free/open DAB+ encoder

[Gro Bil]

Hello everybody

Found of Linux and radio transmissions, and active member of the HAM radioamateur community, i am in charge of finding a multiplexing/broadcasting solution for a pool of associatives radios, i have read almost all your contributions on this group, and the first thing i wish, is to thanks everbody for sharing your infos and experiences here !

i have a lot of questions, don't hesitate to tell me if i'm wrong :

-I saw a lot of excellent "experimentations" here (KANAL PLUS, EBU Radio Week, Geneva Experiment, ...) but, is the Ettus hardware with wbx and mmbtools (plus ampli/filter/antenna) can be a viable and reliable solution for real broadcasting (hangs, radio compatibilities problems, lack of functions like text services or live PTY modifications, etc...) ?

-Did someone tried to use the new free/open DAB+ encoder here : https://github.com/piratfm/fdk-aac-dabplus ?

-Since it does not have the best radio flatness and lacks the "External Reference Input" / "1 PPS input" for SFN synchronization, the USRP1 seems to be easier to install and more stable than the B100. Does somebody successfully use the Ettus B100 to broadcast ? What is the best choice to start ?

-Did someone tried the Matthias Brändli's SFN patch here : http://www.mpb.li ?

thanks

Loic

[Matthias P. Braendli]

Hi Loic,

good to see that you are interested in this technology, please don't
hesitate to ask questions !

On 27. 04. 13 18:06, Gro Bil wrote:
> -I saw a lot of excellent "experimentations" here (KANAL PLUS, EBU Radio
> Week, Geneva Experiment, ...) but, is the Ettus hardware with wbx and

> mmbtools (plus ampli/filter/antenna) can be a viable and **reliable

> solution for real broadcasting (hangs, radio compatibilities problems,
> lack of functions like text services or live PTY modifications, etc...) ?

Well, there have been some transmissions that lasted longer than what
could be called a "just a test", and I believe the open-source mmbTools
are an excellent alternative to commercial equipement, especially if you
look at the price. However, with all things open-source, you have to
expect some rough edges. If you set up your broadcast yourself, you have
to acquire a lot of knowledge, and it can be difficult to find
guaranteed support in case of issues (you could maybe find someone who
agrees to be paid for giving support). We (on this list) are glad to
help, but there's only so much we can do in our spare time.

Mathias Coinchon knows better in which projects/trials the mmbTools
are/have been used, I hope he can give you more details.

> -Did someone tried to use the new free/open DAB+ encoder here :
> https://github.com/piratfm/fdk-aac-dabplus ?

It's quite recent, I haven't had the time to do a test. I've been told
that it works, but I believe no long-term test has been done.

> -Since it does not have the best radio flatness and lacks the
> "External Reference Input" / "1 PPS input" for SFN synchronization,
> the USRP1 seems to be easier to install and more stable than the
> B100. Does somebody successfully use the Ettus B100 to broadcast
> ? What is the best choice to start ?

My impression is that Ettus is still selling the USRP1 only because the
B100 cannot have two daughterboards simultaneously. The B100 is in my
eyes the successor of the USRP1, and will probably receive improvements
(including its driver), whereas the USRP1 will be less and less used.
But I might be wrong.

Most broadcasts have used USRP1s, but I also have had a B100 running for
several days. The next step would be to test continuous transmission
over several months. So I can only invite you to help us with that :-)

> -Did someone tried the Matthias Br�ndli's SFN patch here :
> http://www.mpb.li ?

Yes, I did :-D

I don't know if you've skimmed through my report, but to summarise, I
have not been able to do an SFN transmission using two GPS-disciplined
oscillators, because one of the two was broken. So more testing is
needed for sure.

The other aspect--in my eyes crucial if you want to run a broadcast with
several transmitters--is the monitoring and control of the transmitters
(Receiving alerts when something goes wrong, etc.).


You see that some stuff is still quite experimental, and the solution
needs to see improvements, but it's already useable for many interesting
projects. I hope this answers some of your questions !

Good luck and 73

Matthias
HB9EGM

[Gro Bil]

hi Matthias !

On Sunday, April 28, 2013 7:30:45 PM UTC+2, Matthias Brändli wrote:

Hi Loic,

good to see that you are interested in this technology, please don't
hesitate to ask questions !

i'll remember that one, thanks
 

On 27. 04. 13 18:06, Gro Bil wrote:
> -I saw a lot of excellent "experimentations" here (KANAL PLUS, EBU Radio
> Week, Geneva Experiment, ...) but, is the Ettus hardware with wbx and
> mmbtools (plus ampli/filter/antenna) can be a viable and **reliable
> solution for real broadcasting (hangs, radio compatibilities problems,
> lack of functions like text services or live PTY modifications, etc...) ?

Well, there have been some transmissions that lasted longer than what
could be called a "just a test", and I believe the open-source mmbTools
are an excellent alternative to commercial equipement, especially if you
look at the price. However, with all things open-source, you have to
expect some rough edges. If you set up your broadcast yourself, you have
to acquire a lot of knowledge, and it can be difficult to find
guaranteed support in case of issues (you could maybe find someone who
agrees to be paid for giving support). We (on this list) are glad to
help, but there's only so much we can do in our spare time.

Mathias Coinchon knows better in which projects/trials the mmbTools
are/have been used, I hope he can give you more details.

 

 i love to spend hours (or days) on configuration files...  ;-)

> -Did someone tried to use the new free/open DAB+ encoder here :
> https://github.com/piratfm/fdk-aac-dabplus ?

It's quite recent, I haven't had the time to do a test. I've been told
that it works, but I believe no long-term test has been done.

so, i'm gonna test it

> -Since it does not have the best radio flatness and lacks the
> "External Reference Input" / "1 PPS input" for SFN synchronization,
> the USRP1 seems to be easier to install and more stable than the
> B100. Does somebody successfully use the Ettus B100 to broadcast
> ? What is the best choice to start ?

My impression is that Ettus is still selling the USRP1 only because the
B100 cannot have two daughterboards simultaneously. The B100 is in my
eyes the successor of the USRP1, and will probably receive improvements
(including its driver), whereas the USRP1 will be less and less used.
But I might be wrong.

Most broadcasts have used USRP1s, but I also have had a B100 running for
several days. The next step would be to test continuous transmission
over several months. So I can only invite you to help us with that :-)

ok, B100 WBX bundle will we ordered tomorrow

> -Did someone tried the Matthias Br�ndli's SFN patch here :

> http://www.mpb.li ?

Yes, I did :-D

I don't know if you've skimmed through my report, but to summarise, I
have not been able to do an SFN transmission using two GPS-disciplined
oscillators, because one of the two was broken. So more testing is
needed for sure.

yes i read that, but as this is a long term concern, i will try to test it later
 

The other aspect--in my eyes crucial if you want to run a broadcast with
several transmitters--is the monitoring and control of the transmitters
(Receiving alerts when something goes wrong, etc.).

i've already thought about a php online management page, i have ideas for the linux server(s) and processes monitoring, but for the rf part... maybe someone have ideas about rf dab radio signal and BER supervision with cheap dab usb key or usb connectible radio ?

You see that some stuff is still quite experimental, and the solution
needs to see improvements, but it's already useable for many interesting
projects. I hope this answers some of your questions !

Good luck and 73

Matthias
  HB9EGM

yes, i'm part of the adventure now !

thanks a lot

Loic F4HBG

[Brendan Kehoe]

Hi,

I've been testing the new dabplus codec on our mux for a little while now and it seems perfectly stable. Currently have 3 services using it and the rest (7) with toolame. The audio quality is much more on par with commercial DAB+ codecs which is great news as when we've trialled the CRC codec we were never able to replicate the results yielded by commercial codecs, albeit its over  year since we last used it so maybe things have changed.

B. 

--
You received this message because you are subscribed to the Google Groups "mmbtools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to crc-mmbtools...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Gro Bil]

hi Brendan

good to hear that ! thanks !

It seems that openmokast is MP2 but not HE-AAC V2 compatible as is. Is there a way to test DAB+ ETI with openmokast ?

Loic

On Tuesday, April 30, 2013 9:59:53 AM UTC+2, Brendan Kehoe wrote:

Hi,

I've been testing the new dabplus codec on our mux for a little while now and it seems perfectly stable. Currently have 3 services using it and the rest (7) with toolame. The audio quality is much more on par with commercial DAB+ codecs which is great news as when we've trialled the CRC codec we were never able to replicate the results yielded by commercial codecs, albeit its over  year since we last used it so maybe things have changed.

B. 

[Brendan Kehoe]

Hi,

It's quite some time since I used openmokast as the USB dab receiver I had doesn't work very well but I don't remember there being a problem in listening to dab+ broadcasts with it. The windows software also works with it.

B

Sent from my iPhone

[Coinchon, Mathias]

Hello,

 

Openmokast can decode DAB+ but it has been disabled for licensing reasons. You may be able to re-enable it...

Otherwise, there's also this project for SDR DAB+ decoding: http://www.sdr-j.tk

 

Mc

---

**************************************************
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed.
If you have received this email in error, please notify the system manager. This footnote also confirms that this email message has been swept by the mailgateway
**************************************************

[Ulrik Brinck]

Hi Loic,

>Hello everybody
>
>Found of Linux and radio transmissions, and active member of the HAM radioamateur community,
>i am in charge of finding a multiplexing/broadcasting solution for a pool of associatives radios, i
>have read almost all your contributions on this group, and the first thing i wish, is to thanks
>everbody for sharing your infos and experiences here !
>
>i have a lot of questions, don't hesitate to tell me if i'm wrong :
>
>-I saw a lot of excellent "experimentations" here (KANAL PLUS, EBU Radio Week, Geneva
>Experiment, ...) but, is the Ettus hardware with wbx and mmbtools (plus ampli/filter/antenna) can
>be a viable and reliable solution for real broadcasting

I will say yes, for community radio with just one transmitter site, it absolutely is. Because this is indeed what we do here at Kanal Plus. We have a three years trial license and have been on the air continously since november 2011. The trial is not only about finding ways to build a transmitter, but also to use the transmitter for broadcasting of various sorts of community radio to get experience with that, how is it best done technically and practically? Do people listen to it? Etc.

So, although it is a time limited trial license (valid until 31-05-2014), it is real broadcasting, and we do it pretty much the same way as if it was a permanent license, and some of the broadcasted content is not available on FM.

>(hangs,

We don't have any problems at all with hangs. Since november 2011 we have only had two unexplained breakdowns, and we know that on both days there were workers at the transmitter tower mounting cables for LTE telephone base stations very close to our equipment, so although we don't know exactly what happened, I'm pretty sure that it had something to do with that. Apart from that, we haven't had any breakdowns, hangs etc. at all, except for electricity dropouts (you could place an UPS before the pc and USRP if you want to minimize the risk of that).

>radio compatibilities problems,

We have actually found one receiver, Lingo iMini, which refuses to play three of our channels, while all our other channels can be played.
http://www.lingodab.co.uk/product_imini.php
The behaviour is a bit strange, because one of the channels which can't be played has exactly the same parametres as another channel which CAN be played. But I haven't yet had the opportunity to test it deeply. We haven't heard of problems with any other receiver, and I therefore currently believe it is a bug in the iMini rather than a problem with mmbTools.

>lack of functions like text services or live PTY modifications, etc...) ?

It is not possible to change the PTY, change station name labels or adding new services/stations "on the fly", any of this requires a restart of the multiplex. For community radio, I think that we can live with it being like this, but maybe large commercial stations etc. wouldn't want to. But since CRC-DabMux is open source, anyone who needs such possibilities and has the necessary programming skills, could add them.

Text services (DLS) are added by the encoder. The paid encoder (CRC-DabPlus) supports it, and texts can be live modified, new texts are uploaded by TCP/IP, but this feature is not fully developed and doesn't work very well and can for example crash the encoder if the TCP/IP connection is broken in an "unauthorized" way. So currently we don't use texts so much. I don't know whether the new free DAB+ encoder supports text services, it seems not to be mentioned on it's web page, so perhaps not yet.

So, all in all: Yes, in my opinion Ettus/WBX/mmbTools as it is now is a viable and reliable solution for real community radio broadcasting, as long as you don't use text services (much) and can live without the possibility to modify PTY, names, bit rates etc. or add/remove services "on the fly".

>-Did someone tried to use the new free/open DAB+ encoder here : https://github.com/piratfm/fdk-aac-dabplus ?

We haven't yet tried it, because we already have the paid encoder, but Brendan Kehoes post about a.o. the audio quality sounds interesting.

>-Since it does not have the best radio flatness and lacks the "External Reference Input" / "1 PPS input"
>for SFN synchronization, the USRP1 seems to be easier to install and more stable than the B100.
>Does somebody successfully use the Ettus B100 to broadcast ? What is the best choice to start ?

We are running with USRP1 and WBX (revision 2) and the "classic" (non-UHD) interface in Gnuradio 3.3 in combination with CRC-Dwap.py. This is a rock solid combination and is what definitely worked best for us at the time when we made the transmitter. We also tried B100/WBX with UHD interface and actually broadcasted with it for three weeks, but we had synchronization glitches now and then, knocking some receivers off, and we never succeeded to get rid of this problem 100 percent. So in the end, we rolled back to USRP1 and "classic" interface. But that was in 2011, and the UHD based solution could have become better since then. For the hardware itself, B100 is surely a better box which gives a more clean output.

The situation is:
- "Classic" interface only works with USRP1 and WBX boards and has been phased out in newer versions of Gnu Radio. And I understand from a recent post in this group, that is doesn't work with new WBX boards (WBX revision 3). But it is easy to get a good/perfect DAB+ result with this solution.
- UHD interface comes from Ettus Research themselves and should work with all USRP models and daughter boards, both now and in the future, but my impression is that it currently seems to take a lot more work with tweaking various parametres to get the best result (i.e. suitable for real broadcasting). I however also understand from posts in this group, that it CAN be done.

So your best choice to start depends of what you want (and apparently of whether you can get an old WBX board). Do you want the currently easiest solution ("classic"), or do you want the UHD based solution, which is more future proof, but could take more work to get the best result?
- For "classic" you will need USRP1 and WBX revison 1 and 2 (or perhaps with BasicTX, but that board apparently doesn't support the full DAB frequency band). I don't know whether a revision 2 WBX can still be bought from Ettus.com, you could ask them.
- UHD should work with any USRP model and daughter board, so B100 and WBX could be a good choice here.

Best regards,
Ulrik.

[Gro Bil]

Hi Brendan,

Unfortunately it seems that DAB+ have been disabled on version 0.6.2.

Loic

[Gro Bil]

hi "mc_ebu" :-)

i didn't manage to find something in the source, any clue ?

in addition, i have a compilation error with DABCELT 0.1 (i have created a new topic for this : DABCELT 0.1 source compilation error)

i will try sdr-j, thanks for the info

Loic

[Matthias P. Braendli]

On 30. 04. 13 15:00, Coinchon, Mathias wrote:
> Otherwise, there's also this project for SDR DAB+ decoding:
> http://www.sdr-j.tk

Ah that's cool, haven't heard of it !

mpb

[Gro Bil]

On Wednesday, May 1, 2013 2:14:16 AM UTC+2, Ulrik Brinck wrote:

Hi Loic,

 

 Hi Ulrik

 Thanks for this complete answer !

>Hello everybody
>
>Found of Linux and radio transmissions, and active member of the HAM radioamateur community,
>i am in charge of finding a multiplexing/broadcasting solution for a pool of associatives radios, i
>have read almost all your contributions on this group, and the first thing i wish, is to thanks
>everbody for sharing your infos and experiences here !
>
>i have a lot of questions, don't hesitate to tell me if i'm wrong :
>
>-I saw a lot of excellent "experimentations" here (KANAL PLUS, EBU Radio Week, Geneva
>Experiment, ...) but, is the Ettus hardware with wbx and mmbtools (plus ampli/filter/antenna) can
>be a viable and reliable solution for real broadcasting

I will say yes, for community radio with just one transmitter site, it absolutely is. Because this is indeed what we do here at Kanal Plus. We have a three years trial license and have been on the air continously since november 2011. The trial is not only about finding ways to build a transmitter, but also to use the transmitter for broadcasting of various sorts of community radio to get experience with that, how is it best done technically and practically? Do people listen to it? Etc.

So, although it is a time limited trial license (valid until 31-05-2014), it is real broadcasting, and we do it pretty much the same way as if it was a permanent license, and some of the broadcasted content is not available on FM.

This is very good news

>(hangs,

We don't have any problems at all with hangs. Since november 2011 we have only had two unexplained breakdowns, and we know that on both days there were workers at the transmitter tower mounting cables for LTE telephone base stations very close to our equipment, so although we don't know exactly what happened, I'm pretty sure that it had something to do with that. Apart from that, we haven't had any breakdowns, hangs etc. at all, except for electricity dropouts (you could place an UPS before the pc and USRP if you want to minimize the risk of that).

Very good news too !

>radio compatibilities problems,

We have actually found one receiver, Lingo iMini, which refuses to play three of our channels, while all our other channels can be played.
http://www.lingodab.co.uk/product_imini.php
The behaviour is a bit strange, because one of the channels which can't be played has exactly the same parametres as another channel which CAN be played. But I haven't yet had the opportunity to test it deeply. We haven't heard of problems with any other receiver, and I therefore currently believe it is a bug in the iMini rather than a problem with mmbTools.

We are going to buy some receivers for testing purposes, i will post the result chart as soon as we made the tests
 

>lack of functions like text services or live PTY modifications, etc...) ?

It is not possible to change the PTY, change station name labels or adding new services/stations "on the fly", any of this requires a restart of the multiplex. For community radio, I think that we can live with it being like this, but maybe large commercial stations etc. wouldn't want to. But since CRC-DabMux is open source, anyone who needs such possibilities and has the necessary programming skills, could add them.

Text services (DLS) are added by the encoder. The paid encoder (CRC-DabPlus) supports it, and texts can be live modified, new texts are uploaded by TCP/IP, but this feature is not fully developed and doesn't work very well and can for example crash the encoder if the TCP/IP connection is broken in an "unauthorized" way. So currently we don't use texts so much. I don't know whether the new free DAB+ encoder supports text services, it seems not to be mentioned on it's web page, so perhaps not yet.

So, all in all: Yes, in my opinion Ettus/WBX/mmbTools as it is now is a viable and reliable solution for real community radio broadcasting, as long as you don't use text services (much) and can live without the possibility to modify PTY, names, bit rates etc. or add/remove services "on the fly".

 

yes i think we can "live" without it ;-)

>-Did someone tried to use the new free/open DAB+ encoder here : https://github.com/piratfm/fdk-aac-dabplus ?

We haven't yet tried it, because we already have the paid encoder, but Brendan Kehoes post about a.o. the audio quality sounds interesting.

yep, perhaps we are going to use CRC-DABPLUS. I'm still waiting for the CRC quote...

>-Since it does not have the best radio flatness and lacks the "External Reference Input" / "1 PPS input"
>for SFN synchronization, the USRP1 seems to be easier to install and more stable than the B100.
>Does somebody successfully use the Ettus B100 to broadcast ? What is the best choice to start ?

We are running with USRP1 and WBX (revision 2) and the "classic" (non-UHD) interface in Gnuradio 3.3 in combination with CRC-Dwap.py. This is a rock solid combination and is what definitely worked best for us at the time when we made the transmitter. We also tried B100/WBX with UHD interface and actually broadcasted with it for three weeks, but we had synchronization glitches now and then, knocking some receivers off, and we never succeeded to get rid of this problem 100 percent. So in the end, we rolled back to USRP1 and "classic" interface. But that was in 2011, and the UHD based solution could have become better since then. For the hardware itself, B100 is surely a better box which gives a more clean output.

The situation is:
- "Classic" interface only works with USRP1 and WBX boards and has been phased out in newer versions of Gnu Radio. And I understand from a recent post in this group, that is doesn't work with new WBX boards (WBX revision 3). But it is easy to get a good/perfect DAB+ result with this solution.
- UHD interface comes from Ettus Research themselves and should work with all USRP models and daughter boards, both now and in the future, but my impression is that it currently seems to take a lot more work with tweaking various parametres to get the best result (i.e. suitable for real broadcasting). I however also understand from posts in this group, that it CAN be done.

So your best choice to start depends of what you want (and apparently of whether you can get an old WBX board). Do you want the currently easiest solution ("classic"), or do you want the UHD based solution, which is more future proof, but could take more work to get the best result?
- For "classic" you will need USRP1 and WBX revison 1 and 2 (or perhaps with BasicTX, but that board apparently doesn't support the full DAB frequency band). I don't know whether a revision 2 WBX can still be bought from Ettus.com, you could ask them.
- UHD should work with any USRP model and daughter board, so B100 and WBX could be a good choice here.

 

As you said, it seems that B100 and WBX v3 are more "futureproof"... so we are going to try this configuration

Best regards,
Ulrik.

Thanks again Ulrik, i will post here my experience with B100/WBX v3

Loic

[Gro Bil]

hi everybody

just a little message to inform you guys that we have stopped/paused the project because in 3 months and several mails,  we were unable to obtain any quote from crc about crc-dabplus (in fact we never obtain any answer at all)

too bad... but thanks for everybody's help

loic F4HBG

[Ulrik Brinck]

It's sad to hear that you stopped the project. :( Have you considered trying without crc-dabplus and instead use the new free encoder, as mentioned in your first message in this thread?

Best regards,
Ulrik.

----- Original Message -----
From: Gro Bil
To: crc-mm...@googlegroups.com
Sent: Sunday, June 09, 2013 12:17 PM
Subject: Re: free/open DAB+ encoder

[Gro Bil]

hi Ulrik

i discovered two problems :

-the "new free encoder" can not be used "as is" and i do not have the knowledge to use it in the dab+ creation process :-o
-the "new free encoder" can not handle DLS text, and it is a necessary feature for us

i have another project for the summer, i hope i'll find a solution before october

thanks for your answer

Loic

[Matthias P. Braendli]

Salut Lo�c,

�a viendra, �a viendra :-)

These tools will continue to evolve, maybe in october you'll find them
good enough for your needs.

Good luck for your other projects, we hope we'll hear again from you soon,

73

Matthias

[Rash]

Hi Matthias,

The report I wrote on my DAB experiment is available from here:

http://stakeholders.ofcom.org.uk/binaries/research/radio-research/Software-DAB-Research.pdf

I believe the only typo that remained just happened to be at the end, where you an e other dudes are credited...LOL!

Hope you enjoy! Much more work remains to be done, but the reaction within the UK is very positive.

Best regards,

Rash.

[Matthias P. Braendli]

:-)

Thanks for the link to this research report !

mpb

[François Lefebvre]

Hi Rashid,

Thanks for this great contribution and the credit notes to CRC.

Your paper is generating a lot of interest.

For those who have not seen them, here are two follow-up stories:

http://www.theregister.co.uk/2013/08/06/open_source_hacks_dab_to_the_masses/

http://www.southgatearc.org/news/august2013/sdr_dab_broadcasting.htm#.UgFdgJLVCtM

Best regards,

François.

--

[Rash]

Dear François,

What can I say? Thank you for your kind words, but it is you and your colleagues at CRC who turned an idea into reality. Without your massive contribution, experiments such as mine would still just be ideas!

With gratitude, and warm regards,

Rash.

Sent from my iPad

[Matthias P. Braendli]

Hi Giovanni,

I believe there is quite some interest in that feature among people in
the mmbtools group. Have you been able to progress on this ?

Best regards,
mpb

On 26. 07. 13 17:39, Giovanni Ojan wrote:
> Hi Gro!
> I'm trying to add PAD support to DAB+ and I'm very close to find the
> solution, but i need to ask you some question if you can.
>
> Giovanni

[Kofi Obiri-Yeboah]

Hullo Rash,

I just came across your DAB experiment report.

I noticed that it is a non technical document - will it be possible to get access to the full technical process you used in your tests? We are interested in duplicating what you did. That will be an invaluable help to us as we transition from Internet television to digital radio - specifically DAB / DAB+. While we are physically located in Canada - where DAB/DAB+ - is for all intent and purposes, dead, our operational interest for digital radio is in Ghana and Nigeria and other African countries  BTW background wise I am from (ex) Air Force Avionics / Telecoms

Your report did indicate that you funded the tests completely on your own. That is very noble and inspiring. In that same spirit, we would not mind making a contribution to one of your favorite charities

Cheers

Kofi

[Rash]

Hi Kofi,

On 29 Nov 2013, at 20:13, Kofi Obiri-Yeboah <kofi.obi...@gmail.com> wrote:

Hullo Rash,

I just came across your DAB experiment report.

I noticed that it is a non technical document - will it be possible to get access to the full technical process you used in your tests?

The limitations of the detail published in the report was intentional - please allow me to explain why...

We are interested in duplicating what you did.

Which brings me nicely to the main reason why I did not publish a 'playbook'. The motivation to do the test (which was by no means a first!) was among other things to try and stimulate some innovation in the radio industry and encourage other engineers to experiment, develop, and hopefully create something new and exciting. Sadly (from my perspective at least) innovation has been on the decline in the industry for a while now, and few people seem prepared to challenge the status quo.

Duplication is not really desirable - development is. The only difference between my test and the ones before it (and some that continue to this day) is that I wrote some stuff about it and published it....and the timing just happened to be right which is why I think my test has generated some interest in particular.

That will be an invaluable help to us as we transition from Internet television to digital radio - specifically DAB / DAB+. While we are physically located in Canada - where DAB/DAB+ - is for all intent and purposes, dead, our operational interest for digital radio is in Ghana and Nigeria and other African countries  BTW background wise I am from (ex) Air Force Avionics / Telecoms

You are in a good place there! The mux and mod software was written and open sourced by the brilliant guys at the Communications Research Centre Canada. The UK is in a slightly different situation to almost anywhere else in terms of digital radio - we've had it for a long time now. Had we rolled out on L Band, like Canada, we'd almost certainly have a dead platform on our hands too. I believe around 15,000,000 sets have been sold to date in the UK (the majority of which are not DAB+ compatible) so we have a big legacy to consider. The EBU R.138 also suggests the following...*for Europe*:

1. Immediate deployment be done using DAB transmission as defined in ETSI EN 300 401 with DAB+ services as defined in ETSI TS 102 563 for digital radio broadcasting in VHF Band III;

2. When DAB coverage is not possible, to use DRM as defined in ETSI ES 201 980 for digital radio broadcasting in the frequency bands currently used for analogue radio broadcasting.

   -----

Ergo, DAB is the primary platform in Europe, and DRM the second choice where it has first been proven that DAB coverage is not possible. We need a standard that works across the whole of Europe, DAB is established, has broadcasters on it, half of all homes have at least one receiver, and there are receivers on the shelves at quite low cost. DRM does not have these things in any meaningful way yet. 

DAB is a robust, high capacity system which is quite spectrally hungry (and was specifically designed to be wideband to overcome the multipath problems associated with narrowband systems such as FM) but if one uses interleaved spectrum there is low opportunity cost so it is still spectrally efficient, and in rural areas there are fewer pressures on spectrum - it isn't wasteful to make use of spectrum that would otherwise be fallow, and going to waste! You can also fit a large number of programmes on the same transmission infrastructure - arguably this could save money through reduced CAPEX (1 tx for many instead of 1 tx per programme) and OPEX.

DRM+ is a low capacity, narrowband system which can mimick the coverage of existing FM (requiring the broadcaster to bear the entire CAPEX and OPEX of the plant) the transmitters and receivers are likely to need a bit less power, but multipath is a potential problem too. What I am getting at is that it is not just about DAB as all systems have pros and cons - DRM could possibly fit the requirements of Ghana, Nigeria etc better than DAB does - I don't know those markets very well at all. You probably have a blank canvas though, so do consider that DAB might not necessarily be the best solution for the region. I imagine if I were in your situation I might be inclined to take a look into DVB-T in band III for delivering both digital radio and TV - in fact I'd look at everything to ensure that I made the best possible informed choice for the market. Do look at everything closely yourself, because the proponents of each system will say quite extraordinary things to make sure you choose them, and not the competition! You should do this exercise before deciding that DAB is the right thing for the country.

Your report did indicate that you funded the tests completely on your own. That is very noble and inspiring. In that same spirit, we would not mind making a contribution to one of your favorite charities

Ah, the reality is my motivation was to learn, as well as to try and encourage others and stimulate debate. What I learned I could not have bought at any price so it was well worth it. I love radio, and I want to see it move forward and not stagnate while slowly being eclipsed by the Internet. Getting into a project and getting your hands dirty is the best way to become intimately familiar with any subject. Do that and you'll know more than any book (or technical report) could ever tell you!

I hope this makes sense? If you do decide to press ahead with looking into the open source DAB, then you will find lots of help from the wonderful people in this group, and also don't forget there is the opendigitalradio.org site which has many tutorials and example scripts.

Best regards,

Rash.

Cheers

Kofi

On Monday, August 5, 2013 1:16:58 PM UTC-7, Rash wrote:

Hi Matthias,

The report I wrote on my DAB experiment is available from here:

http://stakeholders.ofcom.org.uk/binaries/research/radio-research/Software-DAB-Research.pdf

I believe the only typo that remained just happened to be at the end, where you an e other dudes are credited...LOL!

Hope you enjoy! Much more work remains to be done, but the reaction within the UK is very positive.

Best regards,

Rash.

On Monday, August 5, 2013 1:16:58 PM UTC-7, Rash wrote:

Hi Matthias,

The report I wrote on my DAB experiment is available from here:

http://stakeholders.ofcom.org.uk/binaries/research/radio-research/Software-DAB-Research.pdf

I believe the only typo that remained just happened to be at the end, where you an e other dudes are credited...LOL!

Hope you enjoy! Much more work remains to be done, but the reaction within the UK is very positive.

Best regards,

Rash.

--

[Ken]

Hi,

Finally I got my B100 working with DAB+ with "aac-enc-dabplus" but have some questions.

1. Is there a difference in sound quality of "crc-dabplus" and "aac-enc-dabplus"? Or is it that DAB+ just sound bad..? 

Retransmit an Internet stream (128k MP3) is out of the question, wav file, or via sound card sounds better. I use the '-a' option.

2. Will there be "DLS" live-text support? I see two possible options, using TCP or that "aac enc-dabplus" regularly read info from a text file.

/Ken

[Matthias P. Braendli]

Hi Ken,

On 30. 11. 13 16:10, Ken wrote:
> Hi,
>

> Finally I got my B100 working with DAB+with "aac-enc-dabplus" but have
> some questions.
>
> 1.Is there a difference in sound quality of "crc-dabplus"and

> "aac-enc-dabplus"? Or is it that DAB+ just sound bad..?
> Retransmit an Internet stream (128k MP3) is out of the question, wav
> file, or via sound card sounds better. I use the '-a' option.

I don't know about the quality difference between aac-enc-dabplus and
crc-dabplus, I never had the right equipment to compare both. (And I
don't have a very trained ear anyway)

However, I can clearly confirm that the cascading of mp3 and aac+
results in a loss of quality. If your input is a webstream, make sure
you have something better than mp3 128k.



> 2. Will there be "DLS" live-text support?I see two possible options,

> using TCP or that "aac enc-dabplus" regularly read info from a text file.

DLS still has to be implemented for fdk-aac-dabplus. crc-dabplus has
some support for DLS.

mpb

[Brendan Kehoe]

There's a great quality difference between crc-dabplus and the open encoder. The open encoder is the closest to the sound of commercial encoders that I've heard, but its not quite there yet but still excellent especially considering its free! It's also 100% stable. I've had 3 or 4 instances of it running with sources via JACK with 6 or 7 standard toolame encoders on our mux for several months without restarts required. 

--
You received this message because you are subscribed to the Google Groups "mmbtools" group.

To unsubscribe from this group and stop receiving emails from it, send an email to crc-mmbtools+unsubscribe@googlegroups.com.

[Ken]

Hi,

Brendan, do you say that the open encoder sound better that CRC's acc encoder?  I think the sound is metallic even at higher bit-rates and with a good audio source. Unfortunately there is no stations broadcasting in DAB+ over here, so I can not compare to a "real" encoder.

Also I can't get bitrates higher than 144 to work, I try with ex 160k my receiver shows the channel and info but I here no sound, have tested several different receivers.

/Ken

To unsubscribe from this group and stop receiving emails from it, send an email to crc-mmbtools...@googlegroups.com.

[Jens Michael]

Hi!

I've been comparing the old toolame MP2 encoder and the current fdk-aac-dabplus encoder for some weeks now and I do completely agree with Ken's evaluation.

If you want to achieve the best sound possible you should stick with 192 kbits/s MP2 because 144 kbits/s AAC sound inferior, whereas professional AAC encoders sound really good down to 120 kbits/s (using HE-AAC even at lower bit rates).

It's a little disappointing because the quality of the fdk-aac can't be compared with other AAC encoders of Fraunhofer, for example the one used by media player software Winamp.

However, I'm convinced that sooner or later even the freeware version will sound better...

Regards

Jens Michael

 On Sunday, December 1, 2013 5:24:10 PM UTC+1, Ken wrote:

Hi,

Brendan, do you say that the open encoder sound better that CRC's acc encoder?  I think the sound is metallic even at higher bit-rates and with a good audio source. Unfortunately there is no stations broadcasting in DAB+ over here, so I can not compare to a "real" encoder.

Also I can't get bitrates higher than 144 to work, I try with ex 160k my receiver shows the channel and info but I here no sound, have tested several different receivers.

Ken

[Brendan Kehoe]

Hey,

Correct, the open source encoder is a lot better. 

Neither are comparable to a 128k+ MP2 stream though which is unfortunate. Albeit even on my car radio which has DAB+ built in, listening to commercial encoders with DAB and DAB+ simulcasts, the DAB+ version sounds 'tinnier' than the DAB version comparing 128/48k. 

The DAB+ encoder allows me to neatly fit in some extra stations on our mux which don't necessarily need to be especially high quality so I haven't tried it at the extremely high bitrates that you've tried...I didn't even think anyone would use DAB+ at such a high bitrate to be honest, 80k is the highest I ever used and think that 48k is quite satisfactory for mainly talk led radio stations. Commercial DAB+ codecs are supposed to give equivalent quality to a 128k MP2 at a 48k bitrate for example. 

Brendan. 

[Ken]

Hi!

Is the codec HE-AAC v1 or v2 ?

I've tried all bitreates and can't even get close to the sound on our webstream even when comparing to a 48k. Have a listen at: http://countryrocksradio.com/  Audio processing is done by a Omnia 9/XE

/Ken

[Kofi Obiri-Yeboah]

Hullo Rash,

Thanks for taking some time to reply and suggest possible paths to follow. I am not very sure of the etiquette here so I will refrain from attaching the  diagram for  the concept that I have in mind.

for now. Will instead email you a copy directly. After you've had a chance of seeing it and if you find it appropriate I will load it then

 I "mispoke" when i said we wanted to duplicate  your work. What i meant was to review what you had done block by block basis so as to know what I could use in my new system design

Since the CRC has their DAB solution in modules, I am interested in finding out in the least possible time which modules could be adapted.

This project is targeted at Africa, specifically Ghana and Nigeria

In this project, I am going to be multiplexing  a lot of audio channels (20 -40) - they are broadcasts of the same topic but in different dialects -being sent out concurrently. Hence at the outset, is the CRC DAB multiplexer going to be more efficient than using any of the available Audio over IP multiplexers? I do have access to satellite. I have also used the Intel RCP WiFi platform as both a back bone and direct client access network before in a previous project

Since the CRC DAB solution is modular, I am trying to understand how each module works and compare using, if practical, some modules from that package. 

Ultimate delivery is to tabletop radios, vehicles, PCs - Windows, Mac OS and Linux, - mobile devices - IOS, Android, and Blackberry

Since I intend to use Satellite and the Intel RCP platforms as part of my solution, do  I need the CRC modulators? I intend to beam up and distribute using satellite and possibly use land based repeaters

I am aware that I am going to need custom radios. I am already in the process of talking to some Chinese manufacturers who are already building "Internet radios" and  I intend to do some deals to have some of their radio models modified to work with the network. I am finding years of being around aircraft radios being extremely helpful here (lol)

Look at this therefore as some kind of replacement for DAB/DAB+ for the poor folks. I cant help but always think outside the box when it comes to telecoms

Any suggestions / criticism  re my cannibalization of standard DAB is really welcome. 

Cheers

Kofi

--
You received this message because you are subscribed to a topic in the Google Groups "mmbtools" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/crc-mmbtools/HY2BBgEYGao/unsubscribe.
To unsubscribe from this group and all its topics, send an email to crc-mmbtools...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
Worrying Is Like Paying Interest On A Loan You Are Yet To Take
Don't Worry, Be Happy!

[Rash]

Hi Kofi,

Apologies for the late reply - it is a very busy time for us in the UK.

Your project looks interesting, but I don't feel very well placed to comment on it - looks to me like something rather different to DAB, so I doubt dabmod will help you much!

Best regards,

Rash.

Sent from my iPad