Today Qualcomm is announcing the new “Snapdragon Sound” branding initiative, essentially an umbrella term that covers the company’s various audio related hardware and software products, promising improved end-to-end interoperability for a better audio experience.

Qualcomm’s initiative is rather vague, but it appears to be to be focused on a certification program that ensures correctly engineered software stacks between a phone’s audio subsystem and the listening device. Qualcomm here particularly looks to focus on wireless audio technologies with greater audio fidelity, most of the technologies surrounding Qualcomm’s proprietary aptX codec and its derivatives. At least one concrete example of an optimised Snapdragon Sound system is Bluetooth audio latency, which would reach down to 89ms in the company’s example.

 

On the hardware side of things, mobile platforms obviously cover Qualcomm’s Aqstic audio codec chips as well as speaker amplifiers, but extends the umbrella out to the company’s Bluetooth audio SoCs which are popular amongst wireless headphone manufacturers.

It’s not clear if the end-to-end optimisations are solely limited to Qualcomm hardware products, or if third-party audio hardware solutions will also gain benefit of the optimised audio stack.

Qualcomm states that the first devices supporting Snapdragon Sound optimisations are expected to be available later this year.

 

POST A COMMENT

43 Comments

View All Comments

  • JoeDuarte - Monday, March 8, 2021 - link

    I don't understand your point on bitrates over 900 kbps. That's a huge bitrate, enough for lossless audio compression. Raw uncompressed CD quality is 1,411 kbps, so FLAC lossless should be between 700 and 900 kbps for most content. Where is the decompression happening? Before the headphones or by the headphones? Reply
  • ZolaIII - Monday, March 8, 2021 - link

    Yup for 44100 Hz 16 bit signal however LDAC on 990 KB is 96KHz 24 bit. Arguably in top implementation with good DAC LDAC tops around 100 dB SINAD (which is more than 96 on 44100/16 signal) but however there are some obvious changes to original signal as it's losy low complexity codec as all of them are. You won't ever need more than 44100 Hz 24 bit (eventually for the compatibility reasons 48000) signal to transport as that corresponding to 150 dB SINAD and best DAC's of today only reach 120~123 dB (best mesured desktop one's). If you like to upsampling for what ever reason any DAC including 1$ one's today can do that to much higher sampling rate (can be useful to drive noise to ultra sonic range or when EQ-ing).
    As always path is encoder - decoder (to PCM) - DAC - analog reciver (hedaphone driver).
    Today there really is no reason to insist on lo complexity codec nor to put money into the pockets of people who sell you a snake oil. Opus should be a next step in BT audio.
    Reply
  • JoeDuarte - Tuesday, March 9, 2021 - link

    Interesting. I read reports that people don't notice 24-bit over 16, and as you said a sampling rate beyond 44,100 Hz doesn't seem to have any benefit. Do you think 24-bit is beneficial?

    Something I've wondered about is why we don't have some sort of metadata augmentation for music tracks that could significantly improve the sound quality of lossily compressed files. The core insight is that almost all of this music is known in advance because it's out there as released music. We're compressing and decompressing it as though we don't know what it is – the encoder is essentially naive to what it is, as is the decoder. But we *do* know what it is. We know it's Lady Gaga's "Born This Way" or whatever. Since we know what it is, it seems like we should be able to analyze it thoroughly, including how MP3, AAC, and Vorbis encoders are treating it. Then distribute some sort of small augmentation file to help the listener's decoder optimize the output.
    Reply
  • mode_13h - Wednesday, March 10, 2021 - link

    > I read reports that people don't notice 24-bit over 16

    The argument I see for > 16-bit is things like digital volume control and Eq. If your 16-bit data is perfectly recorded and mastered, and played back in a good DAC at the right volume level, then I accept that ordinary listeners probably couldn't discern any more resolution.

    > Something I've wondered about is why we don't have some sort of metadata augmentation for music tracks that could significantly improve the sound quality of lossily compressed files.

    Sort of like how MP3 supports volume-normalization, maybe the dynamics of the music could be analyzed to select a more favorable set of compression parameters. That's how I'd do it. If we're talking about bluetooth, it needs to be error-tolerant, however. So, you can't have some piece of metadata that's required to decode the rest of the song and only gets transmitted at the beginning. You could look at the pre-emphasis flag as a simplistic example of that.
    Reply
  • ZolaIII - Wednesday, March 10, 2021 - link

    Well 99% of available material are in 16 bit (true) form. 44100 Hz 16 bit full dynamic range is 96 dB while dynamic range contained in recorded music is usually far less (with some novel examples in complex philharmonic classical peaces and eventually some Dolby not music recordings). Arguably there's benefit in very rare situations (recordings) end even that with great hedaphones with corresponding LPS and for shot analytical listening on such high volume levels. How ever there are benefits of even using 32 bit floating precision in some losy formats such as Opus but that's related how they actually work. I clearly stated how there are some benefits of using higher sampling rates (pushing introduced noise to where you cannot hear it, it's legit method of lovering it in the area where you can). Would be nice for those interested in the topic to read how Opus actually works. Reply
  • Wereweeb - Thursday, March 4, 2021 - link

    Glad to see this kind of initiative. Maybe one day, combined with some novel battery technology, we'll see good wireless earphones. Until then, I'll keep telling people to buy Moondrop's Starfield. Reply
  • ZolaIII - Friday, March 5, 2021 - link

    More snake oil and doungrades incoming. Reply
  • JoeDuarte - Monday, March 8, 2021 - link

    "...but it appears to be to be focused on a certification program..."

    Fix your typos AnandTech. It's incredibly unprofessional for you guys to have typos like this in every article you publish. Hire someone. Use normal word processing software instead of Markdown. Whatever you need to do.
    Reply
  • keremutku - Sunday, March 14, 2021 - link

    Sitemizi ziyaret edebilirsiniz. https://www.viltrin.com/ Reply
  • keremutku - Sunday, March 14, 2021 - link

    Sitemizi ziyaret edebilirsiniz. https://www.viltrin.com/ Reply

Log in

Don't have an account? Sign up now