English Amiga Board


Go Back   English Amiga Board > Support > support.Hardware

View Poll Results: Any full-HD Scart LCD TV in a store displays Amiga PAL RGB fine if you plug it in.
Yes 3 14.29%
All of them did, even before 2010. There was no trouble even then. 1 4.76%
Yes, but there is more to it. It's not quite that simple. (Specify...) 3 14.29%
No 14 66.67%
Voters: 21. You may not vote on this poll

 
 
Thread Tools
Old 02 October 2013, 16:03   #41
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Quote:
Originally Posted by demolition View Post
That will only tell you the screen refresh frequency, not the actual frequency of different frames.

Try yourself - gif animation, four frames, 4 colors - play on Amiga and measure frequency. Lower pattern for photosensor or camera, upper also for "naked" eye.
Beware - content have rapid flashing large bright areas - should be avoided with people with neurological issues (im serious).


Click image for larger version

Name:	320x256.gif
Views:	225
Size:	6.2 KB
ID:	37012

Click image for larger version

Name:	320x512.gif
Views:	211
Size:	9.6 KB
ID:	37013
pandy71 is offline  
Old 02 October 2013, 17:03   #42
demolition
Unregistered User
demolition's Avatar
 
Join Date: Sep 2012
Location: Copenhagen / DK
Age: 41
Posts: 4,158
I was thinking about measuring the frame rate of games, however if you're just testing the SCART connection it would make sense to create some artificial signals on the Amiga as it would make artifacts clearer and easier to measure.
Some/most TVs have adaptive deinterlacing so a test like the above would give different results compared to a situation where two neighboring frames are almost similar.
demolition is offline  
Old 02 October 2013, 18:49   #43
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Quote:
Originally Posted by demolition View Post
Some/most TVs have adaptive deinterlacing so a test like the above would give different results compared to a situation where two neighboring frames are almost similar.
They have adaptive but modes of the deinterlacing - deinterlacing anyway must be performed - adaptivity is related to optimal algorithm for particular video not deinterlace itself as a principle - PDP or LCD are progressive displays and before displaying they must have deinterlaced video, also deinterlacing is mandatory before resizing so unless you provide video with native resolution to display matrix (only digital interface, only DVI, HDMI under requirement that all pixels are shown - overscan show).
pandy71 is offline  
Old 02 October 2013, 18:52   #44
ajk
Registered User
ajk's Avatar
 
Join Date: May 2010
Location: Helsinki, Finland
Posts: 1,323
Well I recorded some Turrican II intro and it's clearly visible that the image updates 50 times per second as expected You can see it just by eye anyway. This TV is from exactly 2010 I believe and I connected the Amiga via SCART.
ajk is offline  
Old 02 October 2013, 19:06   #45
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
The "all 2010+LCD TVs deinterlace progressive modes to interlaced doubleres modes" claim is not mine. In fact, I hope those claims are false. But it needs to be demonstrated.

GIFs cannot be trusted to play at a certain rate without measuring IT. The animation unit in GIF is called a jif and is not divisible into 50ths of a second I think. The examples you provide might work if they didn't animate at a way too slow and unrelated rate.

I'll try to code a TV test program this evening that reveals interlacing to the naked eye. I might use the first GIF pattern but at 50Hz, if it works.
Photon is offline  
Old 02 October 2013, 19:15   #46
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 46
Posts: 25,007
There is no need to show any proof. Some TVs support PAL/NTSC "progressive" mode, some don't. Simple.
Toni Wilen is online now  
Old 02 October 2013, 21:14   #47
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Vlad.H.

Quote:
Originally Posted by Photon View Post
The "all 2010+LCD TVs deinterlace progressive modes to interlaced doubleres modes" claim is not mine. In fact, I hope those claims are false. But it needs to be demonstrated.

GIFs cannot be trusted to play at a certain rate without measuring IT. The animation unit in GIF is called a jif and is not divisible into 50ths of a second I think. The examples you provide might work if they didn't animate at a way too slow and unrelated rate.

I'll try to code a TV test program this evening that reveals interlacing to the naked eye. I might use the first GIF pattern but at 50Hz, if it works.
GIF was used only as i can't export to ANIM - you can split on PC (Amiga) gif to 4 frames (as whole sequence is 4 frames long) - try play sequence with various speeds (50 and 25 fps) - i can provide interlace versions too - this is not a problem.

Pattern is very simple and it works without problems - it was designed to measure frame rate in electronic way (and detect skipped/missed frames).
pandy71 is offline  
Old 03 October 2013, 00:10   #48
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
Quote:
Originally Posted by Toni Wilen View Post
There is no need to show any proof. Some TVs support PAL/NTSC "progressive" mode, some don't. Simple.
Yes, but the question in topic is whether you can go into a store and easily find one that doesn't.

Because the difference is output quality is quite big between the two. One is blurry and one is "neato!"
Photon is offline  
Old 03 October 2013, 00:15   #49
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
OK, some (very strange!) results are in!

Feel free to use the TV Tester software. It uses 640x256 mode. I will not answer any questions about it that are answered in the README.TXT!

I'm not sure it's optimal, but it should eliminate display "afterglow" with the palettes, and it does output 50 distinct pictures in 1 second.

ZeroX took some pictures of his Samsung LCD TV (thanks!!):

Shutter speed 1/30:

Left - 3 numbers in < 1/25 second. May be "afterglow". Strangely, they aren't rasterized, like the checkerboard on the right.
Right - COMPLETELY odd/even rasterized, deinterlaced checkerboard. No flicker, steady as a rock. This means the 640x256 picture IS converted to 640x512.

Shutter speed 1/250:

Left - this is mysterious. Regardless of 25 or 50 pics/s output, there should be only 1 number. Likely explanation: the timing of the shot overlapped a frame switch.
Center - No diff from 1/30.
Right - unevenly flickering Indivision AGA 60Hz output to the 50 or 25 pics/s TV. This is true progressive output, but at a "frameskip" rate.


I was hoping to reveal 25 pics/s TVs by the numbers appearing to "pair together" vertically or horizontally, and that this would be visible to the eye or possible to photograph. (False I would say for this TV?)

Also, that the checkerboard would appear solid - odd/even line rasterized or not. (True - this TV deinterlaces "some things" at least from 640x256@50 to 640x512@25).

On a 50Hz CRT, the checkerboard would flicker smoothly (True), and on a display in a >25 (60Hz) progressive mode, it would flicker unevenly (True). These displays do not deinterlace progressive modes.

Last edited by Photon; 14 October 2017 at 21:37.
Photon is offline  
Old 03 October 2013, 17:52   #50
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Interlace version for previous pattern:
GIF, 2 frames, TFF (Top Field First).
Attached Thumbnails
Click image for larger version

Name:	320x512i.gif
Views:	163
Size:	4.9 KB
ID:	37023  
pandy71 is offline  
Old 03 October 2013, 20:00   #51
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
If this is to test if TVs deinterlace, no pattern is needed? The screen flickers quite distinctly if they don't.
Photon is offline  
Old 03 October 2013, 20:57   #52
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Quote:
Originally Posted by Photon View Post
If this is to test if TVs deinterlace, no pattern is needed? The screen flickers quite distinctly if they don't.

This is to test 2 things:
- TV framerate (as some people reports 60Hz used to display 25(50Hz) source this pattern will produce beat frequency (perceived flashing will have some black/white areas irregulary distributed)
- deinterlacing (after deinterlacing pattern structure will be different than displayed without deinterlacing)

On ECS/AGA video mode with progressive 25Hz framerate can be created easily (on OCS this is also possible by writing to VHPOSW with CPU).
pandy71 is offline  
Old 04 October 2013, 18:51   #53
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
(Feedback, feel free to provide more information on how it works technically and I may use it in a test program.)

But this doesn't reveal if all the 50 distinct pictures/s the Amiga outputs are shown by the TV. It's the negative to half of my test, you could say. Disproving the negative doesn't prove the positive. (Note:On a slow LCD, using black and white may very well hide flicker in the black-to-white timing. Hence my lower contrast color.)


Either that, or it seems to test if the TV can detect and show an interlaced picture straight up? They all can, it seems. It's the deinterlacing of progressive modes we want to detect, not if the TV can display interlaced video, it can because all DVDs output it. This seems just a hi res picture (vertically) being interlaced by the Amiga on output as normal.

I don't know what you mean by 'progressive 25Hz framerate', it's not used in any software and so an Amiga user will never see such a 'mode'.

Maybe the misunderstanding lies between Hz and FPS. Let's call it distinct pictures per second, or DPPS, instead - by distinct I mean that the picture content changes; that it's not the same picture displayed in a different way. Ie. Interlace has 25 DPPS because Denise outputs the same picture twice in succession. (Of course you can change the picture 50 times per second, but that's the problem defined, isn't it? On the "deinterlace-to-25-DPPS-TVs", the changes get combined and you either get choppy banding or choppy blur. On the proper ones you get 50 pictures out.)

Last edited by Photon; 04 October 2013 at 18:58.
Photon is offline  
Old 06 October 2013, 20:34   #54
moijk
License to Bubble

moijk's Avatar
 
Join Date: May 2010
Location: Grimstad, Norway
Age: 44
Posts: 293
The day i'm going to get a LCD tv, I wonder what the guys in the store will say when i have the amiga under my arm and ask if i can test it on my potential new purchase.
moijk is offline  
Old 06 October 2013, 21:19   #55
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
Not any stranger than bringing some Nintendo or Playstation console...? Besides, the public has heard about the "retro movement" by now I think

If you bring my program on a floppy inside the Amiga, you can run it to test the TV. Gaze steadily on the checkerboard on the right.

1) If the checkerboard flickers unsteadily, it's a 60Hz (frameskip) TV, which is bad. (Make sure it wasn't due to you moving your eyes.)
2) If it becomes steady (either completely solid box or steady rastered pattern as in the big photo above) then it deinterlaces to 25Hz which is OK for high resolution work but not for action games.
3) If it flickers steadily, however, the TV displays progressive (320x256 or 640x256) modes properly!
Photon is offline  
Old 07 October 2013, 12:17   #56
pandy71
Registered User
 
Join Date: Jun 2010
Location: PL
Posts: 1,805
Quote:
Originally Posted by Photon View Post
(Feedback, feel free to provide more information on how it works technically and I may use it in a test program.)

But this doesn't reveal if all the 50 distinct pictures/s the Amiga outputs are shown by the TV. It's the negative to half of my test, you could say. Disproving the negative doesn't prove the positive. (Note:On a slow LCD, using black and white may very well hide flicker in the black-to-white timing. Hence my lower contrast color.)
It should reveal all problems except exactly full loop loss i.e. 4 pictures gone, this can be solved by time stamping each frame, also something more nice from eye perspective can be made - like linearly growing horizontal bar together with frame numbers - from 1 to 25 for example.
Even without scope eye should have no problem to detect problem as we (eye+brain) see such thing relatively good (perhaps except younger generations that to be hones don't care about frame jitter or frame rate as they became accustomed to poor PC video quality where 24 then 26 or even sometimes variable video form 20 to 30 fps is common - just check tablets, smartphones - people don't care - more important is that this is glossy, fancy and fashionable... aaah i forget, blink! it must blink and nicely glow then is cool...).

Quote:
Originally Posted by Photon View Post
Either that, or it seems to test if the TV can detect and show an interlaced picture straight up? They all can, it seems. It's the deinterlacing of progressive modes we want to detect, not if the TV can display interlaced video, it can because all DVDs output it. This seems just a hi res picture (vertically) being interlaced by the Amiga on output as normal.
First of all to analyze such problems (perceived frame rate at the TV not at the Amiga video output) you need some higher resolution system than 50Hz - or oscilloscope or high framerate camera - personally i would consider some hi-speed camera like for example CANON HIGH SPEED EXILIM family.
Or ultra low-cost optical mouse sensor.
However i point once again - knowing how business world going today i doubt that they expect on SCART input something else than video signal with BT.601/656 parameters.

Quote:
Originally Posted by Photon View Post
I don't know what you mean by 'progressive 25Hz framerate', it's not used in any software and so an Amiga user will never see such a 'mode'.
Yes, i know that however this will be true progressive 25Hz on Amiga (third - a bit fake mode).

Quote:
Originally Posted by Photon View Post
Maybe the misunderstanding lies between Hz and FPS. Let's call it distinct pictures per second, or DPPS, instead - by distinct I mean that the picture content changes; that it's not the same picture displayed in a different way. Ie. Interlace has 25 DPPS because Denise outputs the same picture twice in succession. (Of course you can change the picture 50 times per second, but that's the problem defined, isn't it? On the "deinterlace-to-25-DPPS-TVs", the changes get combined and you either get choppy banding or choppy blur. On the proper ones you get 50 pictures out.)
Nope, i disagree, in interlace mode, Denise (based on Agnus timing) will flip between long/short line - this introduce "half line) and thus provide valid interlace signal (it is exactly 25Hz), for Lace = OFF (0) there is no switching between LOF and vide is 312 or 313 lines long so slightly slower or slightly faster than 50Hz. And this mode can create serious problems for TV's as they usually trying interpet this mode as Interlace where in fact this is progressive, half number of lines long, twice vertical rate video.
There is no Top/Bottom field as both fields are or Top or Bottom but TV try to interpret one of those fields as opposite to previous thus problem.
pandy71 is offline  
Old 07 October 2013, 17:56   #57
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
Yes, whether it's a problem interpreting progressive as progressive or a chip designed to deliver only deinterlaced doesn't matter though. The screens that don't deinterlace progressive and the screens that are designed to deliver progressive 50 DPPS are the ones you want (although some may want deinterlace to 25Hz for the interlaced modes for Workbench use as well!) Maybe you can't have the cake and deinterlace it too?

The still shots with very short aperture time were an attempt to capture the deinterlacing. In such a case, a deinterlaced picture lasting 1/25s taken with a much shorter aperture time would STILL display all 4 quadrants of the checkerboard and two digits at the same time. See pictures.
Photon is offline  
Old 09 October 2013, 09:45   #58
Solo761
Registered User
 
Join Date: Aug 2013
Location: Croatia
Age: 40
Posts: 341
Ah, it would be nice to have deinterlace issues . In my case I can't even get a picture, this is what it looks like with [ Show youtube player ] and this is with [ Show youtube player ].

Since I recently got them I thought maybe it's because of capacitors were bad so I recapped both of them, no change. I tried few RGB - SCART designs also no change. On old LG CRT TV both work fine so I guess this TV's fault, it's LG 42LW579S. It has one SCART that should support RGB.
Solo761 is offline  
Old 09 October 2013, 20:04   #59
Photon
Moderator

Photon's Avatar
 
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,087
It looks like the Switch pin has alternating current on it, you could try disconnecting it if you can select A/V Scart manually. But please open a support thread for that.
Photon is offline  
Old 10 October 2013, 10:53   #60
Solo761
Registered User
 
Join Date: Aug 2013
Location: Croatia
Age: 40
Posts: 341
I think it's TV issue. I tried making few cables, ian's cable, ian's cable with capacitors, the one from your topic that doesn't have SCART pin 8 connected and it looked the same with all of them. But on CRT it works fine, tried yesterday on one older Samsung CRT and it works fine.

Don't worry, I don't want to hijack this thread . I tried a lot of things in past month and none worked with SCART on that LG, point was just to avoid this LG 42LW579S if you want to connect Amigas to it's SCART, and no, not all modern TV's can display AMIGA PAL RGB correctly . (TV is about two years old).
Solo761 is offline  
 


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
LCD TVs that work on Amiga amigamaniac request.Other 25 17 November 2013 09:05
Using an LCD/TFT display with an OCS/ECS/AGA Amiga Photon support.Hardware 1 15 August 2009 00:28
Hacking an LCD tv for Amiga display Photon support.Hardware 12 21 September 2008 23:11
Amiga (any) and LCD Display wlcina support.Hardware 18 21 February 2005 16:47
Amiga 500/1200 and LCD display wlcina support.Hardware 18 24 September 2004 22:37

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 21:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, vBulletin Solutions Inc.
Page generated in 0.19191 seconds with 16 queries