On Now : Breakfast
Saturday - Martin Handley
BBC Radio 3
BBC Radio 3

Rupert Brun, Head of Technology for Radio, invites you to participate in a surround sound experiment with BBC Radio 3 and BBC R&D.

Radio 3 Audio Innovation

Radio 3 has a long history of pushing the boundaries with new technology, from early stereo broadcasts in 1958 through wide dynamic range high bit rate audio (HD Sound) to binaural presentation (surround sound in headphones). In collaboration with BBC R&D, we have a new experiment for you to try. For two weeks from 15th to 31st March some of our evening concerts from London’s Southbank Centre will be available in surround sound. This is an experiment so it may not work for you, if you just want to listen to the concerts in the normal way, you can enjoy them here.

You can read more about the experiment and find links to the R&D Blog and FAQ further down the page, but if you just want to try the experiment you will find the player here but please remember the experiment is only running during our evening concerts, details can be found here.

What’s new?

There is a new standard for HTML, the language that operates the World Wide Web, called HTML5. It includes an Audio API (application programming interface) which for the first time allows a web browser to play surround sound without the need to download and install additional software. It also includes the Mediasource API, which enables us to use MPEG-DASH, a new standard for media delivery, to get the data to you. Not all web browsers fully support HTML5 (particularly the Mediasource API) yet; we have tested the experiment using Chrome under Windows 7 and Mac OSX, but it might work with other combinations of computer and browser. Although an HTML5 browser can easily play a surround sound file encoded with AAC, we think we are first to get a browser to play a live audio stream without plug-ins but would love to hear from anyone else who has done it.

You can read more about the technical aspects of this experiment on the R&D blog.

What do you need?

You will need a computer with an HTML5 compatible browser (such as the latest version of Chrome on Windows or Mac, or IE11 under Windows 8.1) and a means to replay surround sound. You could use a multi-channel sound card connected to the computer by USB or you may have luck with an HDMI connection from your computer to a home cinema setup. If you set up something specially to enjoy the experiment you don’t need to worry about the centre or subwoofer channels as we won’t send any audio to them, but some systems with small speakers put all the bass through the subwoofer in which case you will still need it connected.

What will you hear?

You will hear the concerts in 4.0 rather than the 5.1 presentation usually used for feature films. There won’t be any helicopters or explosions so we don’t need the '0.1' LFE channel. We won’t be using the centre channel because it's easier for us to create a surround sound balance without compromising the main stereo programme if we only use two speakers at the front. Not all concerts will be available in surround sound, when the experiment isn’t running you will hear a test announcement. If you are listening on a computer surround sound system you are probably nearer the front speakers than the rear ones; in a home cinema you are probably nearer the rear speakers than the front ones, so you may need to adjust the front to back balance to get a good sound. The rear speakers will carry reverberation during the performance, which should give you a more realistic sense of location within the venue than you get with stereo. You should find applause comes from all around you, helping you feel part of the audience.

During the interval, you will only hear normal stereo.

Philip Burwell describes the experiment from the sound balancer’s perspective

There have been many experiments in mixing surround sound over the past 40 years or so. Generally, they involve either rigging special microphone arrays, or processing the sound electronically to mimic the effect of surround. On a Super Audio CD, there will often be two separate mixes embedded into the disc, one stereo and one surround. Many TV broadcasters have to broadcast live on several platforms at the same time, so down-mixers, or ‘black boxes’ may be used to convert surround to stereo without the intervention of the sound balancer.  In our experiment we are using neither of these approaches. For reasons of efficiency, we are using just one mixing desk to provide both the stereo and 4.0 mixes, with some extra microphones rigged in the hall. Our approach, to put it very simply, is to enhance the live stereo mix with some hall ambience in the rear loudspeakers. We hope you enjoy the experience.

How can we give feedback?

You can give feedback by commenting to this blog, or on Twitter using hashtag #BBCR3surround.

What if it doesn’t work?

As this is very experimental I’m afraid we can’t give you individual help getting it working; if you can’t make it work or just don’t like it, please let us know, and you can always listen using our high quality HD Sound internet stream from the Radio 3 homepage.

The surround sound player is here and BBC R&D have created a list of Frequently Asked Questions (FAQ).

Tagged with:


This entry is now closed for comments.

  • Comment number 48. Posted by Chris Poole

    on 6 Apr 2014 18:21

    I'd like to provide some follow up regarding my impressions of the sound quality of the broadcasts. In general, I thought that the quality of the last several broadcasts was excellent. I only experienced a few momentary dropouts. I noted (in a previous comment) a channel imbalance with the rear channels on the broadcast of March 24th, but I did not observe that condition on any of the subsequent broadcasts. Also, I'm probably in danger of losing my "audiophile" bona fides in saying this, but I couldn't reliably detect any significant artifacts from data compression, which was surprising for an internet broadcast. When listening from a central position with my surround setup, I was provided with a good impression of sitting in a concert hall (I can't claim that it was accurate since I've never visited the hall in question, but it seemed reasonable to assume so). The broadcasts also brought to my mind another question: Can you provide any details about the types of microphone patterns and arrangements that were used for the various broadcasts? It's a curiosity of mine and I would be interested to learn what you had found to work well in your circumstances. Thanks again for running the experiment.

    • This entry is now closed for comments. Number of positive ratings for comment 48: 0
    • This entry is now closed for comments. Number of negative ratings for comment 48: 0
  • Comment number 47. Posted by Rupert Brun

    on 6 Apr 2014 16:44

    HI Johnbclemence. I'm afraid we don't have bandwidth on the broadcast platforms for radio in surround sound. The reason we are experimenting with this format rather than those which are already widely adopted is that it runs entirely within the web browser without plug-ins or proprietary software. When the web browsers in TVs and set-top boxes fully support HTML5, they will be able to decode and play the surround stream from the internet, so you won't need additional devices and it will be easy for people to use.

    Rupert Brun
    Head of Technology for BBC Radio

  • Comment number 46. Posted by johnbclemence

    on 1 Apr 2014 12:32

    Having read the material regarding surround sound on Radio 3 I feel bound to ask why you aren't developing the existing proven methods for delivering this sound format, namely cable, freeview, freesat or satellite. I've enjoyed the proms with full surround sound on freeview and cable so why not on the radio 3 channel on those platforms as well. Also, as I already have the AIR connection between my computers and my AV system is there not some mileage in developing that or something similar? Additionally I'm using an Internet Receiver to listen to Radio 3 in high definition sound and, as that is fed digitally to the AV Amp, was wondering why that can't be further encoded to handle Dolby 5.1 surround sound? In short, why yet another system and another connection required between a computer and the AV system? After all, they are not always within close or convenient positions with regard to each other. Not that I want to put you off developing new ideas - just wondered why proven and effective technology can't be developed?

    • This entry is now closed for comments. Number of positive ratings for comment 46: 0
    • This entry is now closed for comments. Number of negative ratings for comment 46: 0
  • Comment number 45. Posted by Rupert Brun

    on 31 Mar 2014 20:46

    Thank you everyone for your feedback, encouragement and patience. I'm sorry some, like Richard Grubb, were unable to get it to work and I've no idea why you kept getting the error message I'm afraid. I will investigate the differences in distribution between the test player and the live concerts. The experiment is now finished, I will review all the comments I have received and the lessons we have learned. I would like to finish by thanking the BBC R&D engineers and the BBC Radio 3 recording engineers, without whom there would have been nothing to listen to!

    Rupert Brun
    Head of Technology, BBC Radio.

  • Comment number 44. Posted by Michael Taylor

    on 31 Mar 2014 19:25

    Only heard about the experiment on Sunday [30 March] afternoon - however managed to receive the Monday evening concert successfully - via Windows + Chrome + a very basic soundcard!

    • This entry is now closed for comments. Number of positive ratings for comment 44: 0
    • This entry is now closed for comments. Number of negative ratings for comment 44: 0
  • Comment number 43. Posted by Richard Grubb

    on 28 Mar 2014 22:09

    Just to finish the story, I did not have the luck here in Colorado that Chris had in Arkansas. I continued to check intermittently at the later concerts but still got the "Stream has Terminated, Reload" message whatever I tried, including turning off my virus checker. I was able to run the little test program under the FAQ link successfully at all times. I am still very curious why that should be when the main player would not function. Perhaps you could comment Rupert. I tried to check the internet delay from here to the rdmedia server. However it seems to have ICMP turned off.

    It would be great to have surround sound. It really can add quite a bit to concert broadcasts if you are using a full scale music sound system. I remember some early BBC HD TV experimental recordings of the proms that surfaced over here on a cable channel in the very early days of HDTV 5 years or more ago. They had Dolby Surround and the total experience was very impressive. Thanks for the R3 surround experiment. I hope it can progress to implementation as the standards for multichannel Internet sound streaming get established.

    Also thank you very much just for the R3 HD stereo. It really is a beacon for classical music listening over here in the US.

    • This entry is now closed for comments. Number of positive ratings for comment 43: 0
    • This entry is now closed for comments. Number of negative ratings for comment 43: 0
  • Comment number 42. Posted by Rupert Brun

    on 28 Mar 2014 15:17

    There are only two more concerts in surround sound as part of this experiment; Friday 28th March (Mendelssohn, Mahler) and Monday 31st March (Ravel, Debussy, Barber, Janacek, Butterworth, Bartok) both concerts starting at 19:30 UK time. Listeners outside the UK please be aware that the clocks in the UK go forward an hour on Sunday morning, so Friday's concert is at 19:30 GMT whilst Monday's is at 18:30 GMT.

  • Comment number 41. Posted by Rupert Brun

    on 27 Mar 2014 11:09

    Thank you everyone for the useful feedback. Chris Poole - the sample rate is indeed 48kHz. I'm sorry those using Firewire from a Mac are having trouble, this is something we will need to think about. Those using 4 analogue outputs from a USB sound card or HDMI seem to have more luck. I like Peter Suthers suggestion of a pink noise calibration test and will keep it in mind if we undertake more experiments with surround sound.

    Rupert Brun
    Head of Technology, BBC Radio

  • Comment number 40. Posted by Peter Suthers

    on 25 Mar 2014 19:40

    like Chris Pool I also found the Right Rear was significantly lower than the Left rear last night, and like him, was not sure if it was my setup or the broadcast. It would be useful that as well as test file that identifies the speakers, you had one with pink noise stepping around the channels at the same level to allow calibration of amps/speakers thus enabling us to hear the same balance as the guys in the truck.

    • This entry is now closed for comments. Number of positive ratings for comment 40: 0
    • This entry is now closed for comments. Number of negative ratings for comment 40: 0
  • Comment number 39. Posted by Chris Poole

    on 25 Mar 2014 13:05

    I have one question about these broadcasts: What is the native sampling rate of the brodcasts? I suspect that it is 44.1kHz or 48kHz, but I would like to confirm the actual rate so that I can set up the rest of my system so that no sample rate conversion takes place. Thanks!

    • This entry is now closed for comments. Number of positive ratings for comment 39: 0
    • This entry is now closed for comments. Number of negative ratings for comment 39: 0
More comments

More Posts