Friday, September 3rd 2021

Qualcomm Adds Bluetooth Lossless Audio Technology to Snapdragon Sound

Press Release by
AleksandarK
 Discuss (60 Comments)
Qualcomm Technologies International, Ltd. today continued to demonstrate its vision and leadership in the wireless audio space with the introduction of Qualcomm aptX Lossless audio technology to its already extensive audio portfolio. aptX Lossless is a new capability of the proven aptX Adaptive technology and a new feature of Snapdragon Sound Technology that is designed to deliver CD quality 16-bit 44.1kHz lossless audio quality over Bluetooth wireless technology. Qualcomm Technologies has taken a systems level approach and optimized a number of core wireless connectivity and audio technologies, including aptX Adaptive, which work together to auto detect and scale-up and are designed to deliver CD lossless audio when a user is listening to a lossless music file and the RF conditions are suitable.

"At Qualcomm Technologies we're excited about the future of sound, and we're continually looking for ways to help our customers deliver new and exciting listening experiences. Lossless audio means mathematically bit-for-bit exact, with no loss of the audio file and up to now the necessary bit rate to deliver this over Bluetooth has not been available. With many leading music streaming services now offering extensive lossless music libraries, and consumer demand for lossless audio growing, we're pleased to announce this new support for CD lossless audio streaming for Bluetooth earbuds and headsets which we plan to make available to customers later this year," said James Chapman, vice president and general manager, Qualcomm Technologies International, Ltd.
To help deliver CD lossless audio quality reliably over Bluetooth wireless technology, aptX Adaptive works in conjunction with Qualcomm Bluetooth High Speed Link technology to help deliver the required sustainable data throughput. Designed to work seamlessly together, these technologies deliver rates beyond 1Mbit/s yet smoothly scale down to 140kbits/s in congested RF environments to minimize any audio dropouts or glitches for a consistent and reliable listening experience.

"Sound quality is the most critical purchase driver across all audio devices according to our 2021 State of Sound survey, which also shows increasing demand for higher quality streaming audio. Over half of respondents are seeking either lossless or high-resolution audio quality, and a massive 64% saying that lossless audio quality is likely to influence their decision to purchase wireless earbuds," Chapman continued. "Currently lossless audio is only supported on client devices such as phones, PCs and tablets. By supporting lossless audio on next-gen earbuds and headphones, we're providing our customers another way to deliver sound the way the artist intended, as well as a significant opportunity to differentiate and be among the first to develop products with this feature."

aptX Lossless features & specifications:
  • Supports 44.1kHz, 16-bit CD lossless audio quality
  • Designed to scale-up to CD lossless audio based on Bluetooth link quality
  • User can select between CD lossless audio 44.1kHz and 24-bit 96kHz lossy
  • Auto-detects to enable CD lossless audio when the source is lossless audio
  • Mathematically bit-for-bit exact
  • Bit-rate - ~1Mbps
For more information and to learn more about our aptX Lossless audio technology, visit here. The technology is expected to be available in late calendar 2021. For the full Qualcomm State of Sound survey visit here. For a whitepaper explaining more of the technology behind Lossless and High Resolution Audio, visit here.

Related News

Add your own comment

60 Comments on Qualcomm Adds Bluetooth Lossless Audio Technology to Snapdragon Sound

#26
mtcn77
TardianI think I said 5 sigma
The scientific term is "Pearson Correlation" rather than a teabag pun, imo. It is interesting not once it is named correctly within the 'freelance' article.
Don't mind me dissing on poor quality journalistic junk. I hate laymen articles.
Posted on Reply
#27
Tardian
The scientific term is "Pearson Correlation" rather than a teabag pun, imo. It is interesting not once it is named correctly within the 'freelance' article.
Don't mind me dissing on poor quality journalistic junk. I hate laymen articles.
One can't understand time dilation unless one has had the pleasure of sitting an MS101 Introduction to Statistics lecture. Listening to my significant other's monologue makes MS101 seem like a Wayne the Rock Johnston action movie after taking cocaine. Before any of you think you can troll me, I just had the pleasure of 'losing' at Hand, Knee, & Foot (cards) to my significant other and youngest son's team. Losing included having my arse handed to me on a plate and being mocked as disabled. In fact, they eventually ran out of disabled puns and nearly wet themselves laughing. I told you before about infinite black.
Even Billie our dog got involved in the mocking. This is on Australian Fathers' Day. I was just about to post and then my evil sister rang and joined in on the general shaming. o_O


I am fairly sure you guys can complete the job. The printer is whirring and an arrow is being printed to remind me which way is up.
Posted on Reply
#28
minami
In the extreme, 24Hz video is acceptable. But audio with a sampling rate of 24Hz is completely unacceptable. Even if it is 24KHz.
As technology continues to evolve, we may one day be able to discard audio cables. Please come while I'm still alive.
Posted on Reply
#29
Aquinus
Resident Wat-man
Huh, okay. It's probably not low power bluetooth. There simply isn't enough bandwidth to drive a lossless codec, so I don't see battery life being ideal wireless headphones. I guess that's a tradeoff.
AleksandarKdown to 140kbits/s in congested RF
While remaining lossless? Sure. :wtf:
Posted on Reply
#31
Dredi
AquinusWhile remaining lossless? Sure. :wtf:
If you read the release, you’ll know that it is lossless only in good rf conditions. In bad rf conditions it switches to use some completely different (lossy) codec.
AquinusIt's probably not low power bluetooth.
BTLE goes up to 2Mbps. That is well enough for this application. For sure it uses more power than some lower bitrate transfer, but it still is classified as BT low energy.
Posted on Reply
#32
Aquinus
Resident Wat-man
DrediBTLE goes up to 2Mbps. That is well enough for this application. For sure it uses more power than some lower bitrate transfer, but it still is classified as BT low energy.
Over the air under perfect conditions. Even in that case it's more like ~1.3Mbps that's usable. That's not a whole lot of bandwidth for handling lossless audio. Let's consider existing lossless codecs, where there is an upper bound to how much you can compress audio data without losing something from it. Even FLAC and ALAC have similar file sizes for what it's storing and it's pretty consistent with others as well. At 16-bit 44.1Khz you're basically at about 1.4Mbps with FLAC (which is one of the better codecs when it comes to lossless compression,) which puts you right at the upper end of the usable range for BTLE. All in all, I honestly don't see how it can be both lossless and be using low energy at the same time. Normal BT probably could do it as it's usable bandwidth is something a bit over 2Mbps short of going into 802.11 land like the faster BT variants. I think that it's far more realistic for it to be over normal BT and that the stars would have to align for there to be lossless quality over BTLE given what we know about existing lossless codecs.
Posted on Reply
#33
Dredi
AquinusOver the air under perfect conditions. Even in that case it's more like ~1.3Mbps that's usable. That's not a whole lot of bandwidth for handling lossless audio. Let's consider existing lossless codecs, where there is an upper bound to how much you can compress audio data without losing something from it. Even FLAC and ALAC have similar file sizes for what it's storing and it's pretty consistent with others as well. At 16-bit 44.1Khz you're basically at about 1.4Mbps with FLAC (which is one of the better codecs when it comes to lossless compression,) which puts you right at the upper end of the usable range for BTLE. All in all, I honestly don't see how it can be both lossless and be using low energy at the same time. Normal BT probably could do it as it's usable bandwidth is something a bit over 2Mbps short of going into 802.11 land like the faster BT variants. I think that it's far more realistic for it to be over normal BT and that the stars would have to align for there to be lossless quality over BTLE given what we know about existing lossless codecs.
They do state that it is bit perfect, so it probably does work.
Uncompressed CD audio is exactly 1.411Mbps, flac at around 0.9. They state that this one is about 1, well within the btle usable 1.3Mbps limit.
Posted on Reply
#34
Aquinus
Resident Wat-man
DrediThey do state that it is bit perfect, so it probably does work.
Uncompressed CD audio is exactly 1.411Mbps, flac at around 0.9. They state that this one is about 1, well within the btle usable 1.3Mbps limit.
That doesn't sound like BTLE though, particularly if they're supporting up to 96Khz/24-bit. Not to mention this little statement.
AleksandarKaptX Adaptive works in conjunction with Qualcomm Bluetooth High Speed Link technology to help deliver the required sustainable data throughput.
Posted on Reply
#35
Dredi
AquinusThat doesn't sound like BTLE though, particularly if they're supporting up to 96Khz/24-bit.
Whether it _sounds_ like BTLE is a bit moot point. It explicitly supports 2Mbps (1.3Mbps usable) data rate. Non LE BT supports 3Mbps and that is not necessary in order to send the audio codec in question. ”Qualcomm bluetooth high speed link technology” is just some marketing name, and I nor you have any idea what it actually uses. I suppose we are not interested enough to find out either.

96kHz/24bit is compressed using a lossy codec, which you would know if you read the qualcomm release. Only 16 bit 44kHz material is supported for this bit perfect codec, and only in good rf conditions.
Posted on Reply
#36
Prima.Vera
Hardly impressive from a lossless 32 bit, 192KHz , or Dolby Atmos....
Posted on Reply
#37
Tardian
Hardly impressive from a lossless 32 bit, 192KHz , or Dolby Atmos....
I'm torn between:
  • Silk purse ... sow's ear
  • The etymology of the word: public.
I am sure you get the point. :cool:
Posted on Reply
#38
Dredi
Prima.VeraHardly impressive from a lossless 32 bit, 192KHz , or Dolby Atmos....
Almost no music exists that goes beyond the 16bit/44kHz quality. And you simply cannot hear even the full 16bits of dynamic range anyway.

For higher end formats, there needs to be some other use than just consumption for them to make sense. For atmos, that applies as the atmos amps do a shit ton of processing in order to do room corrections etc. For music you wish to hear simply ’as is’, higher bitrates and sampling frequencies just get in the way.

some people claim to hear more than is possible to deliver with CD quality, but I’ve yet to see a proper blind test of that, where the results would be favorable to the claimant.
Posted on Reply
#39
Prima.Vera
DrediAlmost no music exists that goes beyond the 16bit/44kHz quality. And you simply cannot hear even the full 16bits of dynamic range anyway.

For higher end formats, there needs to be some other use than just consumption for them to make sense. For atmos, that applies as the atmos amps do a shit ton of processing in order to do room corrections etc. For music you wish to hear simply ’as is’, higher bitrates and sampling frequencies just get in the way.

some people claim to hear more than is possible to deliver with CD quality, but I’ve yet to see a proper blind test of that, where the results would be favorable to the claimant.
Trust me, there is a clear difference between 16bit/44kHz and 24bit/96kHz. I'm talking from experience. There are a lot of music samples out there.
BUT, unless you have decent, quality speakers, with very high range, you are correct, you won't distinguish a thing.
Posted on Reply
#40
Tardian
Trust me, there is a clear difference between 16bit/44kHz and 24bit/96kHz. I'm talking from experience. There are a lot of music samples out there.
BUT, unless you have decent, quality speakers, with very high range, you are correct, you won't distinguish a thing.
The average person can't hear a difference. I can't play a note and can't sing but have pitch-perfect hearing, and there is a clear difference between 16bit/44kHz and 24bit/96kHz and no I have good but not expensive speakers. B05e is rubbish IME.
Posted on Reply
#41
Dredi
Prima.VeraTrust me, there is a clear difference between 16bit/44kHz and 24bit/96kHz. I'm talking from experience. There are a lot of music samples out there.
BUT, unless you have decent, quality speakers, with very high range, you are correct, you won't distinguish a thing.
TardianThe average person can't hear a difference. I can't play a note and can't sing but have pitch-perfect hearing, and there is a clear difference between 16bit/44kHz and 24bit/96kHz and no I have good but not expensive speakers. B05e is rubbish IME.
Trust me, you can’t. Prove me wrong by posting any well made test where the result is as you claim it to be. You simply stating that you can hear a difference means very little.

In the test, the lower bitrate audio clip needs to be produced directly from the high bitrate clip using best practices for downsampling, and the test needs to be a double blind one. Dac used needs to be run at the same bitrate for both clips, with the lower bitrate clip upsampled using best practices on the source device. Otherwise the DAC may be the piece in the signal chain that produces differing outputs.

In many cases music files released at higher bitrates are also mastered differently to any lower bitrate version of the same piece, which explains most of the people who claim to hear the difference.
Posted on Reply
#42
Tardian
DrediTrust me, you can’t. Prove me wrong by posting any well made test where the result is as you claim it to be. You simply stating that you can hear a difference means very little.

In the test, the lower bitrate audio clip needs to be produced directly from the high bitrate clip using best practices for downsampling, and the test needs to be a double blind one. Dac used needs to be run at the same bitrate for both clips, with the lower bitrate clip upsampled using best practices on the source device. Otherwise the DAC may be the piece in the signal chain that produces differing outputs.

In many cases music files released at higher bitrates are also mastered differently to any lower bitrate version of the same piece, which explains most of the people who claim to hear the difference.
What you state, in theory, makes perfect sense, but in practice, I can notice the difference. There is a wide range in human perception outcomes.

I currently don't have access to the right University Library.

dosits.org/science/measurement/how-is-hearing-measured/

For example, people hear best in the frequency range of speech because it is most important for us as apex predators to hear and understand one another.

My clan is the apex predator of apex predators.

Posted on Reply
#43
Dredi
TardianI can notice the difference
Let’s play a game then, assuming that you mean the difference as in audio quality and not in mastering.

You stating that 16bits is not enough, should mean that you can discern differences in volume smaller than what 16bits can differentiate between, correct? 0.1 dB is a much larger difference in volume than what a single value difference is in 16 bit recordings.
What score do you get from these two tests? :) Both are well within what 16 bit recordings can do, and the real limits are way, way smaller.
www.audiocheck.net/blindtests_level.php?lvl=0.1
www.audiocheck.net/blindtests_dynamic.php?dyna=78

There is also a 16 bit vs 8 bit blind test on the site, but I guess that is way too easy for you:
www.audiocheck.net/blindtests_16vs8bit.php
TardianI currently don't have access to the right University Library.
No-one has. That's the problem when trying to find a study that proves the impossible.
TardianMy clan is the apex predator of apex predators.
Or, just possibly, apex sheep? :P
Posted on Reply
#44
Tardian
DrediLet’s play a game then, assuming that you mean the difference as in audio quality and not in mastering.

You stating that 16bits is not enough, should mean that you can discern differences in volume smaller than what 16bits can differentiate between, correct? 0.1 dB is a much larger difference in volume than what a single value difference is in 16 bit recordings.
What score do you get from these two tests? :) Both are well within what 16 bit recordings can do, and the real limits are way, way smaller.
www.audiocheck.net/blindtests_level.php?lvl=0.1
www.audiocheck.net/blindtests_dynamic.php?dyna=78

There is also a 16 bit vs 8 bit blind test on the site, but I guess that is way too easy for you:
www.audiocheck.net/blindtests_16vs8bit.php



No-one has. That's the problem when trying to find a study that proves the impossible.

Or, just possibly, apex sheep? :p
1:1024 chance of being a smartass?

My maternal grandfather used to say: Don'tI can hear the difference enough for it to be material to me personally. It is a perception matter.
Posted on Reply
#45
Dredi
TardianIt is a perception matter.
*placebo matter, but whatever :)
If you like placebo, that is up to you, but don't push your imaginary listening powers as a fact unless you can back it up.
Posted on Reply
#46
stimpy88
R-T-BWe had wireless lossless in 1980?

That's the advancement. This is the first wireless lossless to my knowledge.
#Facepalm
Posted on Reply
#47
Aquinus
Resident Wat-man
Dredi*placebo matter, but whatever :)
If you like placebo, that is up to you, but don't push your imaginary listening powers as a fact unless you can back it up.
I personally would have liked it if that site provided lossless examples instead of using OGG/vorbis though because while its bit depth might have been reduced before being encoded to OGG/vorbis, the actual format medium is going to have a higher depth than 8 bits, so there is some questions as to what the encoder is doing to the input. Another question that needs to be asked about this is what impact does the reduced depth of each sample make with regard to compression. With a lossy codec, what you very well might experience is reduced quality of the 44.1Khz/16-bit clip is to the point where both clips simply sound equally bad. That's more a function of the lossy codec than anything else. To be honest, I can't hear the difference between the two, but they both also don't sound right in the first place either.

So I decided to open spotify, which is also lossy (but a different codec at a higher bitrate,) and played the same song and it does sound better. Bass and mids sound similar, but the highs are noticeably more crisp off of the version spotify has, and I don't even need my flat response headphones and a dac to hear it either. Once again though, I think that has more to do with the bitrate and codec than it does with the depth of the samples.

So I agree with you in the sense that most people can't hear the difference, particularly between 16-bit and 24-bit or north of 48Khz. The problem is that the test itself is flawed and 8-bit vs 16-bit on a lossless codec is a different thing than on a lossy codec which doesn't natively support 8-bit. I think what you'd find is that a good quality audio sample comparing 8-bit and 16-bit on a lossless codec from a lossless source is going to yield different results than using vorbis. However, I don't think it will be an extreme difference, but I do think it'll be perceivable.
Posted on Reply
#48
Dredi
AquinusI think what you'd find is that a good quality audio sample comparing 8-bit and 16-bit on a lossless codec from a lossless source is going to yield different results than using vorbis. However, I don't think it will be an extreme difference, but I do think it'll be perceivable.
I agree!

8bit vs. 16bit is for sure perceivable with a more dynamic song even with a lossy codec, with lossless further expanding the gap. That comparison is more or less just to prove a point that to imagine going above 16 would provide some benefit is just silly. All it does is lower the noise floor below -96dB and hardly anyone listens at their listening rooms noise floor + 96 dB, because they would lose their hearing pretty damn quickly. With 8 bit the noise floor is very audible, at -48dB, which you can try to hear here: www.audiocheck.net/blindtests_dynamic.php?dyna=48

(Should be pretty easy).
Posted on Reply
#49
Aquinus
Resident Wat-man
DrediI agree!

8bit vs. 16bit is for sure perceivable with a more dynamic song even with a lossy codec, with lossless further expanding the gap. That comparison is more or less just to prove a point that to imagine going above 16 would provide some benefit is just silly. All it does is lower the noise floor below -96dB and hardly anyone listens at their listening rooms noise floor + 96 dB, because they would lose their hearing pretty damn quickly. With 8 bit the noise floor is very audible, at -48dB, which you can try to hear here: www.audiocheck.net/blindtests_dynamic.php?dyna=48

(Should be pretty easy).
I thought that the main reason to use 24-bit or 32-bit floating point was for recording to prevent clipping, same thing with a higher sampling rate; so you can more accurately sample down the audio to commonly used formats (the whole idea that more data is never a bad thing when it comes to conversion later.) I had read something, I'm not sure if it's true, that using higher sampling rates can result in harmonic distortion outside of the audible range, but I'm not sure if I buy that. I do have some 96Khz/24-bit content and the fidelity is fantastic, but in reality, if it were sampled down, I would probably not notice a difference. The real difference is that it's 5.1 (6 channel,) and lossless, which makes for a great sounding track.
Posted on Reply
#50
Dredi
AquinusI thought that the main reason to use 24-bit or 32-bit floating point was for recording to prevent clipping, same thing with a higher sampling rate; so you can more accurately sample down the audio to commonly used formats (the whole idea that more data is never a bad thing when it comes to conversion later.) I had read something, I'm not sure if it's true, that using higher sampling rates can result in harmonic distortion outside of the audible range, but I'm not sure if I buy that. I do have some 96Khz/24-bit content and the fidelity is fantastic, but in reality, if it were sampled down, I would probably not notice a difference. The real difference is that it's 5.1 (6 channel,) and lossless, which makes for a great sounding track.
There is a good point about recording and editing being done at higher bit depths and sampling rates, that have to do with clipping and the ease of high frequency recording. For example if you record at just 16 bits and need to adjust gain digitally when editing the track by 30 dB's, the noise floor is going to be at least -66 dB, which someone is possibly going to hear. If you recorded at 24 bits, your quantization noise is still inaudible at below -100 dB. This means that you can set the gain a bit lower when recording, to prevent clipping, and not have to deal with unnecessary quantization noise should you need to push levels later on even by a lot.

For consumption though (as is the case with the aptX lossless), it does not make sense at all to deliver at higher than CD quality. Surround is of course good if you like it, and lossless a mandatory thing IMO.
Posted on Reply
Add your own comment
Mar 15th, 2025 22:19 EDT change timezone

Latest GPU Drivers

New Forum Posts

Popular Reviews

Controversial News Posts