The future of MythTV

[TMonster]

I've just upgraded (to 32), which is about the only time I view the forum, so apologies for the late response

My thoughts... after firstly thanking everyone who's helped to create this fantastic system

Generally everything works fine once it's setup - and that's my main 1st point... my opinion is that the future of MythTV depends on simplification.

This is for everything... the UI, wiki, forum, mailing list, bug tracker all appear to be separate entities - which is probably due to historical reasons as this project has been around for so long

I've updated the Wiki when I've found things "wrong" (out of date), and I've created a couple of bug reports on the old bug tracker, so I've tried to be proactive rather than just a complainer

For us, we almost exclusively use a combined FE/BE for all our entertainment: TV, Videos, Music

It powers up to record TV programs from the internal Hauppauge tuner and shuts down again... We also wake it up via Home Assistant (and use MythWelcome to help prevent it powering back off again for a while) - with the cost of electricity, a permanently on 24/7 box is not an option for us.

With FreeView / DVB broadcasting the schedule data, we don't use Schedules Direct, etc

As many have said, skipping commercials is just fantastic - after watching a film at a friend's house with them trying to fast-forward through adverts (on their paid-for satellite TV system), I was very thankful to all the devs who have worked (past & future) on this great system

Also, having TV programs in profiles is great - I can keep my motorsport recordings in a separate list than my wife's crime dramas and that's just great...

I have never used the "Trailers" feature to watch trailers before a video...

The only plugin we use is MythMusic, which is not intuitive to use, but we stick with it because everything's in 1 place.


Our system runs on Arch Linux, because we had issues with *buntu being so out of date with various components - and in some ways, the Arch Linux wiki was better than the MythTV one when it came to setting up the first install...

As a sweeping statement, the overall UI is dated and overly complex (esp. MythMusic - it's terrible ) - but there's nothing better as a backend.

I've looked at Kodi and while it looked nice with swishy menus, in the end I just wanted a combined system from the same (great) developers. I looked at trying to create my own theme, but got totally overwhelmed with the theming and just gave up - if the UI theming could be improved (or just the documentation?), then I think that would help...


1BE + many FE is also a great feature and I'd use it more if the MythMusic plugin was better as that would be our go-to solution for multi-room audio. Instead we use Volumio for audio (which just clones the MythTV music folder), but that was purely because there was no simply way to use MythMusic on a small device. Volumio has it's own issues, so if I could use multi room audio with MythTV+MythMusic, then that's my preference.

I have no intention of ever subscribing to the streaming services (ie Netflix, etc), but we do watch BBC iPlayer, YouTube, etc - so we just lock Myth, open a web browser and watch them there... if that could be combined into MythTV (ie MythNetVision), then that would be fantastic, but again, MythNetVision had issues (I need to revisit since the upgrade to 32).

I've no idea why the MythTV package includes applications like FFmpeg, when those could use the distro's version (maybe that's due to some distros being slow to update?)

So, again, I think the future is simplification

[minidisk]

been using myth since 2008, loved every minute of it but there are a few things i think need to be addressed

4k HDR support - HLG/PQ/HDR10+, these are desperately needed for videos section and now that atsc 3.0 supports it i think this is a must have. i know there is the android client, that peter created, but a native 4k hdr capable frontend is a big need. the the android exoplayer is a bit finicky with some files and only a few devices support full 7.1 truehd atmos audio.

youtube channel - this is not a specific youtube channel but it would be your accounts subscription results placed in order released. you can play them like a tv channel one after another until they are all watched. make it super simple at first and then ppl can ask for enhancements. this is for the cord cutters that have scaled back off cable tv to OTA and youtube.

simplified transcoding - just need a dummies option, transcode recordings to 264,265,av1, can add commercial cutting later.

[pvr4me]

...simplified transcoding - just need a dummies option, transcode recordings to 264,265,av1

I think the main problem is that there isn't an obvious consensus on the size-quality tradeoff. Some people want to minimize size, say so that they can cram a bunch of movies on an old smartphone. Others want to archive a movie but keep the quality indistinguishable from their initial capture. Others have different target devices and usage goals. Then mix in that encode times are all over the map at least partially due to user's having wildly different platforms used to do the encoding.

Craig

[minidisk]

...simplified transcoding - just need a dummies option, transcode recordings to 264,265,av1

I think the main problem is that there isn't an obvious consensus on the size-quality tradeoff. Some people want to minimize size, say so that they can cram a bunch of movies on an old smartphone. Others want to archive a movie but keep the quality indistinguishable from their initial capture. Others have different target devices and usage goals. Then mix in that encode times are all over the map at least partially due to user's having wildly different platforms used to do the encoding.

Craig

Craig, for my comment on transcoding, its just a disk space saving option for recordings, it doesnt need to be called transcoding, can label it anything. the goal to save disk space by compressing native recordings to more efficient encoding type that some capture devices dont natively support.

[pvr4me]

...simplified transcoding - just need a dummies option, transcode recordings to 264,265,av1

I think the main problem is that there isn't an obvious consensus on the size-quality tradeoff. Some people want to minimize size, say so that they can cram a bunch of movies on an old smartphone. Others want to archive a movie but keep the quality indistinguishable from their initial capture. Others have different target devices and usage goals. Then mix in that encode times are all over the map at least partially due to user's having wildly different platforms used to do the encoding.

Craig, for my comment on transcoding, its just a disk space saving option for recordings, it doesnt need to be called transcoding, can label it anything. the goal to save disk space by compressing native recordings to more efficient encoding type that some capture devices dont natively support.

You've specified part of the equation but you haven't said how much quality you are willing to sacrifice to achieve your goal of saving disk space. You can save massive amount of disk space by dropping to 15 fps, or by cropping to 480, or by sacrificing fine picture detail. There is no free lunch--something of the original will be lost after transcoding. The trouble is that MY tolerance for loss is different from yours and, as I said, there seems to be little consensus. "Ask 6 users and you'll likely get 7 answers!"

Sorry, but I'm just trying to point out that the problem is a little more difficult than you might realize and that the lack of a good general solution may be because there isn't "general" consensus on "good".

Craig

[paulh]

Regarding adding YouTube support to MythTV. It's something I've wanted to add myself for a long time but the uncertain future of the plugins has stalled any developments in that area. My QML frontend does already have YouTube support for example the internal player can play most YouTube video's by itself. Those that wont play can be played in the embedded YouTube player. That's mostly used for the WebCam player but it can be used to play any YouTube video. There is some limited support for subscriptions not exactly like you want but similar for example https://mythqml.net/whatsnew-17.php

The problem with adding similar support to MythTV is the video player wont play YouTube videos's so everything would have to use the embedded player in an HTML web page. While the internal WebKit based plugin will play them OK QTWebKit is EOL in QT. Some distros are already removing it entirely. I don't think QT6 has any support for WebKit at all. There is a replacement for it in QT called QtWebEngine that could be used and I do have a patch to do that but there is an incompatibility between the Qt renderer and the MythUI one that I can't figure out how to fix and no other dev is interested so that looks like a dead end. I've had to admit defeat and thrown in the towel on that one

The other problem for proper YouTube support is how does an open source project like MythTV keep the required API key private so only the MythTV project can use the key. I don't think there is anyway to do that One solution is for any user who wants to use the YouTube support would have to obtain a key for themselves and the plugin would use that but that isn't very user friendly.

[paulh]

Regarding transcoding recordings. Again it's something I've thought of adding to MythArchive. My use case would be to move a recording to Myth's Video section after removing adverts and transcoding it at the same time. Again the uncertain future of the plugins stalled any effort in that area.

My idea would be to have recording profiles similar to how MythArchive has them for creating DVD compatible files but instead they could be used to create any file format mythffmpeg can encode to. The encoding profiles are simple XML based profiles that are simply a template for the parameters to pass to mythffmpeg. From a user point of view they would just pick an encoding profile from a user friendly list and the template would run mythffmpeg using the parameters from the template.

In order to cut the adverts it would be necessary to use mythtransode in fifo mode were it basically plays back the file skipping where necessary which means having to decode the entire file and pass the raw video and audio to mythffmpeg which would then encode the video to whatever format the template has requested. It wont be quick but the cuts should be frame accurate. All sounds pretty easy to do. I might look at how can I add the support to my QML frontend one rainy day.

[paulh]

@TMonster or anyone interested regarding MythMusic. I'm glad someone is using it I've asked this question on the mailing list before and the replies if I'm honest didn't really help me understand what users really want from MythMusic but I'm convinced we get a better class of user on the forum so I sure you will enlighten me with your relies

If you have a clean sheet tell me what you would want from MythMusic. How would the interface look and feel? The only limitation is it must work with a remote or keyboard and work within the confines of the 10 foot interface (it should be usable if you are sat 10 feet away from your TV with your remote).

[heyted]

An easily accessible list of the top 10 or 20 streaming sources based on popularity and location or language would be nice. For the look, I prefer a dark theme. An Android app for MythMusic that plays the locally stored music would be great.

[paulh]

*heyted When you say streaming sources do you mean the radio streams? MythMusic does have a list of 30,000ish radio streams from all over the world. The source we was using to get the stream database does have country and language available and you can search/sort the list MythMusic using those but it does not have any popularity data we can use. The database has gone AWOL so we currently have no way to update the radio stream database. Its was a Russian based site and we originally thought it went away because of the Ukrainian crisis but that may not be the full story. In any case we need a new source for the radio streams database

The MythBuntu theme is pretty dark have you tried that one. Failing that themes are just some xml and a lot of imagination and creativity so anyone, even non coders , can create one

I'm not a big fan of listening to Music on my phone because the sound quality is pretty poor so not sure about music on Android. I assume you mean being able to play music that is stored on the BE machine using the API or MythProtocol? I'm not aware of anything to do that I do have a plugin for VLC that can play recordings, livetv, video, music etc over the MythProtocol that kind of works OK that could be used I guess if there was an Android version of the plugin available. Another thought is if you are good at Angular web development is to create a player in the new WebApp that uses the Services API to download the music tracks and play them in the Web UI.

[heyted]

Yes. I was referring to radio streams. For content stored on my backend machine, I frequently use VLC on my phone to play music using a shared folder. This works fairly well. I will probably keep using this method instead of the myth API or protocol, or I may just copy the files to the phone. The new WebApp sounds interesting. I'm not currently familiar with Angular web development. In my spare time, I'm currently working on a small script that will allow setting a recording from anywhere using titantv.com.

[mattlach]

I know the Open Source community hates it, but the only feature I would add to MythTV is support for DRM so we can record "Copy Once" tagged content while using Cable Cards.

I realize this will probably never happen, but it is something I have wanted for YEARS.

Other than that, I don't really want any features in MythTV.

Honestly, the only thing I use MythTV for is as a dedicated backend, running inside of a Linux container on my KVM/LXC combo server.

All of my frontends run the MythTV plugin under Kodi, and I don't see that changing any time soon.

I think your assessment is partially true, that people are getting increasingly frustrated with the inability to record DRM content, but partially I think it is more than that. The people who have the time to maintain open source systems in their homes tend to skew younger, and younger people are more and more cord cutting, and simply don't have traditional "linear" TV at home at all anymore.

Heck, I'm in my early 40's, and even in my extended circle of friends who are approximately my age, I am one of the dwindling few who still pays for TV service.

Now, the CEO of Netflix is obviously very biased on the topic, but he recently announced that he believes that Linear TV will be dead within 5 to 10 years. This is probably an exaggeration, but I can certainly see tall but dead in the "under 50" age group, and how many in the 50+ age group take time out of their lives to run open source solutions on general purpose PC hardware in their homes?

[nouglywires]

I've been using MythTV for eleven years, for about half of my television viewing. The other half is various streaming services.

MythTV is - by far - the best television experience I have ever used. Its reliable commercial auto-skip, instantaneous jumping, and web interface make for a fabulous, intuitive viewing time. I occasionally copy raw TS files onto a flash drive for travel, and VPN into my home network to add recording rules. MythTV does a perfect job of making my 100+ channels of broadcast TV non-linear. As ATSC 3.0 comes to my area, I look forward to upgrading to the HDHR Flex 4K ATSC 3 and enjoying even more content. I'm ever grateful to the team's amazing achievement.

Maybe I'm short on imagination, but I think the package is wonderful as is. My vote for development direction would be to reduce the number of supported platforms (as long as you include Ubuntu ), simplify setup and maintenance processes, increase self-diagnostic capabilities, and add integrity verification tools. I run into difficult problems once or twice a year - they're fun to solve, but if I had less time or interest I'd have given up long ago. Hardware is now so cheap that perhaps it would be best to target only a couple of reference platforms. With a competent HTPC costing under $300 it might be time to review the economics.

MythTV 32, Ubuntu 22.04.1 LTS. Fanless ASRock (Intel) J5040-ITX motherboard, fanless PSU, 2TB SSD. HDHR Dual tuner.

[warpme]

@TMonster or anyone interested regarding MythMusic. I'm glad someone is using it I've asked this question on the mailing list before and the replies if I'm honest didn't really help me understand what users really want from MythMusic but I'm convinced we get a better class of user on the forum so I sure you will enlighten me with your relies

If you have a clean sheet tell me what you would want from MythMusic. How would the interface look and feel? The only limitation is it must work with a remote or keyboard and work within the confines of the 10 foot interface (it should be usable if you are sat 10 feet away from your TV with your remote).

Paul,
Im using MythMusic regularly and like it.
What is a bit annoying for me is lack of "browse and play"
It might be additional option in MythMusic main pop-up menu and working like i.e this:
-user enters "browse and play mode"
-MythMusic start to present songs collection the same like in "edit playlist mode"
-user can walk/browse by arrow keys, pressing enter starts playing song
-in MythMusic settings user has option: "default mode when MythMusic starts: (1)play playlist / (2)browse and play". When (1) is set - starting MythMusic is like today. When (2) - starting MythMusic gives user initial state like after selection of "browse and play"

Currently i'm doing convolutions to achieve something like this with https://github.com/warpme/minimyth2/blo ... edit.patch and mapping remote key
but this is workaround....

[mikenick42]

I've been checking this thread for a while and wasn't sure what to add, but figured it's about time to at least try to get my thoughts down.

I've been using MythTV since at least 2005, moving from Knoppmyth to MythBuntu, from a single frontend/backend combo to a master backend with remote frontends, to a virtualized backend, and I don't even know which capture cards over the years ending up with a Ceton Infinitv PCIE6.

The MythWeb interface was second-to-none for scheduling recordings, the new version? isn't quite there anymore, and the commercial skipping is great. I also spent a lot of time getting all my optical media ripped into MythVideo.

That said, I find the frontend computer being used less and less these days as streaming takes the place of OTA television. The Video interface is also pretty slow with a large library (maybe I have a bad setting checked off, but I haven't had the energy to go fiddle with it too much). I've been testing out Jellyfin because the interface is a lot faster and smoother when it comes to choosing or finding what to watch, and the ability to cast directly to the television is great. On the other hand, they really don't have the "10 foot interface" worked out nearly as well as Myth, and as a new project it has significantly more bugs. Jellyfin also doesn't support my Infinitv capture card (and likely never will).

Not sure I really have a good suggestion about the future, I think someone mentioned more plugins, or MythTV becoming more of an OTA library that could be a plugin for other programs.

As a longtime user and fan of MythTV, I felt the need to say something and I would hate to see it end, but I'm not sure there's much of a future for broadcast/cable television anymore.