Building 3BP1 tube NBTV Anderson monitor II

Moderators: Andrew Davie, Steve Anderson, Dave Moll

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Harry Dalek » Tue Feb 07, 2017 5:58 pm

Klaas Robers wrote:Yes, this looks interesting. But it would look as if you had a oscilloscope tube with a white phosphor that you could make RGB. This is not true, because white phosphor is not white phosphor, but a mixture of blue and yellow. These two colours together give an impression of white, but there is almost no red in it and almost no green. So an R-G-B colour wheel will give a very poor colour rendering, especially in red and green. People will alway show yellowish faces and skins.

Besides that the colour repetition frequency of about 4 Hz is far too low to give your eye a mixture impression of red, green and blue. The already enoying flicker of 12.5 Hz will be 3 times as enoying.


So Klaas the flicker would be similar to this ...these were used a on web site to show CBS colour but i converted them to anigif...as with splitting the colours on your Avatar ..the flicker is pretty pronounced but for demonstrating its interesting and it think it would be fun to try .

Troy's pretty good at all this ,i think he would try just for the experiment .

Perhaps going non standard might improve the flicker .

So all white phosphor Crts a mix of blue and yellow ? as the Colour again seemed to work using a b/w tvs at the time for the CBS 50's system .

To test what ever crt what light colour it gives sending the light from the crt display via a prism show what colours it could work at ..
Attachments
testcolour.gif
testcolour.gif (316.33 KiB) Viewed 639 times
testcard.gif
testcard.gif (137.63 KiB) Viewed 639 times
anigif4.gif
anigif4.gif (33.91 KiB) Viewed 637 times
11.jpg
11.jpg (13.24 KiB) Viewed 635 times
333.jpg
333.jpg (15.59 KiB) Viewed 635 times
22.jpg
22.jpg (10.13 KiB) Viewed 635 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 3335
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Klaas Robers » Thu Feb 09, 2017 5:05 am

No Harry, the flicker in the pictures of your animated colour GIF is faster than the flicker that you would get if R, G, B follow each other at a speed of 1/12th second, so the Red comes 4 times per second, the Green comes 4 times ...... and so on. Each coloured picture should stand 80 msec. What I see in your posting is R, G and B alternating absolutely much faster.
And then there is the usual line flicker that we are used to for NBTV Nipkow monitors.
As far as I know it is possible for an animated gif to define the time of display. Don't ask me how.....

May be that the black and white gif (is this only red and blue?) looks better. I can count the number of "Reds", and that is how it will look like. Can you make the ËBU test picture with only green and then a black frame in sted of Red and black in stead of Blue? And then 0.08 sec for each frame? This should flicker clearly at 4 Hz.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1291
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Harry Dalek » Thu Feb 09, 2017 8:59 pm

Klaas Robers wrote:No Harry, the flicker in the pictures of your animated colour GIF is faster than the flicker that you would get if R, G, B follow each other at a speed of 1/12th second, so the Red comes 4 times per second, the Green comes 4 times ...... and so on. Each coloured picture should stand 80 msec. What I see in your posting is R, G and B alternating absolutely much faster.


I use a program call photoscape its free ...mainly use it from viewing adjusting photos ...it does animated gifs and i just noticed it shows the time that gif was 0.03 of a sec 30 ms////
I will adjust to 0.08 of a sec 80 ms .
This is really my first ever gos at mixing colours and another mistake is the percentage of red green and blue were on full where you need to mix i think 30% red, 59% green and 11% blue so i will see if i can improve here ....did not look right when i changed the percentage Oh well more learning .

And then there is the usual line flicker that we are used to for NBTV Nipkow monitors.


Yes to me its not to bad Bairds 3 colour 15 line 10 frames a sec must of been much worse ..

As far as I know it is possible for an animated gif to define the time of display. Don't ask me how.....


To me i would have no idea but you gave me some numbers to work with so let me know if it looks closer ....

May be that the black and white gif (is this only red and blue?) looks better. I can count the number of "Reds", and that is how it will look like. Can you make the ËBU test picture with only green and then a black frame in sted of Red and black in stead of Blue? And then 0.08 sec for each frame? This should flicker clearly at 4 Hz.


I will put up the first gif as that .80ms b/w +green
next red + green 80ms
balloon picture used
balloon at 80ms and last same at 40ms which works much better but i see your point at 80ms not so good.....BTW photoscape animated gif fastest is 10ms it can do .
Attachments
80ms.gif
80ms.gif (15.46 KiB) Viewed 614 times
redgreen80ms.gif
redgreen80ms.gif (14.33 KiB) Viewed 614 times
images.jpeg
images.jpeg (5.06 KiB) Viewed 614 times
ballon80ms.gif
ballon80ms.gif (26.87 KiB) Viewed 614 times
balloon40ms.gif
balloon40ms.gif (26.87 KiB) Viewed 614 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 3335
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Klaas Robers » Fri Feb 10, 2017 6:51 am

Indeed, the 80 msec balloons are no more colored. The integration time of the eye is too short for it. The 40 msec is for 25 frames (B+W) per second, and there it is difficult to see the colours, if any.

The attenuation is not needed for colour, then it is 1 : 1 : 1. These factors are the subjective brightness vor a B&W picture made of the 3 colour picures. We see blue as a dark colour, so on a B&W picture blue parts should have a low brightness.
But now you are not making black and white but colour.

Try to make a long exposure photo of your alternating colours GIF and you will see that it is correct.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1291
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Andrew Davie » Fri Feb 10, 2017 9:02 am

In a different endeavour I did a LOT of experiments getting "full colour" from a single-colour display. On the vintage Atari 2600 video game machine, there is but one colour available for pixels (on or off, a colour of your choice but the same for the whole scanline). To get colour images, I did some experiments with alternating colour frames (at 60Hz) so you saw each colour frame 20 times per second, and mixing of red/green/blue effectively gave you 8 colour variations total. You could not vary intensity of the pixels - so, one colour as noted. By combining the colour alternation with stippling (floyd-steinberg) of the original image, the impression of colour intensity could be achieved, and that's where I left the frame-sequential colouring. This worked but was flickery and annoying because the brain sees those huge fields of single colour, pretty much up to 30Hz or so.
The next step was to use line-sequential colouring - that is, alternate each scanline by red/green/blue. Still the same limitations as above, but the colours were varied on each line and the red/green/blue sub-frames were interlaced with each other. This greatly improved the visuals of the image. No longer did we see huge fields of a single colour, and although the frames were displaying at the same rate (60 Hz), the colour flicker was now gone and we had effectively the same coloured image (with the stippling as mentioned). But the vertical resolution was reduced, because you only saw each colour on every third scanline.
The next and final "solution" was what I called "interleaved chronocolour" - ICC. This involves changing the order of red/green/blue lines on each successive frame, but still using the above method of lines having successive red/green/blue colouration. So on frame 1, you would see scanlines coloured red/green/blue/red/green/blue... but on frame 2 you would see them coloured green/blue/red/green/blue/red and then on frame 3 blue/red/green/blue/red.... in other words, the coloured scanlines appeared to shimmer, or if you looked closely, roll. So for any particular scanline, it was back to the frame-sequential rate, but nonethless still included line-sequential colouring.
So, this worked *BEAUTIFULLY* and allowed "full colour" images on the Atari 2600 - a machine capable of only a single colour on any line with no intensity variation.
I've mentioned all this because I wonder if a colour wheel rotating at a speed such that individual scanlines (rather than frames) are alternating colours, and modifying the video signal so that the colours are thus interleaved as described. You will have your colour image at a low frame rate without the annoying frame-sequential flickering.
Cheers
A
User avatar
Andrew Davie
Frankenstein was my uncle.
 
Posts: 1256
Joined: Wed Jan 24, 2007 4:42 pm
Location: Hobart, Tasmania, Australia

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Klaas Robers » Sat Feb 11, 2017 5:58 am

Interesting observations Andrew! I always have to think in these cases of the interlace of the English black and white TV system with 405 lines. When you watch this you see pictures in 200 lines, where the lines scroll upwards or downwards, but is remained a course picture. The interlace only works if you keep your eyes fixed to a certain position on the screen, but as soon as you did this your eyes were caught by the scrolling line pattern. Your Atari-observations were completely in line with this.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1291
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Andrew Davie » Sat Feb 11, 2017 10:16 am

Here is an image of a screenshot of ICC in action on a game I wrote a few years back on the Atari 2600.

1.jpg
1.jpg (151.2 KiB) Viewed 590 times


The little striped red man is NOT subject to the colour limitations, so ignore him. But the rest of the screen is using the RGB interleaved but NOT rolling the colours from frame to frame. The resolution of the '2600 is a painfully low 40 pixels across, but the really interesting thing to observe here is that although the screen as a whole is quite colourful, there is only ever a single colour on any one scanline. It looks like there are different coloured "things" on the same "line" - e.g., dirt, walls, rock - but those "things" are meta-scanlines. In other words, the eye doesn't really care at the level of scanline detail - it looks for objects and thus the single-colour limitation is not perceived. The colours are made by combining on/off pixels and colour changes on successive lines to form "triplets" (not necessarily RGB could be any three repeating colours) to give the visuals that I wanted. This meant completely flicker-free and no colour rolling - of course the vertical resolution was "reduced" because any one pixel required 3 scanlines of colour to define it. But to the eye it WASN'T reduced, because there are still all those individual scanlines, so the eye still sees a full-resolution colour screen even though, as noted, just a single colour of a single intensity on any one scanline.

Here's a bit of a closer look, noting the different "RGB" used that from the above image...

2.jpg
2.jpg (156.06 KiB) Viewed 590 times


There's a few youtube videos of this technique/game, like this one... https://www.youtube.com/watch?v=uiGBVBnFHQ4


So it would be really interesting to see just a static R/G/B colour effort in 30 lines. No rolling across successive frames, but just a single colour filter with RGB stripes. I reckon it might actually look surprisingly OK :)
User avatar
Andrew Davie
Frankenstein was my uncle.
 
Posts: 1256
Joined: Wed Jan 24, 2007 4:42 pm
Location: Hobart, Tasmania, Australia

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Harry Dalek » Sat Feb 11, 2017 5:26 pm

Wasn't bairds original 3 colour system 15 lines 3 of these on the same disk with a colour filter for each of the 3 15 line spirals .
Garys video2nbtv v3 has a this 15 line 10 hz frame colour system but theres no way to view it on pc yet so far you have to build a monitor i suppose to use it .
Good work on the computer game colour programing Andrew thats all beyond me makes my head hurt thinking about it :roll:
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 3335
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Building 3BP1 tube NBTV Anderson monitor II

Postby Klaas Robers » Fri Feb 17, 2017 6:58 am

In the first post of sequence 1 of this thread Aussie Bloke suggested to use the 3BP1 next to NBTV also for standard scan 625 line TV, as well as for SSTV. To my opinion the 3BP1 is not suited for both.
- For 625 line TV the sharpness of the spot is far too low to cope with the fine line structure of standard TV. You will get an unsharp and blurred picture on this small screen. Yes, you can see more or less what is happening, but all details are completely invisible.
- For SSTV the retention (after glow) of the phosphor is far too short. It takes 7.2 sec for a complete picture! You will see no more than a spot running over the screen. Yes, if you make a photo of the screen and you manage to get an exposure time of 7.2 sec, then you get a usefull photo, but our eyes have not the integration capabilities of a chemical photo camera.

And even the long afterglow of the P7 phosphor layer of the radar tubes is too short to get an image of the whole picture. You see a fading band of 1/3 to 1/2 of the total picture height, in which you can recognise the crucial structures. After two or three scans you have got an impression of the total picture in your mind.

But the 3BP1 is a nice tube for an oscilloscope.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1291
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Previous

Return to Building 3BP1 tube NBTV Anderson monitor II

Who is online

Users browsing this forum: No registered users and 1 guest