Dish/DirecTV/Fios whole house distribution

What I do is not exactly the same, but I have all of my satellite receivers in my basement, including the TiVo, and my MythTV box. I modulate the output of the TiVo and MythTV boxes onto cable channels, and stick them on the in-house wiring. Put coax IR devices at each TV, a repeater in the basement, and bingo - I can watch the TiVo and the MythTV box at any TV in the house.

I don't know how that relates to FIOS, tho, and if FIOS can co-exist on one wire with modulated cable channels. But my setup works well for me.

It sounds like you have what I'm looking for. Could you elaborate a bit for those of us that are .... well .... noobs?

What does "modulate the output" mean, particularly what equipment/boxes/software do you do this with?

What exactly is your "in-house wiring"?

What IR repeater do you use?

What is the capacity of your setup with regards to inputs/outputs?

Is a matrix switch involved?

I'm assuming this is a HD distribution setup. What about audio? What about surround sound?

Sure. I threw together this diagram to try and explain what I mean. It's a little confusing, so stick with me.



With the exception of the LAN (at the bottom on the image), black lines indicate RG6 connections. Green are analog video (ie, composite, S-Video, or in one case YPrPb), blue is RS232, and red are IR emitters. Blue boxes indicate computers, yellow boxes commercial devices, and red boxes are a collection of devices that operate as one unit (ie, a simplification to keep the display from getting even more cluttered).

In a nutshell, the DirecTiVo operates as an independent DVR, viewable at any location in the entire house (obviously, two TVs simultaneously viewing would have to watch the same thing, and remote inputs would "compete").

The MythTV setup has two "frontends" - a standard def one located in the basement that can be shared the same way the TiVo can, and an HDTV frontend dedicated to the HDTV location in the house. The MythTV backend records from satellite via the RCA D10 and a capture card, and records OTA HDTV via the HDHomeRun, and uses the storage servers for hard drive space.


I have a 2-channel analog modulator from Channel Vision. With it, I take the composite video from the DirecTiVo and the MythTV Backend and stick them on channels 65 and 68 respectively, essentially creating a private closed-circuit television network in my house. Those are the only two channels on the network.


My condo has exactly two types of "media" wires in the house: RG6 and Cat5e (originally for phone), and thankfully they are all "home run" to a spot in my basement laundry room. Most of my gear sits in the basement, but for one device that needed hard-wired network (the MythTV HDTV Frontend), I repurposed one of the phone wires.


Xantech stuff. It injects the IR on to the RG6, and then I yank it back out in the wiring closet to relay via IR emitters to the MythTV backend and DirecTiVo.


I'm not sure I understand the question. Primarily, the TV system is distributing stereo SDTV signals over RG6. I am operating "at capacity" at both ends - as in, I've got no spare hardware to inject more sources, or more sinks - but I could easily get another RF modulator to add more channels to my "CCTV" network, and adding a new TV is as easy as adding an IR receiver and injector at the TV location.


It is not HD. I only have one HD location in the house, and it was added two years after this network was originally set up. I use ethernet to stream recorded HD data to a computer hooked to the TV that does the decoding and displaying work. So I've got fully digital audio capability, including surround sound, at my HD location.


This is pretty close. Hopefully, what I posted above clears this up.

Personally, I see my way forward into more widespread HDTV that the DirecTiVo is going to become obsolete, along with the entire RG6 network (including IR repeaters). I'll just plop down more HDTV frontend computers hooked to the LAN, and route my media that way. With some of the new hardware for recording HDTV from sat boxes in MythTV, I see little reason to not continue moving in that direction.

Two reasons - it didn't require PCI slots in a computer (I was upgrading my backend because it didn't have the capacity to record enough stuff simultaneously, and the computer I was retrofitting didn't have enough slots), and it was being designed to work with MythTV from the get-go.

I have found that the signal strengths seem to be a bit higher than my old AVerTV HD A180 cards. A pleasant unintended consequence.

As an aside, I think the model number of my Channel Vision modulator in the diagram is incorrect, but I can't be bothered to trek downstairs to check.

Obviously, the big problem with the RF modulation is the fact that it's standard definition video. At least, until commercial HDTV modulators come out (I'm not up to speed on that).

Foosinho thank you for an amazing diagram and explanation of your setup. You clearly put a lot of effort in that response ... and, of course, your setup.

As always, answers lead to more questions.

I am looking for HD distribution, and I'm interested in your comment about using a computer at each HD screen and using your LAN. I haven't seen that approach mentioned before, not that I've read everything there is. Far from it.

It is very attractive to think about simply using a LAN to distribute everything.

Is a gigabit LAN capable of doing all this? What is the wiring requirement? cat6? Just one?

Have you figured out what other boxes will be required? What software?

NP.


Of course! I'm glad to share my experiences with others. I've certainly learned a ton from this place.


It is very attractive. Of course, how I ended up here was a bit roundabout, and my current setup is rather... Dr. Frankenstein. But I certainly know what my goals are for when I build my future home, and the plan is to do virtually 100% IP-based entertainment (unless something really slick comes along - which is why I'll be using conduit!).

The really nice thing about this is that the RG6 network goes away - including all of the IR repeating hardware. Yay!


Gigabit LAN is easily capable of pushing all of this data around. I'm not anywhere near capacity - I'm having bigger problems with fast enough hard drives, TBH. Cat5e is capable of carrying gigabit. Technically, my system is "mixed" because not all of my adapters in the PCs are gigabit capable, so the switch I have will downgrade appropriately. I'm probably about 50/50.

BTW, high-end HD MPEG2 would be in the 27 Mbit/s range. GigE can support 1000 Mbit/s. As a point of reference, the HDHomeRun - that network-attached device with a pair of on-board HD tuners - has a single 100BaseT ethernet port. At full operation, with high-end signals coming out, they are only using about half of the capacity of a standard LAN. For most people, a standard LAN would probably be adequate, as long as it was switched. But if you are going to have several HD frontends running simultaneously, you probably want to go GigE.


So, if your situation is like the OP - too many signals for the wire he has in place - you can do what I've done and relocate all of the vendor-supplied boxes to one location in the house. So my antenna, DirecTiVo, and D10 are all sitting in my basement. And instead of putting a box for each different service at each TV, we put one box at each TV - but it's a box we make ourselves.

Here is my generic recommended setup.

At the wiring closet - where all of the cable/sat boxes are - you place a computer to act as the PVR. It doesn't have to be powerful, as it's mostly going to be shoveling data around (HD comes in already digitized, either from an OTA antenna, via a Firewire interface with a box, or via a device called a HD-PVR, and SD capture cards can be purchased that have on-board MPEG encoders). It will need an appropriate amount of storage for the TV you watch - raw HD from an OTA antenna will run from 6 to 11 GB/hr, SD recordings about 2GB/hr, and HD from a sat box or similar I'm unsure, but probably somewhere in the middle. The storage can be onboard, or in a network-attached device. If you make this PC a little faster than required, it can also act as a commercial flagging machine - this is a specific function of the software package that marks where commercials start/stop, so they can be automatically skipped upon playback. It's video processing, tho, so it takes a little juice. You'll also need the necessary hardware to interface with the video sources. I have a PVR500 to record from two SD streams. I'll probably always keep this around, if just so I can import VHS into my PVR if I want. If a HDHomeRun can serve your needs (I use one for OTA HD, but it can record from other sources), I *highly* recommend it. It's software is tested in the PVR software I use. It works, and it works really well. Finally, for providers who don't want to play ball, there is the HDPVR. It is a box that sits between the satellite dish (or whatever other HD device that doesn't have Firewire out) and the backend and converts component video and optical audio into a compressed digital form and spits it out via USB. It's also very new, and still being integrated into the PVR software. I don't yet have one. The codec is very aggressive in it's compression ratio - good for disk usage, but it requires more processor for playback. I will be waiting until the software is a little more mature.

Then, at each TV you put a small, powerful PC that will read the data streamed from the backend for whatever show you request (including live TV), and decode it for HD output to a TV. This machine is going to require a bit more "oomph" than the backend; decoding HD requires some serious clock cycles. If you are really clever (I'm not), you'll figure out how to get this machine to boot over the network so it can be run without a hard drive.

The real magic here is the software: MythTV. It's free, open-source software PVR project. The number of features and capabilities is absolutely staggering, especially compared to commercial PVRs. As I mentioned before, it has the ability to flag commercials for automatic skipping with remarkable success in most cases (and when it gets it wrong, it's a button press on the remote to skip back to the previous mark and continue watching). It supports some clever disk management schemes - meaning it is no longer necessary to use RAID or LVM to put all of your recording space into one giant volume. Shows from different sources - ie, recorded on DirecTV or via the OTA antenna - show up side-by-side in the recordings list, with no differentiation. In fact, if a channel repeats on different sources, the scheduler is smart enough to know the channels are the same if they have the same callsign, and schedule accordingly. For example, if I have a recording on ABC, it will try to get it off of the OTA antenna because that's my HD source, and I've marked that source as a higher priority to record from. However, if there is a conflict, and a higher priority show will be recording on the OTA antenna at that time, the scheduler will say "that's OK - I can still get this show in standard definition from the DirecTV D10" and do the "right thing".

The drawback? It's a lot more complicated than COTS PVRs, and sometimes the advanced features come with advanced bugs. Generally, people who use this software get a working configuration dialed in, and then don't update the software until a new "must-have" feature shows up. For example, I actually am running an SVN checkout (ie, pre-release development code) because it had a bleeding edge storage feature that I needed for my hardware hodge-podge to work. I've got it working (I do, on rare occasion, get segfaults in the front end - but again, I'm running a development copy of the software), and have left it alone for nearly 2 years IIRC. The current release version is more advanced than what I'm running, but I don't want to mess with it because it will take some tweaking to get it all humming smoothly post-upgrade.

I can say, from experience, having installed my own distro, downloaded the appropriate packages, etc etc the first time, the second time I went with one of the many customized Linux distributions targeted to support MythTV (I went with MythDora) with smashing success. I highly recommend using a MythTV distribution. It's a lot easier.

One other drawback - power consumption. All these hard drives I have eat up a lot of electrons. In fact, the system I have now I've already slimmed down when it comes to the number of computers in use. I don't think I'll try to consolidate functionality any further - but I will be a bit smarter about building stuff in the future that can ramp down it's power consumption whenever possible. (My biggest mistake is that my primary server has a collection of 8 hard drives in a software RAID setup. None of the discs can spin down when not being used. And it's too slow for more than one or two video streams at a time. It's nice to have the redundancies for home directories, but I don't need it for video. I'd rather have less power consumption. But maybe that's just me.)

NB: an important caveat to note is that PVRs don't work without good guide data. Without it, they are reduced to being expensive digital VCRs. No, you need data about when shows are on so the scheduler can do the dirty work for you. Unfortunately, in the world we live in, that data is not free. It just isn't. There is an organization, set up by MythTV developers, which purchases a license to the aggregated data (cost is confidential by order of the provider, but it is "not cheap") and then sells a year-long data subscription to individual consumers to cover the cost of the licensing. This is a not-for-profit undertaking, and the subscription is just $20/yr, which is very reasonable IMO. This fact rubs some people the wrong way, so much so that they refuse to pay for the data. I understand this point of view - I would think that channels would want people to know when their shows are on, and would be willing to provide lineup information in a timely manner and a machine-readable format, but apparently they aren't. So it is what it is. You should know this going in. The software is free. The necessary data to make it useful is not.

If I were starting from scratch today, this is what I, personally, would do:

Because I have my own web and email servers, and the ability to SSH into my LAN, I need a primary server. I've shuttled stuff like the MythTV database onto this box, since I've already got MySQL running there. Also, if you want to make MythTV's web interface available outside your LAN, you'll want to do what I've done - set up an SSL web server that uses Apache to password authenticate and then proxy that web connection to your MythTV backend. Don't directly expose the MythTV backend to the internet. It's easy to set up Apache password authentication directly on the backend, but my way ensures that the password is encrypted via SSL. You could probably hack the Apache config on the backend to get that if you don't have a web server setup on your primary server already, but I'm using what I've got.

Next, I'd set up a Network Attached Storage device. Doing it this way provides a few benefits - and a couple of drawbacks. First, the storage can be shared with your primary server, or any other devices you need. Once disk, multiple machines. Second, using something like FreeNAS will allow you, with the right hardware, to get a system you can maintain (replace dead drives, etc) without taking it offline. Third, you can easily create a giant array of storage nodes that can easily be expanded and upgraded. My plan is to, when more storage is needed, stick in a new terabyte drive and just share it with the MythTV backend by mounting it as a new folder to be recorded to. The primary drawback is that NFS is a bit slower than a local disc. But MythTV is pretty smart about load balancing when picking which folder to record to, so you don't have stutters in recording or playback due to a disk that can't keep up.

Finally, I'd come up with a small, quiet, compact frontend computer, and stick one at each TV location. Rig 'em to boot over the network - pulling their configuration from the MythTV backend or primary server, perhaps - and you've got quiet little diskless frontends that are only drawing power when you've turned 'em on to watch TV.

Good god, that's a novel!

WOW!

and THANK YOU

Truth is: I undersand only a fraction of what you said.

This is definitely a post to print out and slowly digest. I'm entirely a noob when it comes to Linux, and it is clear that there is a LOT to learn to be able to do this. I need a MythTV for Dummies book .... Oh Wait! ... you just wrote one. LOL

Where I live there is no OTA, so my sources will only be DirecTV (HD) x 2, Blu-ray, and music. The most I would have is 3 locations for viewing, and most likely 2 of the 3 would be viewing different DirecTV programs.

I have surround sound at my main HD screen, so I would need 5.1, possibly 7.1 audio capability there.

Me too.

Seriously, there's enough information out there to drown in. There's a Linux forum in the HTPC area here that mostly deals with MythTV, and there is a vigorously active MythTV community online via a mailing list (with searchable online archives).

This is probably a little bit of thread drift from what the OP was looking for. Realistically, tho, there's probably no way to put DirecTV/Dish on the same wire without doing something like this.


MythTV has a built-in music player too. It isn't great - hopefully a better one is in the works - but I've got my entire "My Music" directory from my Windows login shared so that MythTV can see it. I know development is underway with getting Blu-ray drives seemlessly integrated into MythTV, so you can rip movies to your hard drive to watch later, or just play them directly (which you can already do with normal DVDs). And the support for the HD-PVR devices is still very beta (maybe we'll see it stable an in the release by next spring?), and they are expensive items, but that capability is coming.

The way forward for you, if this is something you want to explore, is probably to get a MythDora (or MythBuntu, or whatever) LiveCD, find a Hauppauge PVR-150 card, and play around with the software (recording in standard definition) to see if you like it.


Not a problem. The HD-PVR apparently supports rolling the digital audio signal right into the compressed video stream, and then it's just a matter of having a motherboard that supports digital audio out on your frontend machines. Mine has coax digital audio out, so I just pipe that right into my receiver. If you don't have surround sound at a particular location, you can set up the system to output in stereo, via the normal speaker jack.

Seeing as how my next home will be built to my specifications, I'll be running conduit to TV locations for flexibility, but am planning on using a wired, switched, gigabit (or faster) LAN. Should circumstances dictate a change, I'll be ready.


Wireless is definitely a no-go. You *might* be able to get a single client up-and-running, but HD would probably be stretching things a bit.


Which is why I recommend one of the Myth distributions. They mostly just work. And once you get Myth configured, you pretty much never touch them.


The feature list looks very similar to MythTV. It is unclear to me how active the community is; MythTV's is very large, and very active. Two issues to keep in mind: first, every machine will need a copy of Windows installed on it. Second, a PVR is only as good as it's guide data. Where is the guide data coming from for Media Portal? If you just want it as a "pass thru" for watching shows from your different sources, it could work fine - set up your channel lineup, then just watch LiveTV, and change channels as desired. But you're already 99% of the way to having a pretty awesome PVR then. Something to keep in mind.

I think MythTV is an awesome product, and I love it. But it isn't for everybody. If people just need a normal PVR instead of a SuperPVR, I suggest TiVo. But, for power users like us - guys (and gals) who want whole-house PVRs with all the bells and whistles, it's hard to find anything better than MythTV.