Homebrew Solutions to HDR-FOX Replacement

Black Hole

May contain traces of nut
My longer-term plan is to transition to a mini-PC solution running e.g. OpenElec (disto based on XMBC/Kodi). that would be with PCI/USB tuners for DVB-T/DVB-S.

My reason for going HDR-FOX when they came out was that a PC-based solution was too expensive and there were no DVB-T2 tuner dongles. No doubt the situation has changed over the last 5 years.

Like BH and cdmackay, I have looked into a PC-based solution as an alternative but even just a pair of USB DVB-T2 tuners costs over £100 so a second-hand 'Fox is a better bet.

You can get a dual-tuner for about £80, but yes, it's not cheap, especially if you want DVB-S too, as I do.

Still, had the software been up to it, several years ago, I'd have spent the money doing it properly. But MythTV, then, seemed a terrible mess, so I gave up after a few months. I'm sure Kodi is much better...

Between us we ought to be able to come up with a reference design, and if we all use the same platform and software it will help with support and give others something to follow.

We should go at least one better than HDR-FOX though - three tuners, four tuners?

Does an RPi have enough oomph?
 
Last edited:
Windows MCE is essentially dead, I believe.
The principle with Kodi is to have entirely separate servers and clients.
You have one, or several, Kodi boxes (for example Amazon Fire TV) attached to your various TVs, as well potentially Kodi running on tablets and "normal" computers. These boxes can deliver content from your local NAS, streamed over the Net, etc, but also communicate with a separate PVR server. This lets you receive live TV by streaming it over your local network, set recordings, watch previously recorded content, etc
RPi is highly recommended for the server, but I don't know if it can handle as many as 4 tuners.
Presumably the file storage requires yet another device such as a conventional NAS, since a RPi can't be directly attached to a SATA drive (?)
Several options exist for the backend software, including MythTV - http://kodi.wiki/view/PVR_backend
Overall this looks like a very powerful and flexible system, but certainly not a "one-box" turnkey solution!
 
I would plan initially to have one single box, which provides both backend and frontend.

I had thought that XBMC (now Kodi) could function as both back- and front-ends. MythTV certainly could, but I found its UI & config unbearable. One could then extend to other devices around the house running front-ends.

One nice feature of splitting front and back ends is that you could have a large PC hidden away somewhere, as the backend. However, in practice this doesn't always work out so well, at least for DVB, since most house designs have the DVB (i.e. aerial) sockets near the TV viewing area, not under the stairs where the backend could reside. So the back-end ends up near the telly, and might as well function as the front-end for that telly.

As to whether one wants 6 DVB-T tuners, or 1 DVB-T and 4 DVB-S tuners (for a quad LNB), should be mostly irrelevant: the base design should scale simply. Once you've worked out how to integrate two tuners (of whatever type) adding more should be simple, in theory.

Finally, for those unfamiliar with the software, OpenElec is a standalone distro based on the XBMC software (now Kodi).

I used to be well-up on all this, some years ago, but gave up and haven't followed it recently, so apols for any errors.
 
the reason for the "large PC" as the backend was based entirely on having multiple tuner PCI cards. It was also important to ensure that the motherboard/graphics support included hardware decoding of video codecs (eg. nVIDIA VDPAU in the old days). trying to do that in software results in high cpu use: heat, noisy fans, etc.

It is now possible to get USB tuners, and of course even the rPi has hardware HD video decoding.

However, USB tuners are not always as good as the PCI card versions, especially for dual-quad tuner cards.

For video decoding, a future-proof solution would include hardware support for the newer H265/HEVC standard, which the rPi does not have. Of course, video decoding is a per-frontend capability.

The newer rev of the Amazon Fire TV box does have HEVC support; as has been noted, it also has a Kodi client, so would make a good front-end.
 
Last edited:
Of course, the backend(s) can include many sources/functions:
  • DVB tuners
  • PVR recording & playback
  • playback of locally recorded material, e.g. archived files from Humax
  • other local material: photos, music, camcorder video
  • LAN DLNA providers
  • internet streaming (youtube, Amazon TV, Google Play Films, BBC iPlayer, etc)
  • other net providers (Google Photos)
  • local gateway devices, eg Chromecast, Apple TV
  • general internet, e.g. web browsing
  • video conferencing
and a complex system would include multiple backends providing different combinations of the above, all accessible via frontends (concurrently), ideally with a unified interface.
 
One nice feature of splitting front and back ends is that you could have a large PC hidden away somewhere, as the backend. However, in practice this doesn't always work out so well, at least for DVB, since most house designs have the DVB (i.e. aerial) sockets near the TV viewing area, not under the stairs where the backend could reside. So the back-end ends up near the telly, and might as well function as the front-end for that telly.
Yes, absolutely right on both counts. And of course this "large PC" could also include a number of hard drives as well (possibly RAID) so it can act as your main NAS as well. The NAS needs to be powered up all the time anyway, so combining this with the other functions that need to be on all the time makes a huge amount of sense.

Several times, I very nearly built such a machine, but backed out in the end when I worked out the cost. Plus Windows MCE uses (used) a stupidly large font that is entirely inappropriate on modern HD TVs, while my Unix knowledge is way too rusty for me to risk trying MythTV.
 
Once you've worked out how to integrate two tuners (of whatever type) adding more should be simple, in theory.
You think? Maybe, but I foresee problems with whatever is managing the recording schedule allocating tuners to events, and warning when events are likely to outrun tuners, unless the manager has been programmed to support n tuners (where n is between 1 and a suitable high number).

With 6 tuners and a big enough HDD, you could just spool TV to disk and pick what you want to watch afterwards!
 
As mentioned briefly in #4 .... "Noise, noise, noise"
If it is going to be a platform for all of 'us' then for many it will be sat near the TV (Display Unit) and hence needs to be quiet.
Given the number of complaints about the 2" fan and single HDD in a FOX-T2 (and predecessors) anything louder won't cut it. That needs to be specced from the start - trying to deaden multiple fans and HDD late in the development will be a disaster.
 
The HDHomeRun DVR software is an interesting product that is currently still under development see HERE The server end can be a PC or a NAS (my preferred location) and it will support a variety of clients, including Kodi. Their network attached 2 tuner DVB-T/T2 boxes are also already available, see HERE
 
Yes, absolutely right on both counts. And of course this "large PC" could also include a number of hard drives as well (possibly RAID) so it can act as your main NAS as well. The NAS needs to be powered up all the time anyway, so combining this with the other functions that need to be on all the time makes a huge amount of sense.

Yes, but, you then may have a problem in that you want the backend near the aerial feeds, and as noted by MikeSh you then may have a problem with heat & noise near the viewer. Although that may not be as big a problem as the other half not liking a monstrous server in the living room. Mind you, that's an excuse for a larger telly, to hide it.


You think? Maybe, but I foresee problems with whatever is managing the recording schedule allocating tuners to events, and warning when events are likely to outrun tuners, unless the manager has been programmed to support n tuners (where n is between 1 and a suitable high number).

With 6 tuners and a big enough HDD, you could just spool TV to disk and pick what you want to watch afterwards!

nice :)

Yes, you need to have the software managing N tuners. What I meant was: when you've sorted that for 2 tuners, I think it should scale to N tuners without great change.


As mentioned briefly in #4 .... "Noise, noise, noise"
If it is going to be a platform for all of 'us' then for many it will be sat near the TV (Display Unit) and hence needs to be quiet.
Given the number of complaints about the 2" fan and single HDD in a FOX-T2 (and predecessors) anything louder won't cut it. That needs to be specced from the start - trying to deaden multiple fans and HDD late in the development will be a disaster.

Good points, but is the T2 really that noisy? I use the fan pkg (minimum 45%), to keep the temp at around 45C, and I can't hear it at all, nor the disk. The fan in my LCD TV makes more noise by far. And the telly sound, of course.


The HDHomeRun DVR software is an interesting product that is currently still under development see HERE The server end can be a PC or a NAS (my preferred location) and it will support a variety of clients, including Kodi. Their network attached 2 tuner DVB-T/T2 boxes are also already available, see HERE

thanks, will take a look
 
Several times, I very nearly built such a machine, but backed out in the end when I worked out the cost. Plus Windows MCE uses (used) a stupidly large font that is entirely inappropriate on modern HD TVs, while my Unix knowledge is way too rusty for me to risk trying MythTV.

From my experience, the issue with MythTV was that the UI was so dire you ended up doing most things directly with mysql :)
 
My question would be "homebrew for what purpose?" And for who.
If it's just an interesting project for a couple of hobbyists anything goes.
If its a serious attempt to replace the HDR-Fox a few important questions
that need asking first seem to have been skipped over.
Or have I missunderstood something?
 
I think the focus is on hobbyists, in my opinion. There are any number of consumer devices out there already, if one wants a more modern replacement for the plain T2.

But to replicate the features of the CF, e.g. decoding of HD material, web control, and all these other fantastic features, there's very little out there, other than something like Myth/Kodi.

The key point has to be its extensiblity/customisation: most mainstream products aren't.

That said, we don't want something that can't easily be used by anyone willing to read up a bit. The CF is a great example of what is possible, without it becoming arcane. Whereas MythTV is the opposite, again in my opinion.

As Black Hole says, what are your thoughts?
 
The way I'm looking at it is not as a commercially-packaged unit - but not as a "dedicated enthusiasts project only" either. Yes, the dedicated enthusiasts would have to take the lead in development, but once the trail is blazed it should be easy enough to follow by buying in the specified parts, wiring them together in the specified way, and installing the specified software - "homebrew PVR by numbers".

If the RPi is powerful enough, that would be a good starting point because it is cheap and readily available with a strong builder community and OS/software support, and not tied into proprietary hardware/software. Kodi is a good software starting point but may require customising later - at least it should get something off the ground, and early results are essential for cultivating enthusiasm.

Tuners - I've not looked into them. Are tuners available as plug-in modules for the RPi, or would we have to settle for USB? That may limit the number of tuners and/or the number of programmes that can be recorded simultaneously. One of the biggest challenges I foresee is finding a way to connect the aerial through to multiple tuners, unless maybe they have pass-through.

I presume a SATA interface is available for RPi. Displays, LED driving, button inputs, etc etc - all no problem in the RPi project world. The RPi has an HDMI video output as standard.

Ultimately, some enterprising person might even come up with a housing for the assembled parts - in the form of a 3D print file for example. Print out and paint to taste.
 
How about a pi running tvheadend with a hd tuner (+/- storage). Then have a plex setup PC with your media on that. I believe there's a script to feed tvheadend to Plex.

Then all you need are plex clients. These can be setup on nearly anything: Pi, android box, ios....

That means your tvheadend can be discrete and you can have your powerhouse decoding server in a cupboard! Throw in a few homeplugs for speed and stability.
 
That's OK for people with fast networks. Might struggle over HomePlug though (mine can't do HiDef streaming without glitches).
 
Back
Top