Why not COAX Cable Vs HDMI/Others at home

I'm not sure what you are asking. Are you saying your house has HDMI, component etc wires run throughout? Most new houses these days will have those run, in ADDITION to coax cable.

You only need a 15$ or less HDMI cable if less then 15'
HDMI broadcasts digital video and audio instead of RF frequency content which is what COAX tranmits. You can convert HDMI to and from Coax.

Different cables for different needs/requirements. Just like a telephone uses RJ11, ethernet uses RJ45, composite uses one coax for video, S-video a din connector and separates luminance and chrominance and so on...

Look at your PC. How many different types of cables are you using?
A "simple" coax will/can not work for all requirements.

If you are paying more than monoprice or Blue Jeans Cable charges for an HDMI cable you are paying too much.

There is no reason HDMI could not have been implemented using coax. And if HDMI would have been implemented with coax, it would have made it a lot easier to run it long distances. But the resulting cable would have been much larger. Figure three coaxial cables for the three color channels, another for the clock, and one or more for the other functions. That would be some bundle, even using miniature coax. I doubt it would be cheaper than reasonably price HDMI cables currently. You couldn't get by with a single coaxial cable even if you redesigned HDMI to use a single link. There isn't enough bandwidth in a gigahertz ethernet connection to handle the full bandwidth of HDMI.

When HDTV is carried on coax, it has been encoded by the program provider and modulated into ATSC or QAM by the broadcaster or cable co or a different method by the satellite companies. STBs and BR players provide an uncompressed HDTV signal that is carried by component or HDMI cables. If you had the encoding and modulating equipment yourself, you could send those signals along the coax also. The simple RF modulators that were in VCRs were the analog eqivalent of that equipment. Unfortunately, they are not easily available yet in the digital era.

Coax is a type of cable construction: center conductor with cylindrical shield and a dielectric in between. You are probably thinking exclusively of something like RG-6 when you write "coax". Some of the other cable types you mention also use "coax". Most audio interconnects have coax construction. Composite and component video cables use 75 ohm coax, but with RCA or BNC connections instead of the screw-on F connector. S-video typically uses a twin-axial (2 center conductors) or dual coaxial cable.


Single conductor coax can only perform 2-way, full-duplex transmission with some fancy electronics (with echo cancellation on the receiver to remove the signal just transmitted) on both ends. Otherwise coax is used as a one-direction transmission line. For 2-way transmissions, multiconductor cable is preferred and used: one set of conductors are assigned for one direction, another set of conductors are assigned for the opposite direction. Twisted pair, either shielded or unshielded, is often used instead of coax for multiconductor cables but with less cost and bulk (e.g. USB cables and the phone lines).


Actually the opposite is true: good RG-6 coax has a much higher bandwidth (DC to maybe 3 GHz) than all of the other cables you mention except fiber optic cable. (But cheap Toslink does not have the same BW as the optical cable that telcos use.)

BTW coax such as RG-6 is best suited for very low level signals in the microvolt to millivolt range. Think of the weak TV signal that a TV antenna receives. A tabletop DVD player with a power supply connected to the AC power grid can generate a video signal a million times stronger.

BTW 2: Coax seems to be the cable of last choice. The coax is bulky and stiff; its connectors are large . For example, if you are old enough to remember, Ethernet used to use coax and BNC connectors. Today Ethernet uses UTP (unshielded twisted pair) cable and RJ-45 connectors.

Two words, Copy Protection.

Nonsense...

Copy protection isn't affect by the medium over which the data is transmitted.

True.


Those are connector types, not types of cable.

Puhleez....
Based on the opening question, sometimes one needs to dumb down the "science" aspect of the forum.

How long can this be dragged out? We can "correct" each other ad nauseum (as usual) to see who's smarter than a fifth grader.

You are correct. Have a cookie.

While Coax has enough raw capacity for multiple uses the electronics needed to multiplex the different services is not inexpensive.

Decisions over interconnect are typically trade-offs between cost, availability, complexity and ease of use. For example it is possible to use RG-6 coax for TV and Data using MOCA http://www.mocalliance.org/. Verizon is using it so they do not have to install new cabling to set-top-boxes when they install FIOS. On the other had if you already have Structure network cabling (Cat5/5a/6/6a) it is more capable and less expensive then using coax.

Another example S-video. It uses a cheap mini-DIN connector and carries baseband video and sync signals. In order to send those signals over coax requires a complex modulator and demodulator. Plus since coax is typically used in a multipoint network, rather then S-video point-to-point it needs to be able to carry multiple S-video signals and there needs to be some way to select the desired one. All in all much costlier then the S-video we all know and love.

I'm old enough to remember when ethernet used a transceiver clamped to the cable with a vampire tap. The cable was very thick and stiff and had marks on it where to place the taps so they'd be multiples of the wavelength apart.

To get back on topic...the answer to most "why" questions concerning the media industry can generally be answered by looking at who will profit from it. Coax could have been used (and is used in professional settings with HD-SDI), but it wouldn't have made much money. HDMI is licensed (by Hitachi, Panasonic, Philips, Silicon Image, Sony, Thomson, and Toshiba) at 4 cents per device plus a yearly $10000 per manufacturer, and HDCP is licensed (by Intel) for 0.5 cents per device plus a yearly $15000 per manufacturer.

Coax is analog and works best for radio frequencies. HDMI is digital and works best for short distances.

For digital it's a lot easier to stuck another wire in a cable to send more data than try to find some way to piggyback that data onto existing wires however high speed serial technologies like USB and SATA have made us old timers stop assuming that "serial" means "slow".

Nonsense. Early ethernet used a single coaxial cable to run 10Mbps 500 meters under 10base5 and 200 meters under 10base2. 10baseT used two twisted pairs to run only 100 meters.

HDMI would have better distance capabilities if it were implemented with coax.

By modulating an analog RF carrier to transmit the data.

10base-2/5 was totally digital, Manchester encoded with ECL single-ended drivers.
http://en.wikipedia.org/wiki/Manchester_code

From a purely technical point of view there no reason not to have used a single coax for digital HD distribution but the cost of re-encoding that signal to the display instead of using parallel high speed transceivers to send the uncompressed data might have been a design factor in the past.

It's not money. Different cables are used for different applications. It's like saying why are there different tools in a tool box. The different cables fill different requirements. And as noted, copy protection has recently come into play.

Yes, there could be a much greater degree of standardization, but once you have more than one cable type with more than one connector type, you might as well have specific types of cable and connectors for every different application.

Uh oh, this is a matter of interpretation and you can argue both ways. I'll argue one...

While it did send out digital pulses, it was really phase-modulating a fixed frequency carrier (or clock if you believe it's digital) transmitting at RF frequencies. When the signal on the coax was positive it didn't necessarily represent a one or a zero. To me that means I'm not looking at a pure digital signal but a carrier being modulated by a digital signal. Lots of people will strongly disagree with this interpretation. Despite sending out square waves (i.e. digital signals) on the transmitter end, at the receiver end it looked just like an analog RF signal on an oscilloscope. It was still close enough to a digital signal that you could use a simple ECL driver to demodulate it and that's how you can argue it had been digital the whole time.

ISDN used a similar trick to transmit "digital" data on a phone line and, man, people argued and argued about whether or not ISDN was a truly digital signal.

It doesn't help that signals that carry digital information are (always) referred to as "digital" signals by everybody. But if you substitute "quantized" for the word "digital" (which is legitimate), then these other people should realize that "quantized" levels (or states) cannot exist in an analog signal (or world) without transitions (however short or fast). These measureable transitions prove that these signals are not "digital".

But I agree, it is impossible to prove to some people that all signals are basically analog. It's the transmitted information that is analog (continuous) or digital (quantized).


BTW when a waveform has analog information and is sampled, that is called analog-to-digital conversion.
But when a waveform has digital information and is sampled, then that is called demodulation.

The bottom line is that it was sampled on the wire for only the 1 or 0 voltage levels. The harmonic distortion of the signal due to the transmission line was just an artifact of of the drivers RFI/EMI slew rate control due to current limiting to a few mA of drive on a heavily loaded coax.

What about SDI and HDSDI? These are uncompressed digital video signals over coax! But I do recoginize that any "digital" stream over coax is basically an RF signal.

HDMI is actually a QUASI PARALLEL interface. It uses four twisted pairs, R,G,B and CLOCK. It is the skew between the pairs at long lengths that causes problems. Pure serial interfaces over coax or fiber do not have this issue. HDSDI can be sent over 100 meters on RG6 coax. Try that with HDMI!