[TomCat]

...I couldn't figure out how it's possible to use one tuner to get 4 networks at the same time, but in other places it sounds like it isn't recording them in the traditional sense.

Conventional ad hoc digital sat delivery uses 4 slots (directly adjacent slices of the bandwidth each on a separate frequency) per transponder, typically fed ad hoc by 4 separate uplinks from separate sources, which allows 4 different clients to timeshare the transponder on demand at any given time, simultaneously. Since DBS uses all of the xponder all of the time, they instead send a single MPTS (multiple program transport stream) which is one wideband carrier with the modulation holding the elemental programs from 4 sources all in the same stream. DISH leverages this by arranging the spot beams for the big 4 on the same xponder in each market.

A conventional DVR/IRD demodulates the MPTS signal and then demuxes the program signal of interest, discarding the rest. That then goes to the HDD for playback. IOW, even a conventional DVB tuner can (has the bandwidth to) see the entire transponder at once, but conventionally, only a fraction of the MPTS (a single program) is recorded.

The Hopper does not demux directly after demod, but sends the entire MPTS to the HDD. Then, at playback, it does the demux.

So there is nothing new as far as the technology goes, but moving the demux to playback is what makes PTA possible (that, along with fast CPUs and large HDDs).

This, of course, implies that it is impractical to save one show from a recording that actually holds 4 simultaneous shows since the file size is 4 times larger, and why to save a show past the 8-day expiration they have to basically do a playback/rerecord in the background from that partition to the user partition. Pretty slick implementation of conventional pro technology for consumer use, if you ask me. DirecTV will be close behind.

So no, not in the traditional sense (as far as consumer technology goes; TV stations have used this for backhaul for years). But not exactly magical, either.

 

[TomCat]

Just to note:

This part of your post describing "Conventional ad hoc digital satellite delivery" actually sounds like "Frequency Division Multiple Access" (or FDMA).

Therefore shouldn't the last portion quoted above read as;

It's both. Each of the four slots can be leased out for time segments as short as 5 minutes to individual customers. That's time division multiplexing, per slot. The slots themselves share 4 narrower slices of the wider transponder, each slice dedicated to a single slot. From the point of view of the transponder, that's frequency division multiplexing.

Sorry for any confusion. The point is that since the DBS company owns or leases the entire transponder, they can remotely reconfigure it to have one wide carrier centered on a single frequency rather than 4 narrower carriers adjacent to each other, and that is received by a single wide-band tuner, whereupon the 4 separate signals are separated out (demultiplexed) at eventual playback, or separated at reception for a conventional DVR.

So wideband single carrier (containing multiple programs in a single stream) vs narrow multiple carriers (with a single program in each) is the difference between how DBS sat and conventional sat works (with DBS usually having more transmit power to illuminate the smaller reflector), but the only real difference between a "Hopper"-style IRD/DVR and a conventional IRD/DVR is that the Hopper demuxes at playback but records all of the programs at once, while the garden-variety DVR demuxes at record to minimize the amount of HDD space needed for a single program.

 

[TomCat]

Other than them moving your big 4 channels to the same transponder (and only if necessary) nothing changes on the transmit end, so no change is made to the bit rate and quality. The tuner is already wide enough to accept the combined bit rate of the entire stream, and your HDD can already handle a stream that big as well.

The difference is in where the signals are separated out, either at record or at playback. The demux stage just separates the signals, and does not compromise bit rate or quality either.

But, for instance, if you are only interested in keeping Alcatraz out of the four programs recorded at the time that airs, the recorded file still holds all 4 and will take up 4 times the space on the PTAT partition (but 8 days is about 500 GB, so there is room).

If you want to keep the program longer, the DVR has to stream the entire file in the background to the demux which separates out the program you want, ignores the rest, and copies that part of the file into a separate location on the user partition, where it takes up the same amount of room it would take if you had originally recorded it on a conventional DVR; and again, this does not compromise quality either.

 

[TomCat]

...if each viewing device/decoder (Hopper or Joey) is accessing the same tuner that is recording PTAT, is the Hopper able to extract a different channels from the same stream using the same tuner to send live to the decoders?...

Playback paths never, on any DVR or STB, access the tuners directly. The tuner has one job, which is to accept the L-band carrier it is tuned to, and pass that signal on to the demod, which converts the signal to baseband (there may also be an IF stage). From there it goes to demux, and from there to the HDD or playout path (HDD then demux for Hopper).

So whether a DVR can pass multiple programs to multiple playout paths at the same time is based entirely on how sophisticated the demux is and how it is programmed. Theoretically, individual streams can be routed to individual paths, but of course in this case only if they programmed it that way. Since the Hopper does not demux at receive, but at playback, it is unlikely that it can demux the signals for live viewing also.

Think of it this way; if the demux could output separate paths for the 4 programs and was early in the chain like a conventional DVR instead of later in the chain at playback, there is no reason why they could not each be recorded to the HDD separately. It appears that they are not, so that does not bode well for the opportunity to feed these signals live to separate client boxes. Since you can simply start playback and watch "virtually" live, it really does not matter.

 

[TomCat]

I've always owned all of my DISH receivers. I'll probably have to bite the bullet and lease these when they come out.

1. Will it be possible to purchase the Hoppers and Joeys outright? I know this problaby doesn't make sense financially, but for the sake of arguement, can it be done?

2 Will it be possible to do a do-it-yourself install? I've never needed an installer's help before, and can't see why I would now.

Thanks.

Since the only real technical change is in how the DVR works internally, there is no technical hurdle making installation any more difficult. You could swap them easily. But there is a paradigm shift in how folks would interface with the new DVRs, so there may be some customer training in order. Limiting them to pro installs gives them the platform to accomplish this training at the same time, and so my best guess would-be that this is how they will do that.

 

[TomCat]

I would say - 'could', but normally signal to TV(s) coming directly from video/audio decompression engines inside of 'CPU' if you are not using DVR keys (pause, etc).
Remember: dish (not DTV) DVRs can works OK without a drive.

"P" makes a point. For DTV DVRs, the decoded signal bypasses the HDD directly after a channel change and is routed straight to the output channel (the signal actually goes to both, but the user sees it live). This is a clever way of shortening channel acquisition time, which became somewhat longer once MPEG4 began to be used (it eliminates the write/read delay of the buffer). Only when you use trick play features does the signal route through the HDD, and the switch is seamless.

This is such a good idea I would have assumed that DISH was doing it too. Are they not?