Separate names with a comma.
Discussion in 'DIRECTV Installation/MDU Discussion' started by k2tech, Sep 1, 2009.
Recording "server", single RG6 to various receiver "clients" and with DECA, "you're done".
I will go one step further and say that the next generation of receivers will have it built in too, so that you only need to plug one box somewhere in the home into the internet and all boxes will be automatically hooked up to the internet, without any additional devices...
I wasn't really trying to address whether or not the whole transponder bandwidth had to be available in order to extract a single video channel... if fact, just the opposite, I went out of my way to state very clearly that such a restriction could very likely be a limiting factor on how small they could make each SWM channel.
What I was saying was that the current tuner hardware would be able to handle a smaller SWM channel, provided that all the data that is needed to extract a single video channel was available on that smaller SWM channel.
Actually, your proposed solution is cramming more into the same cable, or more eloquently stated.... reducing the wasted and unused bandwidth in each of the current SWM channel allocations.... which is a good thing.
That said, I'd be willing to bet the farm that the SWM16 will be nothing but two of the current (less efficient) 8 channel SWMs with built in DECA to SWM blocking and built in DECA cloud to DECA cloud bridging, all integrated into an LNB assembly that has two coax outputs.
Don't bet the farm unless you're really wanting to part with it...
I'm not saying this will or won't be the approach, but sure seems the risk is pretty high...
The risk of having two separate SWM outputs is way lower than trying to completely re-engineer SWM, which has been tested and already has hundreds of thousands of units in place (LNBs) and likely millions of compatible receivers. They would have to go through an insane amount of testing if they reconfigured the channel bandwidths, etc. There would be much less testing required of an LNB that has an extra SWM switch inside.
I see now. I had assumed that the tuners used pre-defined 102mhz channels and that the SWM channel map simply provided which channel number (of the 8 possible) contained which transponder. You are suggesting that the SWM channel map actually contains the frequency ranges defining the channels. If that is the case, then what you are saying makes complete sense; all it takes is th ability to use less bandwidth per channel and a new SWM is all that is needed to get more than 8 channels. Thanks, VOS!
Does this still leave room for the DECA (or does that overlap the OTA band)?
This made me think about a possible solution in case the current tuners can't handle more channels. Maybe the system could use two SWM feeds, but then stack them onto the same single run of coax. Of course this would require a destacker at each receiver, similar to the BBC's. Since a single SWM "partition" still wouldn't be able to handle more than 8 transponders, it would require two sets of destackers - one for the low partition and one for the high partition, and no installation could have more than 8 of each.
Sorry, I'll be more clear. The risk I referred to was in his bet, not in the technology. A farm seems like a very large thing to bet on the little amount of information we have.
This is why I'm "leaning" towards the narrower 70 MHz bandwidth.
DECA is down in the OTA band and the non H/HR23 receivers can't receive anything below 950 MHz.
For this [SWM16] to be backwards compatible, it needs to work above 950 MHz.
The guide channel is at 974 MHz. If this and all other SWM channels are spaced 72 MHz [70 for the channel and 2 for separation], they'll all fall within the frequency range of the MPEG-4 receivers.
Thanks for the clarification;
But I still see no way for the SWM to actually reduce the data flow from a transponder beyond going to something like a full packet switch where the SWM only sends the packets pertaining to the specific channel on the transponder requested by the receiver instead of the current method of transferring the entire stat-mux stream and letting the receiver handle the demultiplexing.
Can the SWM do this without extensive redesign and increase in complexity?
[Or maybe "packet filtering" instead of "demultiplexing" would be a better choice of words since the receiver simply ignores all other packets in the transport stream(s) it is receiving outside of the desired ones at any given time.]
If the largest transponder is 65 MHz, then a narrower 70 MHz bandwidth, would work.
I agree VOS;
That seems to be the most sensible design approach for such an expanded multiswitch by simply reducing the individual channel bandwidths to around 70 Mhz.
But I was simply responding to cartrivision's speculation of having the SWM act somehow as a sort of channel preselect by reducing the bit rate from a transponder to just enough for a receiver to correctly extract the desired channel. And as I said I don't see how this is possible with statistically multiplexed TDM streams. If the individual channels on a transponder were laid out in frequency maybe (FDMA), but not in time this way.
The only option in this case would be for the SWM to become a full packet switch and only route the applicable packets for a requested channel (HBO, SHO, TVLand, etc.) on it's RF channels to the receiver particular tuner ala AT&T's U-verse approach.
This part is a "can of worms" that seems not worth opening [besides not being my specialty].
My bet (very small bet at that, no spare farms on hand to bet) is on a "double SWM8" in a single package. Two outputs each capable of supporting 8 tuners. Same frequencies, channel spacings, etc., as the current SWM8 has. As to a 16 port SWM LNB, don't know but if one is developed I would expect the same, two output ports each supporting 8 tuners.
Time will tell.
Regarding the other SWM-16 speculation though about having two SWM-8s in parallel, you mentioned the two outputs being combined in a "crossover bridge." Pardon my ignorance if the answer should be obvious but how exactly would such a bridge work to prevent co-channel interference on a single cable following the bridge since the two SWM-8s are using the same RF channel frequencies?
Now you're asking in my specialty.
A DECA pass band filter between "the two SWM8s" would have one DECA cloud, and if between "this filter" and each SWM were the [current] DECA bandstop filters, then each SWM wouldn't react to the other [no common path].
This is very simple filter design. Think of this like a diplexer that separates OTA from the SAT signals.
Perhaps we've made a case that the title should be changes to "rumored"...
But isn't this still two physical "brighed" cables coming from the double SWM-8 arrangement in the LNB to be distributed throughout the residence?