Multistream tuner conflicts

[Arcy]

I have a new Octagon SF-4008, delivered this week with 5.0.025 flashed by the board sponsor. It has two dual DVB-S2X tuner modules (i.e. three usable tuners) which i have connected to sundry LNBs via DiSEqC switches, and I've run into a tuner conflict problem.

If I start a recording on a DVB-S/S2 channel, I find I can't then record at the same time from a multistream channel (such as on 5.0 W or 12.5 W) via another tuner. The box displays the multistream channel, but if I try to record it, it pops up a message: "Could not record due to a conflicting timer <programme title>". Conversely, if I record a multistream channel, I can't then record anything else (multistream or otherwise) until it has finished. It doesn't seem to matter which tuner I make the recording from -- the effect is the same, whether it's on the same stream as the channel being recorded, on a different transponder or even on a different satellite. But no conflict arises with plain DVB-S/S2 channels -- I can record those freely from up to three different satellites simultaneously, as you'd expect.

I hit this problem previously with my Vu+ boxes (Duo2 and Solo4K) when using a TBS5925 USB tuner: recording a multistream channel from this tuner would prevent the box from recording anything else. It was suggested at the time that problems existed with the TBS tuner drivers (certainly my boxes ran more slowly when the TBS tuner was attached). So it's disappointing now to find the same rather serious limitation with the Octagon receiver. Wanting to be able to receive two (or more) multistream channels at the same time must be the main reason for choosing this model, so I am sure that other users must have noticed it. Is there a solution?

[ccs]

How have you configured the tuners?

[Arcy]

Tuner A is in Simple mode, DiSEqC 1.0, with satellites on ports A, B, C, D (5.0W is on port D).

Tuner B is in Advanced mode, with a combination of DiSEqC 1.0. and 1.1 which includes all the satellites on Tuner A plus some more (including 12.5W); its DiSEqC 1.0 switch is cascaded from port 1 of the DiSEqC 1.1 switch.

Tuner C is the same as Tuner B but with some additional satellites on the DiSEqC 1.1 switch. This setup may sound a bit ambitious but it works perfectly with my Vu+ receivers except when I try to receive a multistream channel, as I described in my previous post.

I tried swapping the Octagon tuner modules over, but it didn't make any difference. I haven't tinkered with the tuner priorities.

What's happening is that the box reports a tuner conflict where none exists, when one or more tuners remains free to do the task.

[twol]

So out of curiosity and perhaps simplicity what happens if you only configure Tuner A to include 5.0W and remove it in the configuration for the other tuners and set prefered tuner to B or C?

[Arcy]

Thank you for the suggestion, twol, but I don't see how that would work. Tuner priority is set on a per-satellite basis, not for the tuner as a whole. I only mentioned tuner priority because I fiddled around with it on my Vu+ boxes; the TBS USB tuner appeared as Tuner E, and I wanted to make sure it had a low priority so that it would be left available for any jobs which only it could do -- i.e. multistream services.

However, I've tried running a recording on Tuner B of the SF-4008 -- by starting one on Tuner A first and then stopping it as soon as Tuner B was recording -- and then selecting a new channel, whereupon it appears on Tuner A, as expected. But the box won't let me record it if any of the channels in use is a multistream one.

My DiSEqC system has been working with my Vu+ boxes without problems for more than 18 months and it checks out fully with the DiSEqC tester plugin, so I really don't believe it can be part of the problem.

[twol]

No ... Menu/Setup/Tuner config/settings/prefered Tuner is receiver based
So setting to B or C will make all but 5E go the prefered tuner if only A is configured for 5E.
Just trying to take it back to basics, and see if that would work with 5E on the "simple" config setup
I realise thats not how you want to configure, but how about seeing if it works and then work back to youre prefered config

[Arcy]

Sorry, twol -- I must have overlooked that menu somehow.

I've now deselected 5W on Tuners B and C and have set the preferred tuner to B. When I start a recording of a multistream channel on 5W it's assigned to Tuner A. While the recording is in progress, all other channels on 5W are greyed-out in the channel list except for those on the same stream as the one I'm recording. If I try to record one of those at the same time, I get the conflict message. Meanwhile, I can view channels from other satellites, including multistream ones on 12.5W, but I get the conflict message if I try to record from them. So I don't think the box is behaving any differently like this.

It's as if adding the extra parameters for multistream transponders has made the tuner conflict detection routine too aggressive.

[twol]

Yep interesting. Think it needs someone like Huevos involved, Really don't understand why this multistream would do this ..... which it obviously is doing.
In my simple view of things 5E and multistream is sectioned off to tuner A, so the others should be able to do whatever..... but I guess (?) until it starts transmitting it doesnt know how many streams are in the signal or therefore how many tuners are needed? Maybe I am misunderstanding multistream?

[Huevos]

I hit this problem previously with my Vu+ boxes (Duo2 and Solo4K) when using a TBS5925 USB tuner: recording a multistream channel from this tuner would prevent the box from recording anything else.

Have you tried in other images? e.g. OpenPLi or SatDreamGR?

[twol]

Don't think OpenPli have an image for the SF4008,
Nearest is Openeight which is kind of the "factory" image and a Openpli clone but part of the OE-A

[Arcy]

I decided to start with the download page at the Octagon-Germany site for an 'official' image - first, 'Firmware 5.3 Octagon_SF4008-UHD_DE_04-01-17', which turned out to be an OpenATV image dating from January. Next, after a factory reset, I tried the latest OpenATV 6.1 build, linked from the same page. Then, after another factory reset, the latest build of Openeight-6.0. All of these produced a tuner conflict error when I tried to record a multistream channel plus any other channel in parallel.

I've been exploring the problem mainly by launching ad-hoc recordings with the red button on the handset. But if I program timer recordings in advance by picking items off the EPG, any future conflicts are flagged up there by the timer sanity check.

So I'm now wondering what to make of this. The SF-4008 is the first box of its kind to have DVB-S2X built in -- yet it seems pointless to have this capability in a twin-tuner or triple-tuner version if recordings are going to fail this way because one of them happens to involve a multistream transponder.

[twol]

I kind of expected ATV to be the same, but Open8 means OpenPli as well.

Are there any multistream on 19.2E or 28.2E? If so I can run a test on my HD51 S2x and see if it is a generic issue

[Arcy]

Unfortunately not, twol. The multistream transponders I'm aware of are on 12.5 W and 5.0 W (lots of them) plus one on 9.0 E which I can't receive here. There's a transponder identified as DVB-S2X on 33.0 E, but it's off the end of my toroidal dish.

On 4.8E there are four transponders in T2-MI mode, which is another kind of wrapper. Someone on the Octagon EU forum has helpfully posted a recipe for receiving these using an experimental version of the satdreamgr image, which I'm trying at the moment. But it looks to me as if 12.188 H (with streams 10 and 20) is not on the air at present, while on 12.341 H stream 30 is present but empty and only stream 40 carries any channels; I can receive these channels without the use of the recipe.

The satdreamgr image, like the others I have tried, reports a tuner conflict if I try to record anything else at the same time as a multistream channel.

[twol]

If it changes let me know

[Huevos]

Don't think OpenPli have an image for the SF4008,
Nearest is Openeight which is kind of the "factory" image and a Openpli clone but part of the OE-A

I meant using the Vu+ and TBS.

Fault must be in TimerSanityCheck.py. If you turn off conflict detection everything works fine.