24" Widescreen CRT (FW900) From Ebay arrived,Comments.

spacediver said:
Wait, I'm not following. What would cause the distortion?

Keep in mind that my measurements are being taken based on a cropped image dead center of the camera sensor - (sensor is 4290 x 2856, and the cropped image is 87x 253).

Also, because I'm achieving perfect focus with both ronchi imaging, and phosphor imaging, the distance between the target being imaged and the lens are identical.
Click to expand...
k if you crop that much it should be fine. mathematically it matters, but probably it doesnt matter in practice.

http://images.tutorvista.com/cms/images/83/snells-law-example.PNG

if the center in perfect focus, that means the optical path length is equal at the centers
https://en.m.wikipedia.org/wiki/Optical_path_length

cant find a good image to explain... ill draw an image tonight
 
ah, I think I understand now. As the viewing angle of the phosphors (relative to the "eye" of the sensor") increases, there will be more distortions due to refraction of the light coming through the glass of the CRT.

Yea, as you say, mathematically important, but probably not practically. Though it would be interesting to measure the distortion by measuring the exact same pair of spots on the screen but at different points on the sensor.
 
As an eBay Associate, HardForum may earn from qualifying purchases.
igsux3 said:
I was told I can ask for any dimension with max width of 50cm. So for the first try I would rather order 50x40. If I fail, then I am going to order something like 50x150 and will surely be willing to sell what's left ;)
Click to expand...
can you give me contanct where I can buy it too?
 
ilovemyfw900 said:
The amount of time and money you've wasted with theses connectors and whatnot, jus NOT worth it!!!

Fury or bust buddy! They work flawlessly, go on ebay! This is one investment WORTH the $$$s!
Click to expand...
I have absolutely no complaints about image quality of converters I use.
No banding, no noise, no lag, support both 16-235 and 0-255 (the way it works is rather crude but ok for CRT) and other than Playstation4 (fixable) I had no issues with compatibility.

Later HDFury models like 3, 4 and 4s have some interresting features but I do not really need stereoscopic 3d support, scaling, or other such things. All I need is to display 720p60 for PS3 and 1080p60 for PS4. Not a big requirement, any converter will do as long as its D/A converters are decent which they are. I did comparison of HDFireFury output compared to native VGA output and it is only slightly worse. I would say it like early 00's GPU output quality like from old Radeon 9200. Same can be said for my other chinese adapter.

What guearantee do I have that later HDFury models are any better than this? And if they are is it worth it? I do not think so...
 
Last edited:
Just picked up a Hitach 46w500 rear projection CRT TV. For the fabulous price of free. :D So far I can tell it hasn't been used much as it's bright as hell. Hope to calibrate and work on it in the future. I can tell already it has the potential for some sweet image quality.
 
Interesting - as I understand it rear projection crts actually have three separate CRTs.

Wonder how the image would compare to an XBR960 (which is direct view).

Congrats :)
 
Because of the three separate CRTs, the geometry and convergence are the best I've ever seen on a CRT television. Rivals that of the GDM-FW900. Thanks! I can tell that with calibration and cleaning of the optics, this tv would probably look amazing. According to Hitachi, the tubes can resolve 1260 horizontal pixels.

Although limited to 1080i, high definition on it looks smooth and clear with zero motion artifacts. A shame these televisions went the way of the dinosaur. The place getting rid of it is my in-laws' church. They simply couldn't find anywhere to put it and have since "upgraded" to LCD long ago. It was only used for classes and the occasional youth evening. So in other words - almost never.
 
spacediver said:
Wonder how the image would compare to an XBR960 (which is direct view).
Click to expand...

I have not personally seen the XBR960, but here are my guestimates based on what I've seen from Sony direct-view CRT televisions (both standard and high-definition).

The XBR is probably sharper than this television. Its mask can resolve 1440x1080 pixels, which is very respectable for a television of its era. Most competing televisions (notably LCD and DLP) could only display 1280x720. Why Sony didn't make the XBR a TRUE 720p television, I'll never know. It's a shame they didn't because it could do it effortlessly.

Anyways - to make it short - I would guess that the XBR:
+ Better blacks (Sony TV's come with DAMN good blacks out of the factory)
+ Sharper picture
- Worse geometry
- Worse convergence (it can only adjust dynamic convergence on the horizontal axis)
- Smaller picture

Overall, I would guess the Hitachi rear projection TV is a better home theater television that the Sony based on size and the other factors mentioned above. Plus, you get all the inherent benefits of CRT - natural fleshtones, linear greyscale, deep blacks, etc. Compared to my Sony 3LCD rear projector - this television decimates it.

HOWEVER - if I ever come across an XBR with a stand and in good shape, I'll find somewhere to put it and let you know. :D Personally, I would LOVE to get a 36 inch XBR - that one also had the 1440-capable screen but in 4:3 form.
 
I wonder if you could take some of the IC's from a PC monitor and make a frankenstein rear-projection TV that could run at various frequences and resolutions, instead of only 1080i/540p.
 
EnhancedInterrogator said:
I wonder if you could take some of the IC's from a PC monitor and make a frankenstein rear-projection TV that could run at various frequences and resolutions, instead of only 1080i/540p.
Click to expand...

I don't see why you couldn't. All CRT monitors, in the end, work the same way. You could strip all of the color decoding bullshit out of it too. Most of the top-end rear projections could easily sync to 1080p.

I have to admit, thinking about having pure RGBHV for my televisions sounds awesome.

But, I'd have to do some digging to see if this is even possible. I.E. - what causes a monitor to not be multi-sync?
 
I know that 1080i Sony TV's would just recognize the different standards (480p, 720p, etc.) and use digital processing to either center in a 540p window and zoom (480p) or downscale to 1080i (720p).

Figuring out which boards/IC's to replace would be really hard, if it's even possible to craft a working multisync CRT this way.
 
EnhancedInterrogator said:
I know that 1080i Sony TV's would just recognize the different standards (480p, 720p, etc.) and use digital processing to either center in a 540p window and zoom (480p) or downscale to 1080i (720p).

Figuring out which boards/IC's to replace would be really hard, if it's even possible to craft a working multisync CRT this way.
Click to expand...

Not necessarily. All you need is a service manual and a working knowledge about monitors. There are other forums available too. I have a book on CRT monitor repairs and pretty much all monitors work the same way.

The question is - can the deflection board in the televisions do a multisync? What you're describing above suggests that the televisions operate in one resolution only.

I *believe* - from reading other forums and such (filled with actual Sony technical data) - that the DRC (Digital Reality Control - fancy term for scaler) runs in 960i for interlaced and 480p for progressive modes. But I can't confirm this, nor do I know how this affects the final output. :( It could very well do 1080i and then zoom for interlaced, and then for progressive - 540p and zoom.
 
if it have hard-coded resolutions then changing them is not possible
sad and unacceptable but by all means true

if someone have the skill to reprogram electronics then I request mod for FW900 to allow more refresh rate (raise 121KHz/160Hz limit)
 
XoR said:
if it have hard-coded resolutions then changing them is not possible
sad and unacceptable but by all means true

if someone have the skill to reprogram electronics then I request mod for FW900 to allow more refresh rate (raise 121KHz/160Hz limit)
Click to expand...

Sad and unacceptable would be true (if the deflection circuit can't do more resolutions). The top-end Sony CRT televisions can do some pretty high res. Reprogramming them would also allow for true 240p to be displayed for some old school goodness.
 
Legendary monitor, used in post production houses in the film industry. The HP Dreamcolor displays are sort of a modern day LCD based successor.
 
Minor update on focus testing (previous update here). Wanted to see how sensitive and noisy my equipment is. Set up the camera to image the ronchi ruling, and took 20 measurements at 20 different subject distances, each of which was separated by 10 microns (total of 400 measurements).

Was tricky to set up without vibration, but I got it very close to perfect (i.e. I was able to adjust micrometer with hardly any visible wavering of the image).

Below is the setup. The bright area is a lightbox, to which is taped the ronchi ruling (invisible due to bright light).

o91ugp.png


Here are the data. Error bars are standard errors of the mean (n = 20).

I hid the first few data points as there was some noise at the beginning. Note that the angle of the slanted edge was almost 7 degrees. If I use a steeper angle (say 2 degrees), I'll likely be able to get more precision due to increase in the supersampling factor. But even with that edge angle, the image processing is clearly able to distinguish subject distances of 10 microns. Tomorrow if I have time, I'll do the same test but with one micron adjustments. This will be trickier as you need very fine motor control with your fingers, and it's tricky reading the vernier scale on the micrometer.

jpy7vo.png


Anyway, upshot is that this shit works. Next step is to set up the camera so it can image the monitor without any vibration. Will have to move monitor to the smaller table to do this. Also, the ronchi ruling measurements can serve another purpose: Since the ronchi ruling is a perfect square wave target, that means when I image it at peak focus, I can calculate the modulation transfer function (MTF) of my camera (which is due to a combination of the lens and sensor). When I image the display, the image is going to be "filtered" through the MTF of the camera, so it's not really the objective image that the display is generating. In other words, if the display was able to show a perfectly sharp transition from black to white, the image itself, even when perfectly focused, would be slightly blurred (the exact amount of blurring is what the MTF of the camera represents). So in principle, I can factor out the camera MTF and provide objective measurements of the display. This is not so necessary when comparing two conditions (i.e. 2 displays, 2 video cards, 2 refresh rates, etc.), but it is necessary if you want to objectively quantify things like the way a single pixel is rendered (i.e. how focused is that pixel).
 
The phosphors degrades over time hence life is measured in hours
To find one operating at peak performance with no degradation is tough
They stopped making them around 10 years ago, mines a 2001 with low hours
Good tech stuff
 
jbltecnicspro said:
Sad and unacceptable would be true (if the deflection circuit can't do more resolutions). The top-end Sony CRT televisions can do some pretty high res. Reprogramming them would also allow for true 240p to be displayed for some old school goodness.
Click to expand...
as a warmup change FW900 limits from 30-121kHz to like 15-150kHz
if it can be done then I could believe in reprogramming other TV's

as I see it with HDTV you have things to solve:
- there is a scaler to get rid of
- deflection circuit made for single resolution might be simplified and if so would need serious rework


So no simple firmware reflash. FW900 on the other hand have chip somewhere which is probably simple I2C flash and could be easily reprogrammed with with anyone with most basic electronic skill and cheap equipement. All that is needed is someone to do it first, make tutorial and modified firmware available.

If you can do it then I would personally take your opinions on reprogramming HDTV's to support multi-sync from 240p up to 1080p (or higher) much more seriously :eek:
 
Last edited:
repeated the experiment but with single micron adjustments. 17 diff subject distances, with 20 measurements at each. Very tricky adjusting the micrometer with that precision, as the vernier scale is in an awkward position relative to the rest of my setup, so it's hard to see it clearly. The first few measurements were smooth, as were the last few, but in the middle I had to make a few readjustments, and the data seem to reflect this, although I think around peak focus, there is very little change in the edge spread function compared to when you start to go away from peak focus.

note: the sigma values are larger compared to the previous experiment, as the edge angle here is much steeper. Because of this steeper angle, the supersampling factor is greater, which means the edge spread functions consist of many more data points, hence the larger sigma values.

Also, this data is impressive considering it's being generated from 8 bit image data. Ideally I'd be able to take RAW images and automatically process them but that would a seriously difficult challenge, and I'm not sure it's worth it in my case, given the high precision I've already been able to obtain.

For what it's worth, when using my eye, rather than the image processing, I can discriminate focus changes with about 20 micron precision. So the supersampling code gives me a huge boost in precision compared to using eye alone.

nfmq1z.png
 
Last edited:
XoR said:
If you can do it then I would personally take your opinions on reprogramming HDTV's to support multi-sync from 240p up to 1080p (or higher) much more seriously :eek:
Click to expand...

Dude - relax. :) I'm just dreaming out loud. I haven't done any looking in to this yet. I don't even have a Sony HDTV to try this so the point is moot. :(
 
XoR said:
can you give me contanct where I can buy it too?
Click to expand...

http://ovro.pl/, jak pytałem przez telefon to mówili, że przycinają do dowolnego rozmiaru z maksymalną szerokością 50cm.


------

Thanks to your advices guys I found my very old laptop with Windows 2000 SP4 installed. As supposed, WinDAS works perfectly with no single symptom of wrong readouts from monitor that I had on win 8.1 machine. Only issue I have is wrong scaling of some fonts as seen below:

windas_problem.jpg


I have also had this issue on Win 8.1 machine, but they were readable after all. On Windows 2000 I can't see a damn thing. Any tip where to look for it?
 
having some success converting my RAW file into RGB's. This processing is all done from scratch, using custom transformation matrices. Still gotta work on expanding the full dynamic range, and will experiment with dithering to try to maintain the 14 bit information in the 8 bit image, but at least I'm in the right ballpark.

Here's the original image that I loaded onto the display, and took a photo of:

2nw1fzp.png


And here's the processed image of that photo that I took. I do find it weird that in the camera image, the phosphor stripes are clearly visible, but when I view the other image I don't see them, even though the spatial scale is roughly the same. Might be due to the raised black level in the image or something. Or perhaps it's an artifact from the way I subsampled the Bayer array. I'll do some more tests, taking photos of things other than my CRT, and see if the artifact persists.

I'd be curious to see someone else take a regular photo of the original image and post their photo here. Here's the original full version.

1199l6o.png
 
Last edited:
Hi, How can I check the "time of usage" information (the last line)? When I press menu button for a while I am getting "information" but with only first 3 lines. I managed to get one FW900 and 4 units of GDM F520.

How I can make this thing pop out? Picture from internet
m3PJS9c.jpg
 
idenieide said:
Hi, How can I check the "time of usage" information (the last line)? When I press menu button for a while I am getting "information" but with only first 3 lines. I managed to get one FW900 and 4 units of GDM F520.

How I can make this thing pop out? Picture from internet
m3PJS9c.jpg
Click to expand...

These monitors will display the manufacturing dates, coded as MANUFACTURING YEAR (2003) - MANUFACTURING WEEK (08), and not "time of usage".

There is a procedure for accessing the information window and it is explained on the user's manual (you can download a free copy online). If you can't find the info in there, PM me...

Hope this helps...

UV!
 
igsux3 said:
http://ovro.pl/, jak pytałem przez telefon to mówili, że przycinają do dowolnego rozmiaru z maksymalną szerokością 50cm.
Click to expand...
Jak to fajnie zobaczyć w tym wątku prawdziwy język :p

widzę że mają folie o przeźroczystości 73% więc znacznie większą niż ta którą używam która ma 48% czyli czerń w dzień będzie 2x jaśniejsza ale za to sam monitor będzie ok 50% jaśniejszy a więc i nie trzeba będzie go tak żyłować aby uzyskać te 100cd/m2. Imho słaba zamiana ale przynajmniej da się kupić bez kleju więc będzie można łatwiej założyć taką folię. Oryginalna powłoka FW900 ma chyba podobną przeźroczystość do tego polaryzatora.

Może zadzwonię jutro i się zapytam o te folie dokładnie co i jak choć narazie nie będę kupował bo mam jeszcze dużo tej folii z klejem do wykorzystania. Dalej używam tej co założyłem za pierwszym razem. Wyszło to chu... ale jak gram to jestem tak skoncentrowany na grze że nie widzę żadnych wad :) Być może da się ten klej co jest jakoś łatwo usunąć jakimś rozpuszczalnikiem, choć tą opcję zostawię na ostatni kawałek jeśli druga próba okaże się porażką.

Rozumiem planujesz kupić wersję bez kleju, tak? Jak to będziesz przyczepiał tj. czy masz już znaleziony klej. Pytam z ciekawości.

jbltecnicspro said:
Dude - relax. :) I'm just dreaming out loud. I haven't done any looking in to this yet. I don't even have a Sony HDTV to try this so the point is moot. :(
Click to expand...
Ok, I get it :p

it is just that since certain user with nickname beginning with "r" and ending with "7" I am more oversensitive to dream stories about making monitors, tweaking monitors and such kind of stuff :eek:

XiP said:
Wow after all these years this thread is still going! :cool:
Click to expand...
It is because FW900 is still best gaming monitor ever produced :)
 
XoR said:
There is visible difference in flickering between 85Hz and 96Hz. Eyes lifespan is always more important than monitor lifespan. And its better in games to run fastest as it is possible at given resolution.
Click to expand...


There is no visible flicker at 85 Hz for me at all. I'm 45 and have been staring at 75Hz to 85Hz for over 15 years daily and my eyesight is only now starting to deteriorate slightly.
I can see like an eagle from 2 feet out though, lol, but that's due to age. Years back i ran at 1680 x 1050 at 75Hz and there was a little slight flicker, but at 85Hz there is none whatsoever.
I also don't stare at white screens which makes a difference, all my Windows UI is custom dark, most programs i reskin the UI if possible, my Web style sheets everything is dark or black, like this forums background, so that also helps huge.
As far as running games, my video card only does about 85 fps max in anything in the last 10 years anyway so i run vertical sync at half refresh in the driver and it's basically locked at 42 fps in modern games, so unless i'm playing Quake or something old then i unlock it.
If i had my refresh at 96Hz then i would get stutter more often, so it's a sweet spot. 42fps is perfectly fine for me in rpgs and such.
 
LAGRUNAUER said:
These monitors will display the manufacturing dates, coded as MANUFACTURING YEAR (2003) - MANUFACTURING WEEK (08), and not "time of usage".
Click to expand...

I managed to find service manual. To get this "information" bar I have to press menu for few seconds.

I found this Japanese auction of F520 where seller is saying that the monitor was working only for 3.5h and he is adding this photograph of extended information bar with extra 2 lines saying that the last one stands for "time of usage". That's why i asked. http://page.auctions.yahoo.co.jp/jp/auction/208038573
 
Last edited:
2nd don't forget to adjust your focus pods with this tool thru the air vents in rear

Good Luck to all enthusiasts
 
idenieide said:
I will have for sale 2 of them in Poland for the same price I bought them.
Click to expand...

Oh, I see we have some nice Polish team in here ;) Where do you live? I might be interested in seeing these F520s.

XoR said:
Jak to fajnie zobaczyć w tym wątku prawdziwy język :p

widzę że mają folie o przeźroczystości 73% więc znacznie większą niż ta którą używam która ma 48% czyli czerń w dzień będzie 2x jaśniejsza ale za to sam monitor będzie ok 50% jaśniejszy a więc i nie trzeba będzie go tak żyłować aby uzyskać te 100cd/m2. Imho słaba zamiana ale przynajmniej da się kupić bez kleju więc będzie można łatwiej założyć taką folię. Oryginalna powłoka FW900 ma chyba podobną przeźroczystość do tego polaryzatora.

Może zadzwonię jutro i się zapytam o te folie dokładnie co i jak choć narazie nie będę kupował bo mam jeszcze dużo tej folii z klejem do wykorzystania. Dalej używam tej co założyłem za pierwszym razem. Wyszło to chu... ale jak gram to jestem tak skoncentrowany na grze że nie widzę żadnych wad :) Być może da się ten klej co jest jakoś łatwo usunąć jakimś rozpuszczalnikiem, choć tą opcję zostawię na ostatni kawałek jeśli druga próba okaże się porażką.

Rozumiem planujesz kupić wersję bez kleju, tak? Jak to będziesz przyczepiał tj. czy masz już znaleziony klej. Pytam z ciekawości.
Click to expand...

No miło, szczególnie, że wychodzi na to, że idenieide też od nas :) Co do polaryzatora to tak naprawdę nie widzę praktycznie żadnej korzyści z jego właściwości polaryzacyjnych, ale jak już przyciemnia i nie kosztuje nie wiadomo ile to wezmę. Mam monitor ustawiony na 115cd i z testerami jasność spada do ok. 46cd, więc jest nawet gorzej niż deklarowane 48%. W tym wypadku chyba też się zainteresuję tą 73%, ale na stronie mają niezły bałagan.

To jest folia 73% bez kleju:

http://ovro.pl/sklep/folie-filtry-optyczne?product_id=61

73% z klejem:

http://ovro.pl/sklep/folie-filtry-optyczne?product_id=67

Ty wziąłęś tę ostatnią? Czy 48%? Mógłbyś pokazać, jak u Ciebie wygląda ten problem z klejem? Ja myślę o folii bez kleju, którą po prostu mocno naciągnę na szkle i przycisnę ramką monitora. Jeżeli nie da rady, rozejrzę się za klejem UV. A gdzie mieszkasz? Jeżeli w okolicy to na start mógłbym odkupić na próbę tę klejącą się :) Ja stacjonuję we Wrocławiu.


-- international readers please forgive us for this Polish chatting. We discuss about polarising foils that are sold in Poland and I doubt that anyone would be interested in them. But if so I would be glad to help. When I get mine I will of course share my results.
 
XoR said:
it is just that since certain user with nickname beginning with "r" and ending with "7" I am more oversensitive to dream stories about making monitors, tweaking monitors and such kind of stuff
Click to expand...

Hello.
 
Have taken some phosphor images at centre, mid centre, and edge. Will analyze them later this evening for dot pitch measurements. For now, here's a teaser:

2ciicfa.png
 
You must log in or register to reply here.
Back
Top