what is possible resolution for component video?

[FredThompson]

Would someone please direct me to a reference for component video that would show the upper resolution it should be able to handle?

[jggimi]

For analog capture, or for another purpose?

For capture, I would expect it's the limits of the capture chipset. The cheap cards most of us use are fairly limited, typically 720x576 for PAL, 720x480 NTSC. And, because they're cheap, they usually are limited to S-Video input.

[mpucoder]

Or are you asking about CGA, EGA, VGA, SVGA ?
If not, which components? (typical video switchers like GVG-200 can handle analog RGB, YCbCr, YPbPr, YIQ, etc)

[Atamido]

A lot of high end CRT's used to use component inputs, (5xBNC connectors) so those could obviously handle high resolutions. However, for standard A/V you have just 3 connectors for RGB. The possible clarity/resolution is supposed to be higher than what is capable with S-Video, and I even saw a direct comparison once, but I cannot recall where. Though, I don't remember a big difference between the two.

Does anyone know if there is a generic name for the 5xBNC connectors used for computers CRT's, like the name "component" used for split RGB in A/V?

I have owned a 5xBNC 21" CRT, but it was old and so I can't comment on the quality of the signal. I currently own a 21" Trinitron CRT with both a 15DB connector, and a DVI connector. I am waiting on the delivery of the DVI cable in the mail as the super cheap 15DB VGA cable I am currently using is causing some ghosting of the image. I am confident that a DVI cable will give me a perfect signal.

[^^-+I4004+-^^]

fred,you almost asked a question...almost....

[Atamido]

That was funny enough for me to blow air out of my nose, but not quite chuckle. Funny stuff.

[FredThompson]

ok, ok, I'll try again.

wires/cables have an upper frequency limit they're rated to handle.

since we're talking about video, that frequency limit will correspond to a screen resolution, more or less.

I'm interested in the RGB standard. Didn't know there's a five-wire one but it makes sense.

Capture isn't what I had in mind. I was wondering how to tell what type of resolution could be expected from the more recent ATI cards with their RGB adapter.

27" HDTV tube TVs are around $600-$700 and up. The Act Labs video-out light gun needs S-Video connections and doesn't work with plasma, LCD, or other non-tube displays.

I'm wondering if there's a way to have a tube TV that could also function as a large monitor. It doesn't have to be perfect, just needs to handle 1024x768. At the same time, I'm trying to determine what the lowest common denominator might be. An HDTV display makes sense as a long-term investment, especially since DirecTV will be expanding that, but I'd prefer to not pay much more than is necessary for the display.

The computer system I'm using for games could also be used to edit streams extracted from my TiVo or camcorder stuff without going to the office. Would make far more sense for the "game" system to run compression than my usual workstation so the "real" work doesn't get bogged down. The quality needs to be high enough that I could really see what's going on with the video and do other work without the display being all mushed.

S-video won't provide enough resolution and the next step up is component which I know nothing about. Hence, the question.

[int 21h]

Composite is 200-220 lines of horizontal resolution
S-Video is 400-420 lines of horizontal resolution
Component is 500-525 lines of horizontal resolution

You can figure out the source's lines of resolution by doing a simple calculation. For example, DVD resolution of a 4:3 picture is 720 / 1.33 = 540 lines of horizontal resolution (usually unless you own a very expensive DVD player, you're given a little bit less than that, due to digital-to-analog conversions in the electronics).

[FredThompson]

500 lines isn't bad. Guess I could live with that. Wonder if OmegaDrive's custom drivers would give a proper resolution so there's no "mushing."

720x480 for NTSC DVD is raw, the pixels aren't square. Although, you did make me think of something worth investigating.

The TVs I saw listed at Best Buy's site said they are HDTV with 1080i support. If the connections are RCA, S-Video and composite, how can they display at that resolution if they can't take the full res as input? Do they have a flat connector like the LCD display?

[int 21h]

I would imagine the HDTV scales the output from the DVD Player.

[FredThompson]

That's what I was wondering about.

"Sure, this TV does 1080i" (on the screen, you just can't feed it that resolution...)

[Atamido]

BEWARE

You are really going to have to put research into each unit individually. Just because a display says it is HDTV doesn't meant it can display a very high res. A display may only be able to 720x480 progressive and call itself HDTV because it can display more than the standard 720x480 interlaced.

Also, just because a display says it can display 1080i, doesn't mean it can physically display that resolution. A lot of displays will say that they support a higher resolution because it scales the image down to fit. I have seen displays that are 800x600 and say they support 1080i because it can take the input and scale it down to 800x600.

Here is a quote from BestBuy's web page: "42" 16:9 plasma display panel with 480p native resolution (853 x 480 pixels) — accepts inputs in all major DTV formats (1080i, 720p, 480p and 480i), plus PC video"

What you need to look at is the "native resolution". This display accepts 1080i and 720p, but it just scales it down to 853x480. If you hooked it up to your PC, you would be getting far less than 800x600 displayed. Thats not to good for a display that you just dropped $3300US on.

My personal favorite is the $10,000 Pioneer 50" 16:9 Widescreen. Look at its stats:
Vertical Scanning Lines (Native Mode)
1080i
Horizontal Resolution
1280 lines
Pixels
983,040

But wait, 983040/1280 = 768. The display is 1280x768, but it says the native "Vertical Scanning Lines" is 1080i. So shouldn't it be either 1280x1080 or 1280x540? WTF?

On a lot of these you will need to take the model number to the manufacturers site to find the actual specifications and see what resolution it really supports. If you can afford it, find a display that has a DVI connector so that you can get a perfect signal to your display.

All is not well in HDTV world. There is a lot of not-completely-honest advertising out there, and people constantly fall victim to it.

[FredThompson]

Oh, that's a real crappy thing to do. It's like the big lie of 56K modems, isn't it? The one I was looking at claimed 800 horizontal lines. Boy, this really stinks. Why is it so hard to get anything over 17" at a decent price?

For those who don't know, a phone has a 4K frequency spread. Can't remember the exact formula but, essentially, 28.8 is about the upper physical limit that can be carried by that frequency. Windows does a great job of telling you the connection speed the modem drives claims it has, not the real throughput. Almost nobody gets anything even approaching 56K because of line noise, extra handshaking as it tries to move more data, etc. After I'd upgraded to 56K (victim of the hype, I admit) and learned how to test the real throughput, I found more data actually moved through the 28.8 connection.

[Atamido]

Originally posted by FredThompson
The one I was looking at claimed 800 horizontal lines. Boy, this really stinks. Why is it so hard to get anything over 17" at a decent price?
I purchased that 21" Trinitron Perfectflat CRT w/ DVI connector for about $340US after shipping. But it was used, so its a risk. Mine works fine, but your results will vary. Of course, that has a native resolution of 1600x1200@85hz and it is comparing apples and oranges. Honestly, I prefer this for use with my computer. For a TV though, the larger, lower resolution displays are fine.

[FredThompson]

Yeah, I guess I'll just have to "settle" for monitor-like behavior. Will be fine for games and simple video editing but not for cleaning or compositing. The idea of a 27" screen to play MAME and also edit video is very intoxicating to me ;)

[^^-+I4004+-^^]

>Composite is 200-220 lines of horizontal resolution
S-Video is 400-420 lines of horizontal resolution
Component is 500-525 lines of horizontal resolution

not true.......composite is such only on VHS recordings,but the composite signal itself has all the bandwidth as s-video (cca. 5Mhz) (ie. svhs won't get you image that's better (resolution wise) than the broadcast video with composite connector used....but pure s-video will have less luma/chroma crosstalk with proper source .....)

also,guys you're forgetting one thing:
classic tube won't last as long as a monitor tube!
you'll have "phosphor burn" in short time....(some patterns stay inprinted all the time...)
monitors use more resilient phosphor that can take constant image (always the same patterns on the screen as in PC) for longer durations of time.....
so classic tv as a monitor is not so good idea.........
(dunno if projection tv's are affected too...)
user manual for my panasonic tv clearly states this...(although it's not quite new tv-set or so)

also,don't worry about the cables,they won't limit you...input/output devices will.........

>After I'd upgraded to 56K (victim of the hype, I admit) and learned how to test the real throughput, I found more data actually moved through the 28.8 connection

not quite fred,not quite:56k is DOWNLOAD speed,and upload speed is the value you suggest (cca.28 or a bit more..i usually have 32k upload though)
why is this?because the phone company can provide infrastructure with fewer a/d (or d/a?heh) conversions and therefore they have nice link to you...but you don't have nice link towards them,as you're stuck with that extra d/a conversion (when your digital data is converted to 3k phone bandwidth) so no fast upload for us 56k folks....
i'm dloading at 5 and sometimes even 6kb/s and that's much more than 28,8k,isn't it?
(wether u use 8 or 9 bits per byte....(modems use 1 "stop" bit ))

/ivo

[FredThompson]

Use will primarily be video output, not monitor. Would be nice to work from the counch instead of the office all the time.

String and chaos theories, aside, physics basics haven't changed. There's only so much that can be pumped through a 4k spread and it's less than 56K, very close to 28.8. Throw in line noise and added handshaking for overhead and you get a lower actual bandwidth.

we have a 4k spread, not 3. Don't know about the rest of the world but it's got to be pretty close. That gives 16K when 4 lines are multiplexed. Don't remember the reason that's important. It's another physics issue related to wire. If you've only got 3, you're limited even more than dial-up here.

You're probably confusing apparent throughput with actual. Apparent would take into account data compression which si part of the 56K standards.

[Ookami]

@Fred

I would say, worry about the cables, as in my experience the difference between using different cables is very visible. But, I've never used the component cables.

Of course just because it's the the cable/implementation is the same standard doesn't mean it's the same quality like other cable/connector etc..

E.g. there have been even cases reported where the s-video output (or input) was giving a lower quality than the composite one :scared: etc. I think that Pamel wrote a similar thing *cough* . So, I guess, it's best you test it for yourself.

Yummy, big screen :) .

Cheers,

Mijo.

[ppera2]

I see here lot of shallow explanations, and starting question is wrongly made.

5 BNC connector on 'high end' CRT monitors is simple for R,G,B, +H +V sync signals. No DDC :angry:

What determines possible resolution is not connection type in fact, but video standard. Or bandwith (max freq.).

I don't like termin component video, because it's not enough specific. But under it is usually considered YUV connection, what we can see on DVD's, projectors and expensive TV's.
It's nothing better than RGB, actually is worse little.

But resolution depends from max. freq. of video, so it is by actual standard.

[FredThompson]

I guess if we want to get completely accurate we should say the possible resolution is infinite because cables are just cables and they will have an upper frequency capacity. How that data is interpreted has nothing to do with the cable, per se. If the refresh was 1/sec...

You're right, I'm still not sure what the "proper" question is but, more and more I realize it has to do with whatever standards exist for the delivery of video over certain types of wires.

I'll look into that. Will also ask ATI what comes out of their adapter. What if it can send 1280x1024 and a sub-$1000 27" digital TV can properly display that?

That would be quite nice.

[^^-+I4004+-^^]

Originally posted by FredThompson
String and chaos theories, aside, physics basics haven't changed. There's only so much that can be pumped through a 4k spread and it's less than 56K, very close to 28.8. Throw in line noise and added handshaking for overhead and you get a lower actual bandwidth.

we have a 4k spread, not 3. Don't know about the rest of the world but it's got to be pretty close. That gives 16K when 4 lines are multiplexed. Don't remember the reason that's important. It's another physics issue related to wire. If you've only got 3, you're limited even more than dial-up here.

You're probably confusing apparent throughput with actual. Apparent would take into account data compression which si part of the 56K standards.

i luv offtopics!hehe

fred,get a grip man!i'm not saying some numbers i have invented......
nor am i a man that doesn't know what dload speed he has!
in good conditions i'll dload 300kB per minute (5kB x 60sec) or 1,2MB dload in 4 (four) minutes.......i know folks that go even faster,and even i sometimes have 6kB/s speed.......
and no,i'm not believing windows..i can measure the time myself and see how big file dloaded is.......
when the connection slows down to 28,8k(this happens sometimes) i can feel it ok..it crawls....it's scary.....

and this is for zipped files (so just forget data compression of 56k standard!)



check out this link (and it's in australia quoting some USA numbers too,so...)
http://www.technical-services.com.au/BITS/modemlines.html
see especially the sectiton "POTS (Plain Old Telephone Service)"
(a quote:"Very high-speed connections assume that there is a substantial amount of digital technology in the system (thereby normally increasing the usable bandwidth). Some users are going to find that their local phone system effectively limits them to 16.8K or 19.2K.")
this web suggest that more than 0-3kHz is needed for higher speeds,you're right there,but the 3kHz figure is the usual one mentioned for phone lines....(it depends on the hardware in "POTS")
that's why i mentioned it

also there are tests that can prove you some things about your connection (see this http://homepage.tinet.ie/~leslie/testpage.htm
one) if you have poor connection then i believe you have 28.8....but i have 40-48kbit/s dload most of the time (mentioned 5kB/s dload speed)..i have never suggestted that 56k is exactly 56 as it's not....some folks get 52kbit/s...i have 40-48kbit/s most of the time
this proves that my ISP indeed has most of the stuff digital(in his domain) and my modem is no master of d/a conversion....(but it still goes up to 32kbit/s....that's chechked too in case you wonder..real number from the real world)

>That gives 16K when 4 lines are multiplexed. Don't remember the reason that's important. It's another physics issue related to wire.

?
physics issue related to the wire?
"twisted-pair" wire is capable of ADSL and that's some mighty bandwidth (over 1MHz as some paper on ADSL by my ISP suggests)
3kHz limit is made because of the hardware that produces the sound/cheap telephones etc..ie. it's "legacy" hardware so it acts like it acted 60 and more years ago....(ok,probably lil better..hehe)

perhaps you didn't benefited from 56k just because you have poor line quality,but i appreciate the difference between the upload at slower speeds (up to 32kbit/s) and dload at higher speeds (up to 48kbit/s in my case) so i wouldn't ditch 56k so easily.....

see this link http://www.56k.com/ too.......

[and i thought nobody has dial-up in this new age..hehe]

(probably better suited for PM,but anyhow........display question is not interesting anyhow....get that big screen tv and burn it with some PC video edititng...hehe......you won't burn it only if you display VIDEO ONLY on it....if you edit you still have PC graphics on..)




/ivo

[FredThompson]

Sheesh, you're right, this all came from a side comment and it's spun into it's own Comedy Channel skit.

I'm on a cable modem, actually, have been for about four years.

Yeah, 3K usable would make sense if 4K is allocated for analog use.

That's the kicker, analog. You're absolutely correct, the vast majority fo provider equipment is digital shortly after the customer connection.

Guess I'm showing my age, I was thinking about old copper. It's been 15 years since I did any work with basic electronics...

Sure there's limits on frequency based on wire construction. Only the surface of the wire carries any charge and there can also be crosstalk problems. I remember a few years ago, oh, okay, more like 8, I'm getting old, which makes me babble, is it cold in here?, having to specify CAT 5 when I picked up cables because the primary LAN cables at the consumer stores were some lower standard that worked at 10 but not 100...mine goes to 11 (old man humor, there....saw it on a videotape before you youngsters were born...harumph...where's my teeth?)

Good point about twisted pair. An isolated pair of wires is probably capable of a lot more, consistently, than our standard four-wire if both lines are active. A line in use generates magnetism that's picked up by the others. At least, I think that's what happens. If somebody is talking on one half of a shared-four here it will interfere with the other half but the single-pair isn't affected.

That's where the upper limit comes from. A wire can only carry so much frequency, consistently, based on its metalurgy (sp?) and interference.

[^^-+I4004+-^^]

>That's where the upper limit comes from. A wire can only carry so much frequency, consistently, based on its metalurgy (sp?) and interference.

uhmmm...did i hear you shout "skin-effect" right about here
>Only the surface of the wire carries any charge
?

this wire is thin,and on hi-freq. all hell breaks loose....
so it pushes the current to outer sides,lil space is left so less of anything goes thru....
but as is 1MHz is enough for 256 tones of DMT (discrette multi tone) modulation........
cable connections should be capable of much more as they use better (rf antenna type same cable is used for tv and internet services...) cables which practically have no limits...(if i got all the terminology right)
but obviously are limited by the providers bandwidth towards internet backbones....
imagine connection with a rf antenna cable that uses the complete spectrum of that cable...(this means all the tv and radio channels and then some)
that would be much more than todays PC's can accept or process...

to top off this "Comedy Channel skit. " (heh it was fun anyhow..)

cheers

/ivo

[mpucoder]

If you are talking about analog connections to an HDTV, they will be NTSC or PAL standard (legacy connectors). This includes a yellow RCA (composite), S-Video, and even 3 or 5 cable component. There is no analog HDTV, it is digital. HDTV receivers are required to display SDTV (Standard Definition TV) as well, at least for the transition, and possibly forever.

Edit: RG-8, RG-58, and RG-59, the cables commonly used for video applications, are good up to at least 900MHz.

[FredThompson]

Hmmm...

What about the DVI connector? That would be a digital output, yes?

If so, is that a standard connector on a HDTV TV?

If not, is there a way to do some kind of 1:1 adapter?