Hdmi/dvi/component

The most benefit people derive from using a digital video connnection is if they have a digital display device (LCD/plasma/DLP/etc). In that case using component cables involves the signal going through a D/A and then an A/D conversion to be displayed. Since you have an analog display device, that aspect itself won't help you. There are, however, a few benefits that you might see:

1) If the TV has a better video DAC than the Sat receiver then using a digital connection would be better.
2) If you're in a noisy environment and the component cables are picking up interference then using a digital connection would be better (unless you've got so much noise that the digital signal is undecipherable).
3) If the TV digitizing input signals for processing then you'd avoid an extra D/A+A/D conversion by using a digital connection. I don't know how common that is.
4) I don't know about your Sat receiver, but many DVD players will only upconvert video onto DVI/HDMI outputs, and not component. If you run into such a situation then the only way you'd be allowed to get 720p/1080i video would be to use a digital connection. If this doesn't apply to you then ignore this reason.
5) You can be cool and say you're using the newest/hottest technology.

You'll have to figure out which of those applies to you. Easiest way is probably to buy an inexpensive DVI/HDMI cable and give it a shot. If it looks better then it is better

Hope this helps,
-- Dave

 

Right one David....

In this crazy display world. You really need to try both and decide which is better to you for sure.


It is just not as clear cut as it seems it should be, no matter what type of display your using.

 

Digital is naturally going to give the sharpest resolution.

The conversion to analog for component introduces softening; how much depends on the cables and, more importantly, the quality of the electronics on BOTH the source and the display device.

Your best bet is to get an HDMI/DVI adaptor. They are pretty widely available for between $40 and $100.

HDMI is signal compatible (same signal) as digital DVI - all that differs in the pin specification of each. That's why it is so easy to adapt.

 

I've just heard about this recently and want to make sure I understand it. If I have an upconverting player and a HD display without a DVI/HDMI input, I'm stuck with a 480p picture, period?

 

yep. apparantly it's rare/impossible to now get a dvd player that outputs hd via anything other then dvi/hdmi. so, if your tv does *not* have those connections, you're out of luck.

ps - supposedly there is a hack for the zenith 318 that will reverse the latest change and allow you to output via the component.

edit - however, i believe just about all hd display gear has at least one of those connections don't they? heck, every hd tv we sell at bb has (at a minimum) a dvi connection. the more recent models have hdmi.

 

Okay, so just about any HD set is gonna' have at least a DVI; or at least I should not get one without a DVI connection (if I want to upconvert, that is).

Is HDMI better in any way than DVI, or are they fairly comparable?

Can you tell I'm in the market for a HD set... ?

 

Greetings

They are the same ... HDMI also allows audio signals ... where applicable.

HDMI partly came about because of the concerns from the industry about running DVI cables because their connectors are so bloody big ... and therefore more difficult for wiring a home ...

Regards

 

They're very comparable, and adapters are available. HDMI can be thought of as a superset of DVI and with a smaller connector. Here's a rundown of the differences I'm aware of:
1) DVI can carry an analog signal equivalent to VGA (DVI-A and DVI-I). This is relatively unused in consumer electronics.
2) HDMI can carry digital audio signals
3) I've heard rumblings that HDMI can go longer distances easier. I don't know if this is because of stronger transmitters, more sensitive receivers, or better connector, or even if it's really true.
4) HDMI can carry video data in the YCbCr colorspace (as well as the RGB that DVI uses). Since DVD's (and most everything else) are encoded in YCbCr (it's more efficient for compression algorithms), this would allow the display to do the YCbCr->RGB conversion instead of the source. This *may* allow a better conversion as the display knows it's color gamut better and can optimize the conversion.
5) HDMI can carry video data at 10 bits per pixel per plane (as well as the 8 bits that DVI uses). I've heard that if using YCbCr and crominance subsampling then 12bits per pixel per plane can be used.

I don't think that (4) or (5) are actually used yet, but I wouldn't be surprised if an HD-DVD or Blu-Ray player (optionally) used them.

-- Dave

 

Thanks, fellas, one and all. (even though, Dave, you started losing me at 4), and pretty much totally lost me by 5)...! )

Is this implying that a DVI input is not neccesarily a HD input, or a component input?

Edit: oops, DVI stands for Digital Video Input, doesn't it? Sooooo, I'm guessing there are some DVIs that aren't component, right?

2nd Edit: No, Jim, DVI stands for Digital Visual Interface, and HDMI stands for High Definition Multi-Media Interface.

 

dvi and component are two different types of connections. (check out the beginner's primer..i think there's a thread about the different connection types). i think what david was trying to tell you what that, in the past, some dvd players that could upconvert the resolution via component no longer do that.

 

Yup, Ted, that's what I meant



Well, then I can explain what I was talking about, even though it may very well never effect you in real life
Feel free to ignore it if you don't find it important.

Quick primer on how to represent colors:

In order to represent a color you need to have defined a colorspace, just as to represent a number you need to having a number system. For instance nine in arabic numerals is "9", while in roman numerals it's "IX". They represent the same thing, but looked totally different. So it is with colorspaces.

One very common colorspace is RGB, and it's variations. RGB stands for Red Green Blue, and a color is defined in RGB by how much red green and blue light is in it. It's a recipe sort of like baking a cake. For instance a light aqua type color may have 0% red, 50% blue, and 80% green (each is on a 0-100% scale). White is 100% red, 100% blue, 100% green, and black is 0% for all.

A colorspace used by almost all color printers is CMY, which stands for Cyan Magenta Yellow. Just like mixing paints, if you mix those 3 colors together you can make any other color. Almost all printers also add a black ink, because it's so common and hard to do well with just CMY inks, so you often hear this referred to as the CMYK colorspace.
CMY white would be (0,0,0), CMY black would be (100,100,100). IOW as you add ink you take away reflectivity, and so reduce the brightness.

The colorspace YCbCr is different from both of those in some very important ways. It also has 3 parts to it, Y, Cb, and Cr, but they are not mixed in the sense of RGB or CMY. the Y channel is the luminance, or brightness channel. If you turn a color image into a black and white image you get luminance. The Cb and Cr channels are two color components which don't match up to anything in the natural world; they are purely mathematical constructs.
This was actually how color TV came into being. They already have the luminance channel of the black and white broadcasts. They added Cb and Cr channels, which the TV then converted into RGB to display. Existing B/W TV's simple ignored the Cb and Cr, and looked at the Y channel which was the same thing they always did. That way you get color and B/W broadcasts in the same signal, and everyone is happy.

YCbCr has another property which makes it very useful. Because of the construction each of the channels varies more slowly than it would in RGB or CMY. This is obvious with the Y channel, as you can see most images the brightness doesn't change very quickly. You might go from a bright blue to a bright yellow, and the luminance would change only slightly, while the amount of red, green, and blue would each change very much. This is very helpful when compressing an image, because high frequency information (alot of change in a short space) takes more data to represent. High frequency information is often the first thing to be thrown out when using lossy compression such as JPEG, or MPEG2 (or MP3 for audio). This is why you might notice some "mosquito noise" around sharp lines in a digital photograph if you zoom in or use a low resolution. This is a big reason why "edge enhancement" on DVD's is bad, because you're adding sharp edges to the image, which the MPEG2 compression does a bad job on and creates mosquito noise around the edge.
Another benefit of YCbCr is that the human eye is much more sensitive to brightness information than color information. This means that you can throw out some of the Cb and Cr channels (called chrominance subsampling) with very little impact on the perceived image quality. Commonly there is half as much of each Cb and Cr as Y (often written as 4:2:2 subsampling), or the is 1/4 as much (often written as 4:2:0 subsampling). If you use 4:2:2 subsampling then you end up with 2/3 as much data as you started with, and if you use 4:2:0 subsampling you end up with 1/2 as much data as you started with. This is very often used as a first step in compressing an image.

So long story short YCbCr is a much better colorspace for compressing images than the RGB that displays actually use. So someone has to convert the YCbCr data into RGB. If you're using a component connect the display would do it, because component connection uses the YCbCr colorspace (actually it's called YPbPr in the analog form, but that's details). DVI only supports passing RGB data, so with DVI the source would do it. Since it's best to do the colorspace conversion before the digital to analog conversion, that's not a bad thing. HDMI can pass YCbCr as well, which would allow the display to do the conversion, and to do it in the digital domain. Good all around, I think.

As far as bit depth, that's luckily a simpler topic
Let's take as an example the RGB colorspace. If you want to represent a color in RGB digitally, you need to assign numbers to each of the channels. IE how much red, how much green, and how much blue to use. If you use 8 bits for each number you have 256 possible values on your scale. If you use 10 bits for each you have 1024 possible values on your scale. If you have 12 bits for each you have 4096 possible values on your scale. So more bits = (alot) higher precision. You may hear of something like "24 bit RGB", which would normally refer to 8 bits for each R G and B, not 24 bits for each. You might also hear "8 bit RGB" to mean the same thing. Usually the size of the number makes it clear whether they are talking about bits per pixel (aka 24 bits total), or bits per channel per pixel (aka 8 bits for each R G and B).
The same concept applies to all other colorspaces. As an aside, you need at least 3 channels to represent the gamut (possible colors) of the human eye, so most colorspaces have 3 channels.

This is probably alot more than you ever wanted to know, but there you have it anyways

-- Dave

PS. There are a great number of other colorspaces out there in many different forms, but variations on RGB, CMY(K) and YCbCr are by far the most common for various reasons. I didn't want to get too far off track talking about them, but there is one or two interesting things yet out there...

 

Thanks David

If that is the quick primer on how to represent colors, I'd just hate to see the full version . Does anyone have an aspirin?

Artie

 

Heh, you caught me, it wasn't as short as I originally intended. But isn't that how it always goes?

These things are obviously easier to grasp visually. If you search around the web I'm sure you'll find some fancy explanation with all kinds of graphics and applets. Then you can save the asprin for another day

-- Dave