). Then position the galvo mirror setup so that you can bounce a cheap laser beam off of the two mirrors, onto a wall or something. 10 dollar laser projection system. Now take into consideration under/overshoot of the system (axis lag? I have no idea what the official term is) caused by the elastic not being stiff enough (slack = overshoot), or too stiff, and the motors aren't driven hard enough, or are too weak (undershoot, or scaling errors), andonce you compensate for that, you're a good 200 dollars into a professional laser show setup. Now add a blue and green laser (trivial with the vga signals, which is a brilliant idea btw), hook the lasers up so that they can be varied in intensity, either direct by varying input power, or indirect, by using a (really) high speed stepper motor with a diffraction grating (or cheaper, a transparency sheet with a 0% - 100% opaque gradient printed on it) to limit beam intensity. Now you're a good 400 dollars into it. Closed captions are really awesome. Adds a lot of value to your video's. And I would not mind ads at all. In fact, why not go the subbable.com or patreon route, so people who want to throw money at you have a nice target to hit? ;)
It would take some circuitry to boost the power from the 1v signals on the GPU, and I'm really bad at that sort of opamp stuff. I think a fair bit of the over/under shoot stuff could be handled by control loops on the data, which isn't too hard to run. I'll keep doing the closed captions, and ads on my how-to's as they get older. I think I may stick a full screen ad on here in a few days. You're the second person to suggest patreon route... I need to get back into the swing of regularly releasing videos... then I suppose that would work out. Maybe I will just go for it. And honestly, some of my hacks are getting to be more complicated and expensive, and trying to pinch pennies leads to heavy delays :(.
If you can have feedback back to the software side, then yes, most of the over/undershoot is definitely fixable in software. I was just thinking about it in an "as cheap as possible" way too, i think. I don't know if the cost of doing it in software would be smaller than fixing the mechanical parts. (i'm not really good at estimating time investment) You're doing a lot of stuff that people really like, so i'm sure people (at least some) are more than willing to compensate you for the time / money you invest yourself. Just put ads on everything you can monetise. The people who don't like ads have adblock plus installed anyway. And as for the patreon /subbable route; Well, any dollar received is good, right? Cheers man ;)
Indeed... I guess my thing with these projects is I am only willing to make certain amounts of an investment in unknown territory. I expand my territory of possibilities, but slowly.
good call. RU-vid combined with an agile mind often equals spending tons of money to show cool stuff. I guess there IS an incentive to cull enthusiasm :)
i absolutely love this concept as a way of outputting vector graphics! i'm still somewhat of a beginner when it comes to programming but i definitely want to give it a shot using the code and schematics you assembled and see if i can figure it out enough to adapt it to something new. i've been dabbling in analog vector monitor graphics for the past year or so, as well as beginning to study the fascinating field of electronics, but haven't found too much into developing vector graphics renderer software to make the signals to feed into the display/scope, though this seems like this method has a number advantages to using audio output DAC setups like a lot of people in the area seem to use, since those are usually limited by sound card sample rates among other things. i tried to get the code running on ubuntu a while back on an older laptop in August but to no avail, but having recently gotten a much better system recently, i'm hoping i have better luck with it now. would i be able to contact you if i run into issues or have questions on the code or setup? Again, i just wanted to say this is a fantastic idea!
+SamBlye The biggest challenges are (1) making sure you can trick Linux into accepting another fake monitor. (2) Getting your xrandr lags right to setup a screen to your liking, that will let you set to that resolution and blank and not freeze your GPU. (3) getting the software to work. Feel free to contact me on google for quicker help.
