Main > Reviews
Ultimarc Arcade VGA2 comparisons using a LCD for PC
Patent Doc:
Genesim
--- Quote ---Lets say you are right. Does it make sense to interpret the code wrong first??? Why would anyone choose 640x480 over 352x288 if they had a choice???
--- End quote ---
OK, again...I will preface my comments by saying that I am so not qualified to know the reality here, but I think what everyone else has been saying is that the entire equation affected the display. Sure a programmer would have designed for VGA rather than CGA given the choice, but that's just it...they likely didn't have a choice. Making the code for VGA has to be more expensive than CGA (I'm guessing...I don't know Pascal, Fortran, or any other language of the day), which would be one reason the programmer would create a lesser design. Additionally, there may have been monitor capable of 640x480, but those were expensive and not the ones being used. The programmers would use the monitor that would be destined for the game and that choice would have been driven by economics as much as anything. No one is saying that if a VGA monitor was going to be used the code would be written to display a lower resolution. They are saying that CGA monitors were being used and those monitors are limited to what 352 x 288. So the programmers would have taken this into account. Chips may have been expensive and processign power a factor, but OMG CRT's were expensive back then ...particulalry color ones (remember early games were balck and white and if they had color achieved it by a colored display on the front (see Space Invaders)). The monitor was easily the most expensive part of the cab and a company wanting to make a profit would use the cheapest possible monitor to maximize profits.
Regarding the analog v DVD, FYI pops are caused by static and hiss is due to overamplifying poor signal (i.e., total harmonic distortion). Neither are experienced with good equipment which as I stated earlier takes in excess of $10K. The hiss and popes are not part of the record nor the recording. Good equipment does not mask anything. Poor equipment will not have the ability to reproduce the frequency nor dynamic range of its high priced cousin without derious distortion. Remember, the discussion is the medium not the equipment. Yes records due degrade, friction is a ---smurfette---...but that's why the comparison is with a new record. Scratches will also cause a problem, but the quality of sound argument is not about the fragility of the medium but its ability, when operating optimumly, to produce the recording. Regarding the dynamic range and the hum to which you speak, this is more a factor of your equipment than the media. Dynamic range refers to the differences in volume levels noticeable not the quality of the sound. Due to how records are displayed, they do have less dynamic range because the background noise is higher. The better the turntable the lower the noise and better the range. Admittedly, even the best record with the best turntable will not have the dynamic range of a CD or DVD but that doesn't speak to the harmonics missed by the digital media. By the way, the approximations that occur in digital media occur over the entire frequency range not just at 100kHz which is about 80KHz above human hearing and in isn't even within the capabilities of the microphones used to record to pick up. By the way, I have experience as a studio musician and know how things are recorded. Rather than speak without experiencing just give me a fair test. go to a audiophile equipment store and try a blind study. You may be suprised. I say may be suprised, because not everyone can hear the difference....you may not. However, I bet that you won't say records are worse...you may not say better, but it won't be worse. And for the record, the warm sound isn't an urban legend...I can hear the difference...and I used warm to explain it ...its like incadescent lights and fluorescent lights....its hard to explain, but the difference is noticeable...you just have to experience it.
Anyway, I've spent to much time rambling.
genesim:
ahofle,
--- Quote ---Nope. Most people around here are simply trying to reproduce as accurately as possible exactly what they saw in the arcades of the day, while you seem to be after some fictitious, nonsensical blocky version of it that was never ever viewed by anyone in the 80s -- not even the game programmers themselves!!!
--- End quote ---
What do you think is happening...LCD's are picking up something that isn't there. :laugh2: Any "blocky" pixels, are an accurate representation of the original code. I doubt I will ever get this through to you. But yes, PIXELS ARE SQUARE!!! They are not little round "dots" as you think they are. Truth is truth, and your ignorance of this is astounding. Even Randy doesn't dispute this. It is all how they are handled after being presented is what is being debated here.
--- Quote ---It's very obvious from your posts (and your lack of understanding my joke earlier) that you don't have a clue about CGA CRT monitors (maximum of 300 or so lines of horizontal resolution). Hint, if you turn them 90 degrees, then they become "tall and narrow"!
--- End quote ---
I have perfect understanding, and what you don't seem to get is that the display was made specifically for the game. If you think that monitors were limited by 300 lines it shows even more ignorance on your part. The tall and narrow display was a manufacturing choice not a hardware limitation.
You say that LCD's have a horrible algorithm. Do you know how idiotic that sounds? It displays what is put in from the source. Even CRTS aproximate because of the fixed resolution(provided by windows and the limtations of refresh rates being able to display the low resolution...i.e. with out good software it would draw the screen so fast that the refresh rate has to be out of the world to display it!). The difference is that good software rectifies the inherent problems. The difference with LCD's is all the other flaws that I mentioned are not paired with it. Your little fantasy that LCD's are somehow inferior comes from a fundamental lack of knowledge that can be seen directly from your post. There isn't any lacking, rather it is too good! That is why as I said a hundred times, good software that doesn't use simple upscaling, but real pixel interpretation can fix this....it isn't rocket science.
--- Quote ---You may prefer it that way, but please refrain from claiming your incorrectly-scaled, blocky version of Mortal Kombat is more accurate than running on the exact same display device as used in the arcades.
--- End quote ---
Have I ever disputed this? The problem....who the hell wants to use a 400x256 monitor taht is utterly useless for anything but Mortal Kombat if you care about quality. BUT on the Arcade VGA they actualy have a close resolution of 401x256 which is a hell of alot better then windows defaulty. That and, I don't like scan lines. It isn't part of the original hardware, and while Pacman could have a point, I absolutely don't believe the Midway creators preferred it! Having MK be my favorite games, I can even hold up for this even more.
--- Quote ---Utter nonsense. Even the mamedevs will tell you the effects are at best a meager approximation of the display of an analog CGA CRT.
--- End quote ---
Again, it is lack of software rather then ability. Meager....that is a good one though. Shows more of your limit.
The only thing "meager" is using default 640x480 to represent a resolution of 224x288. The very same things that many on here claim are a limitation to the LCD are even further a problem for any monitor that is not an exact 1:1 ratio.
--- Quote ---Hmm you say displays were made for the limitations of the "chips", and then say that displays never changed despite improvements in the "chips" over the years. You just completely contradicted yourself.
--- End quote ---
In what universe is this a contraction?? My point was displays didn't change(until later years) in that they were capable from the word go, but chips had to catch up to the capablities of displays that were used back in the 50's. You know why TV's are somewhat good(despite the fact that an S video is used which defeats all gains) monitors, because they were almost exactly the same ones used 10 years later!!! The chips were the bottleneck. I cannot say this enough. The displays could be made inferior as a cost cutting measure because all the extra resolution was not needed! Why on earth would anyone use a display with over 500 lines of resolution(which again was readily available long before there ever was a Pacaman) when the chip set would not benefit from it.
Patent Doc,
--- Quote ---Making the code for VGA has to be more expensive than CGA (I'm guessing...I don't know Pascal, Fortran, or any other language of the day), which would be one reason the programmer would create a lesser design. Additionally, there may have been monitor capable of 640x480, but those were expensive and not the ones being used. The programmers would use the monitor that would be destined for the game and that choice would have been driven by economics as much as anything. No one is saying that if a VGA monitor was going to be used the code would be written to display a lower resolution. They are saying that CGA monitors were being used and those monitors are limited to what 352 x 288. So the programmers would have taken this into account.
--- End quote ---
First of all, this is totally untrue. Just revert back to what I said to ahofle.
Also you missed the meaning of my quote. When I said if people have a choice I was referring to NOW. The video card makes a discrete display by using multiple pixels to create a set of resolution that is more closely related to the old one. i.e. 352 x 288 to represent 224x288 of Pacman. The extra pixels is the black bars on the sides. So many on here are spouting that 640x480 is being used for their arcade display(which of course is windows default) so in esense you are getting that much more approximation. I don't care how many pixels are being used to achieve the 352x288 as long as the ratio is staying congruent, then the resulting display will be more accurate to the original vison! The reason the games look so "blocky" is because that is exactly what they were. Now Randy says they were never meant to be displayed that way. Well this is a point of arguement, but I counter that they were never meant to be displayed at 640x480. What he is referring to is in regards to pretty much one thing. SCANLINES. That is what smooths out the resolution...that and the fact that you have a crappy analog connection which blurs the image anyway...and of course the nature of a CRT that distorts the sides and gives a camera lense effect. Even with a flat screen you cannot get away from the fact that the nature the tube makes every pixel unequal. Oh yeah..can I mention color bleeding yet again.
The digital image shows everything that is there in its ugly glory. Many on here do not like this. That is fine, but like records it is nostalgia over reality. In this case there is even more on my side because the original code is of course DIGITAL, so you cannot say that anything is lost in the traslation.
But still if you want the "effects" you can have that too. Good artificial scanline implemenation can produce the same result. That part is easier. Why...because you are only dealing with the virticle. You get the aproximation of how wide they are, it is the same as masking your display with just about anything. But rest assured, it was still an effect, and it was NOT part of the original code. Again, get that part represented right first and then you can build up from there.
Oh yeah, the records. Yes some pops are static, and alot of them are SCRATCHES(as well as debris). They are inherent on the record from the physics of a piece of metal having contiued contact with the surface. With every play you lose quality. That is why analog cannot be captured correctly.
--- Quote ---Admittedly, even the best record with the best turntable will not have the dynamic range of a CD or DVD but that doesn't speak to the harmonics missed by the digital media. By the way, the approximations that occur in digital media occur over the entire frequency range not just at 100kHz which is about 80KHz above human hearing and in isn't even within the capabilities of the microphones used to record to pick up.
--- End quote ---
Your right on the first part, but you fail to see that even over the entire range it is all covered....more then enough(like 64 times).
I appreciate your discussion, but I hate the Randy approach. You being a musician has nothing to do with it. We all have experience, but it isn't really pertinent to the discussion.
Harmonic range is not even an issue when convered in numbers we are speaking of. Like I said, the fact that Records don't even capture all of what is in hearing range(i.e. low bass as well as high pitch sounds), but because of the hundreds of other problems, you lose hearing what is range because of physical problems.
Scratches can be heard on any player and you and I both know it. If you can hear the scratches logic should tell you that you are losing part of the sound curve. A loss that FAR out does what is in theory lost from a digital sample. A human cannot hear 24 bit sampling limitations on the harmonic range.
RandyT:
--- Quote from: genesim on July 20, 2007, 08:43:14 pm ---What do you think is happening...LCD's are picking up something that isn't there. :laugh2: Any "blocky" pixels, are an accurate representation of the original code. I doubt I will ever get this through to you. But yes, PIXELS ARE SQUARE!!! They are not little round "dots" as you think they are. Truth is truth, and your ignorance of this is astounding. Even Randy doesn't dispute this. It is all how they are handled after being presented is what is being debated here.
--- End quote ---
I guess I need to dispute that then. Here's what I wrote earlier:
--- Quote from: RandyT ---The images stored within the electronics were very boxy, but you never ever saw them that way in the arcades. Those boxy image representations were hand tweaked by the artists to take advantage of characteristics of the display and were never intended to be seen block for block. I pushed plenty of pixels in the 80's and that is the first thing I learned.
--- End quote ---
I probably could have worded that better, but I didn't. I have the benefit of knowing that graphics for early arcade games were created on graph paper. In fact, I even referred to that context when I wrote the following just a couple sentences later in the same post:
--- Quote from: RandyT ---You may have compromised what the artist put on his graph paper, but his graph paper was never intended to be viewed as such.
--- End quote ---
Those graph paper images were what the programmers used as a reference to calculate bit values for the graphics and modified those values once they saw how they were represented on the actual CRT. In my mind, I too equate that pattern of 1's and 0's with a graph paper matrix, mainly because I actually used it to create graphics in those days. However, the fact is that there was no square grid in reality. It was just data arranged in "X,Y" co-ordinate fashion inside a memory chip. Each 1 or 0 didn't need to be represented as a square. It was just convenient to do so. It could just as easily have been associated with a grid of circles, or diamonds or even fuzzy little kittens. The square was not what was important, it was the grid.
Those 0's and 1's were just data until they were transmitted to the CRTs that would be seen by millions at their local game rooms. Each 1 in data was not a square on those screens. It could not have been. Research what a coarse dot pitch, dot-triad picture mask looked like, scale it appropriately and superimpose it over a scanline. Then cut a chunk out of it and add some fuzziness. That was a pixel.
I just know that you won't address this question, but if you would, you would know why even if the displays of the day could show a true square, there's no way all pixels could have been square (all sides roughly equal.) Take a look at how many arcade resolutions there are for the same monitor in the same orientation. All of those resolutions filled the screens, yet each resolution had varying pixel counts. How could this be? The screens didn't get larger or smaller based on the resolutions, so only one thing is possible; the pixels could not always have the same vertical and horizontal dimensions.
A horizontally oriented arcade monitor screen has a 4:3 aspect ratio, so to fill it with "square" pixels one would need the resolution to be in the same proportions. Therefore, a 320x240 resolution would be based on pixels that had roughly equal vertical and horizontal dimensions. The aspect ratio would be as follows (320 / X) : (240 / X) which comes out to 4:3. Now lets look at what happens when we use a different resolution, say 256 x 224? Using the same calculation method, this gives us a 4:3.5 pixel ratio, which means that the pixels would need to be wider than they are tall to fit all of them in the same screen space. Given that visually perfect rectangles were an impossibility on the displays used, these pixels looked like little horizontally oriented oval blobs.
RandyT
genesim:
Now we are getting somewhere.
Maybe I should have prefaced better as well with PROGRAMMED pixels.
But we still have to look at a few things you are saying.
First of all, the one thing that programmers back then cared about(and now as well), cost, speed, memory.
The reason why they didn't draw "kittens" is because it took too much thinking to do so. They could have given a rats ass as to a circular pixel looking better then square because it was too much money/programming, over getting the product out on the shelves. I can understand that.
What I don't understand is the silly concept that they purposely drew the pixels that way to make them look better on a CRT display! For crying out loud. Displays like that could show movies at full resolution. The "Splash" screen on many video games looked a hell of a lot better then the moving picture. That was the bottle neck, not the display.
If one would believe what you are saying, it is kind of like saying that Doom 3 looks a certiain way on the regular game, so it is obviously the limation of the display....YET one ignores the fact that the full motion video that serves as a intro often look better on most any game. Why....because that doesn't have to be programmed like the in game...it is just played as a movie.
I do believe they made the square pixels in Pacman on purpose because of easy coordinates to work with. i.e. Maybe Pacman himself took about 8x8 pixels to move around(I don't know for sure, this is a guess). When drawn out on graph paper this was a nice number to program with accounting for dots and pathways etc.
That was the reason why it looked "blocky" and not because of programming to a supposed meager monitor which was WAY more capable then anything the programmers produced at that time or years to come.
Randy, why do you even debate this when you know all I am saying is true? You guys keep attacking LCD displays when yet the same is true for a CRT display that is at a higher resolution.
The facts are that I do know that the LCD being displayed with multiple pixels declaring one, is an approximation...BUT it is a very good one. Boxy is absolutely what the programmers meant to do, because it is what they had to work with....given chip limitations, coordinates...money....laziness....time...whatever!
I repeat, if resolution could have been better given those limitations......THEY WOULD HAVE because the displays at the time blew away anything that was put into them! This isn't that hard to grasp. This is what was true at the time, and still true now(looking back that is). Only until recent HD times have chips finally surpassed what a display can produce. This was absolutely not true back then, and any bit of normal reasoning would tell you what is true.....or I guess I wasn't watching movies when I was younger after all. Perhaps they were all just blocky displays because that was all that could be shown. :laugh2:
The arcade vga rectifies the one problem with drawing low displays....not out of this world refresh rate because every pixel of the native resolution is being used to the best advantage to the original vision. As opposed to the default MAME of given a black border when there is direct draw, or the smudgy upscaling(that of course many on here like because of the ILLUSION of better picture) of direct 3D.
The facts are that I want my resolution as close as possible to the original vision. And as sorry as it is, BLOCKY is the way it was meant to be seen, because that is what was programmed. When all the fuzzyness of an ANALOG connection and the scan lines are eliminated, this is what you get.
This isn't the LCD somehow magically making it appear like ahofle thinks. Geez dude, do you believe in the Easter Bunny as well?? Maybe there is are little gnomes that actually intercept the code and magically makes it appear blocky just for pacmans and donkey kongs. But donkey kong uninhibited was fully rendered CGI quality that would make even the real Kong faint from the realness of the graphics. ::)
xxxxxxxxxxxxx
Randy, I do appreciate your civilized post, and I am not trying to avoid anything you have written. I actually find some of it to be quite facinating. But seriously, you and I know both know that displays could have been better given an unlimited chip size/capability. The splash screens as I mentioned earlier, would often dissapoint me on some games because like in E.T. on the ATARI you would get this cool picture...then when you got to the moving game it was so much worse. Now do you honestly in your reasoning going to tell me that is why E.T. looked blocky because of the frickin' display limitations given the splash screen I mentioned?????????!!!!!!!! Or another time you actually blamed the RF connection.
This is real number evidence as opposed to the graphics on the side of the arcade that you used as an example.
RandyT:
--- Quote from: genesim on July 21, 2007, 11:56:15 am ---The reason why they didn't draw "kittens" is because it took too much thinking to do so. They could have given a rats ass as to a circular pixel looking better then square because it was too much money/programming, over getting the product out on the shelves. I can understand that.
--- End quote ---
You still missed the point, so I'm going back to the pictures. First, a reminder of what I stated in my last post so things stay in context:
--- Quote from: RandyT ---Those graph paper images were what the programmers used as a reference to calculate bit values for the graphics and modified those values once they saw how they were represented on the actual CRT. In my mind, I too equate that pattern of 1's and 0's with a graph paper matrix, mainly because I actually used it to create graphics in those days. However, the fact is that there was no square grid in reality. It was just data arranged in "X,Y" co-ordinate fashion inside a memory chip. Each 1 or 0 didn't need to be represented as a square. It was just convenient to do so. It could just as easily have been associated with a grid of circles, or diamonds or even fuzzy little kittens. The square was not what was important, it was the grid.
--- End quote ---
Here is the data for a 7x10 pixel arrow graphic in an 8-bit system:
8, 28, 62,127,28,28,28,28,28,28
This is all that exists in code for a bitmapped image. Kind of hard to see what it's supposed to be, yes? So lets fix it:
0,0,0,1,0,0,0,0,0,1,1,1,0,0,0,1,1,1,1,1,0,1,1,1,1,1,1,1,0,0,1,1,1,0,0,0,0,1,1,1,0,0,0,0,1,1,1,0,0,0,0,1,1,1,0,0,0,0,1,1,1,0,0,0,0,1,1,1,0,0
Hmm...that still doesn't tell us much, but we can at least start to see where the pixels are. Next step:
7 x 10 Binary Data Grid =
Not bad. We can start to see the shape of the arrow based on our data. While not the best visual representation, it is the most accurate in terms of our code. So now lets see what the same data looks like on graph paper, and what it would look like on any display using purely rectangular pixels:
7 x 10 Graph Paper / LCD Grid =
Well, we can certainly see the pattern more clearly, but to say that it distinctly resembles an arrow would require a fair stretch of the imagination. Regardless, it too is a perfectly accurate rendering of the data. Speaking of perfectly accurate renderings of the data, let's talk about fuzzy kittens. Remember when I said it didn't matter what shapes were used as reference when laying out the data?
7 x 10 "Fuzzy Kitten" Grid =
Awww, aren't they cute? Once again, perfectly valid references for that line of code above. If there existed "Kitten graph paper" where a designer could leave the kittens representing 1's and erase those representing 0's, he may have used this tool while coding (especially if he really liked kittens!) Interestingly enough, due to their shape, they actually create a nicer looking representation of an arrow than the normal graph paper does.
And speaking of better representations, the "coup de grace":
7 x 10 CRT Scanline Grid =
This is the most accurate visual and numeric representation of the code shown above. It was this representation of the data that was ultimately tweaked to make the best looking images possible on those displays. See how very different this one appears as opposed to the graph paper, with the very same data? When they modified the data to better the appearance of an image portrayed in this fashion, it might (and often does) look quite poor on graph paper / LCD / High-resolution monitors with enlarged rectangular pixels.
If this were not so, there would not be special graphics cards or programs like this one designed to allow these graphics to be viewed as intended. The MAME devs would also not have spent so much time adding effects that attempt to approximate that appearance.
But as others have stated before me, if playing games as they would appear on graph paper is your bag, that's fine too. But it's not "authentic" and you don't need a special video card to experience it. The photos I posted of my LCD panel being driven by a low-end motherboard video chipset is the proof I have offered. And I didn't even use the recommended method of applying the "prescale" options in MAME when I did it. Just plain old hardware stretch.
RandyT
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version