RNZI - DRM vs AM audio bandwidth

I've done this comparison before but here is a quick double-snapshot of RNZI in DRM and in AM.

This was done with the "Music 101" program host Emma Smith IDing the show, first in DRM on 17675 kHz and then in AM on 15720 kHz. You can see that DRM gains you 5 kHz of audio bandwidth. At the audio bitrate of 15.5 kbps using AAC+, RNZI sounds great - especially when you consider that it is reaching me from nearly 7000 miles away.




No argument, RNZI in AM does perfectly well here, but it is clear that DRM delivers the goods. Also note the level of audio compression on the AM signal. In spreading out the DRM signal, we also gain delightful dynamic range, which is what the recent CD loudness wars have been all about. Imagine how much better xHE-AAC would be!

More on Opus codec and my dreaming

This guy seems to "get it" after having spent some time investigating Opus for use in imbedded systems. I continue to see the practicality of implementing these new codecs (such as Opus and codec2)
in DRM and Digital Voice applications for ham radio and other services.

What excites me most, is the ability to multicast via HF DRM. These codecs make it more than possible to offer multiple voice channels with decent audio quality and bandwidth.

Somewhat off topic but...How about we propose a North American DRM service from the new and improved WRMI? What would be ideal is a DRM - AM simulcast transmission with 2 to 4 audio channels available in DRM.

WRMI is looking to lease more airtime. This would be a great way to do it! Their website says they can offer time for as little as "$1 a minute"! Do we know if WRMI is DRM capable now? Kickstarter anyone?

(Thanks for the Opus article link Benn!)

AIR produces Tender for 800 trial receivers

Charlotte from the Consortium has informed us that All India Radio (AIR) just produced a tender for 800 receivers. You can read it here.

The set they are looking for is fairly basic, but adds a few newer features including xHE-AAC compatibility and the DRM Warning System protocols.

I'd love to get my hands on one of these sets when they hit the real world. Stay-tuned to see if I do!

New implementation guide and xHE-AAC codec added

The Consortium has released a new implementation guide available here in PDF.

In addition to the updated guide, the DRM standard has gotten even better with the addition of  xHE-AAC compatibility from Fraunhofer. E(x)tended HE-AAC "improves significantly the audio quality of music and speech particularly at very low bit rates of 8 kbit/s and more, and is compatible with HE-AAC streams." Don't just take their word for it, listen to the audio on this short slide presentation.

From what I hear, this is an amazing improvement and may help robustness through stations being able to reduce audio data payload. I'd love to see someone do a side-by-side comparison with AAC+ at similar bitrates. The 12k example would appear to rival what we now hear via AAC+ at about 20k!

Schaa says "Opus codec for DRM"

Update 20130512: Rafael - our man about town in Brasil - indicates that the newest version of DReaM software supports Opus codec! Now for the addition of Codec2 and some hardcore experimenting! I'd love to see one of our favorite International broadcasters and super-power outlets experiment with DX and multiplex using these new, highly effective/efficient codecs.

Update 20121024: Read this Opus (CELT) proposal for DRM by Michael Feilen. Michael also added some clarification on rights and pricing of AAC+ via the DRMNA Yahoo Group. Read his post here.

Our friend Stephan Schaa (of HCJB Global Germany and Pappradio fame) wrote to the DRMNA Yahoo Group this morning..." 

Adding Opus as a free codec to the DRM System would give many benefits: At the moment you have to pay royalties for the use of the "better" codecs AAC+ for example. This has to be done if you produce receivers (for less than 1000 it's free, maybe this is the reason that the number of units for DRM Receivers are so small everytime), but the "bigger money" is made on the encoder side: about $4000 (or more) have to be paid to dolby if you want to use AAC+. If - for example - Brasil would opt to the DRM System and insists on a opus as free codec, I think this would be a good idea..." 

Relay of National news broadcast using Codec2

"Relay of National news broadcast using FreeDV in SE Australia next Sunday"  
Read the article here.

A ham software application has been developed that implements the new super-efficient Codec2 audio codec. The software is called FreeDV and is freely available with the open source codec. Speech is compressed down to 1400 bit/s then modulated onto a 1100 Hz wide QPSK signal which occupies half the space of a conventional analogue SSB transmission.

We've covered alternative codecs in previous DRMNA posts. What this means is that not only can digital voice be sent and received using off the shelf equipment, but if this were implemented into DRM, a multiplexed signal with with even more voice channels would be possible. The improved intelligibility, higher compression and narrower bandwidth betters all previous voice codecs by a significant amount.

Here is an opportunity to push-the-envelope and make an even more robust and higher-utility implementation of DRM. Will it happen? We shall see. It does follow the "open source" rubric we always hear from the DRM Consortium.

What I Would Like To Hear (on digital shortwave radio)

DRM is a versatile technology, loaded with features. As many traditional shortwave outlets sunset, We need new broadcasters to take up the reins and try some creative initiatives. I know, no receivers! Well, we are getting closer to eliminating that excuse with the Newstar DR111.

Here are a few of my crazy ideas:

Codec2 speech codec

Codec2 specs and samples here. What is remarkable is that it will work down to 1400 bit rate! Check out this YouTube video. Imagine a fairly robust DRM setting being used and then 4 or more voice channels with weather, traffic, news, what have you. WOW! Open source and ham radio friendly too! I'd love to see more development and implementation of this. The image from the video explains that 1 hour of voice fits on a 1.4M floppy!