https://bugs.kde.org/show_bug.cgi?id=387078

--- Comment #3 from Mauro Carvalho Chehab <mchehab+...@kernel.org> ---
(In reply to knossos456 from comment #2)
> Created attachment 108960 [details]
> attachment-14044-0.html
> 
> Yes of course it change the IF frequency and signal strength
> Consider the frequency 11642  on Horbird.
> For lo switch 11700, this gives a IF frequency on cable between lnb and dvb
> card of : 11642-9750 = 1892 Mhz which is a high frequency
> For lo switch 11600, this gives a IF frequency on cable between lnb and dvb
> card of : 11642-10600 = 1042 Mhz which is a more lower frequency

Ok, per your answer, I suspect that you're using an Universal EU LNBf, as this
is the only currently supported LNBf type with has a range of ~200MHz that both
low and high LO can convert into IF.

Considering that usually attenuation is higher for high values of IF, it is
probably a good idea to change libdvbv5 to use 10,6 GHz as the range switch
frequency.

> I think this libdvb5 is to rework completly.
> Most of other DVB software havet this feature, including VLC

Most other DVB software aren't capable of supporting some LNBf types that
libdvbv5 supports. There are several products out there that use more than 2
LO. By moving the LNBf code into the library, we can easily add support for
them on all applications that use the library at once, as, currently, the only
parameter that an application needs to pass is the LNBf model. They don't need
to know the dirty details about each model, or about the cabling system.

It is even possible to add support there for DiSEqC 2.0, with would allow the
library to autodetect the LNBf model and the antenna system configuration,
without changing the API.

Now, we could eventually add some extra API there to allow setting a custom
LNBf. It shouldn't be hard, but probably the best would be to store it on a
configuration file, in a way that people could set it just once for any
libdvbv5-based application.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to