[zolkis]

On my search for a affordable bluetooth headphone, I've noticed people on Head-Fi peer pressuring others to jump up to a $1,300 bluetooth headphone that is the best headphone ever.
I saw posts that Dali was advertising that their expensive headphone sounds just like an electrostatic headphone and with the lowest distortion in a headphone.
I've attached a video below.

@2:36 - 6:08
The Dali employee here is saying almost all headphone drivers on the market are "cost effective, commoditized, no focus on the driver" they have the solution that no one else has, with the lowest distortion possible.

I felt like Sound Science was the best place to inquire about this.
I sent a message to Gregorio about this topic, and he gave me really good insight of his thoughts! There are some red flags in their marketing material such as not fully publishing evidence that their driver is actually audible in reducing distortion. Greg also mentioned that Dali's marketing is "in this case Eddy Currents and Magnetic Hysteresis but more commonly something like jitter or skin effect and then make some claim about your product being better because it reduces or removes the effect but the consumer can’t verify the claim because it’s a complex effect that only scientists/engineers specializing in those specific areas will fully understand."

I wanted to post in the Sound Science to get more thoughts on this style of Audiophile marketing that utilizes Scientific Concepts.
Is this a way just to squeeze money out of people? Creating a problem and then selling the solution?

https://www.dali-speakers.com/en-us/sound-academy/tech/patented-soft-magnetic-composite-smc/
I've attached a link to Dali's webpage about the patent for the Soft Magnetic Composite driver.

I wanted to add pricepoints for a couple of products.
Dali IO-12 Bluetooth Headphone cost $1,300 USD
Dali Kore Flagship speaker utilizing SMC driver costing $120,000 USD

I definitely am not interested in purchasing a headphone for $1,300 USD.
I am worried that people are being squeezed out of their money due to peer pressure or having to fit into a crowd to have the best headphone.

Replying to the original post... (rarely visiting here, sorry for the late reply).
Here's what I figured out of the Dali IO-12 drivers after a few months of ownership:
- They are far, far better then than drivers in the Dali IO-6 and IO-4: bigger sound stage, more layered, better resolution in the whole FR range, better dynamics. That is likely due to the improved magnetic circuit, so there is ground for Dali's SMC claims.
- However, they are far from being even close to good electrostatic drivers (Stax and others). Dali should clarify the reference there.
- They are close, but ultimately not as good IMHO as the drivers used in the closed back Denon D9200 and D7200, and Fostex TH900.
Here I make a note that the D9200 and TH900 drivers use almost-too-strong (1.5T) magnetic field (close to the theoretical 2T limit), i.e. the drivers work closer to magnetic saturation, which is the likely reason for their problems in the treble rendition, but in turn they offer much better musical resolution. The IO-12 drivers are closest to the rather neutral Denon D7200 drivers, only a bit more shouty without pads and in passive mode. Both don't have any treble issues.

However, in active mode (hifi setting), I find the IO-12 drivers very good, currently the best in wireless headphones (IMHO), and that includes the T+A ST, ML 5909 etc. - YMMV (some or many people still prefer the T+A ST, which is fine and legit, just not for me).
If Denon would make the D7200 an active wireless headphone with or without ANC, I'd buy it immediately.
Until then, I am more than happy with the IO-12.
Even though I have world class e-stat systems which are far better, several classes above, I am still often lazy enough to not switch them on, and just listen wireless via the IO-12... they are that good, at least after some ear pads optimizations to my ears. Since that, the overall experience has been close to the top passive closed back headphones with very good amplification (with some differences here and there). Considering the total price of those systems, IMHO the IO-12 is a screaming bargain for the musical experience and convenience it offers.

 

[KinGensai]

What do we have in the bottom line? Correct me if there is any inaccuracy.
In theory
LDAC 990kbps - better audio quality (lossless audio can be obtained), "lossless high resolution".

For FLAC, ALAC, WAV, use LDAC. Other codecs remove content past 14kHz.

AptX HD, AAC - excellent sound quality, but lossy "high resolution".

For MP3, AAC, OGG, whatever other lossy encoding, use these or LDAC, makes no difference on the codec end. Choose whatever makes you hear your preferred sound.

UAT, SSC are dark horses capable of overtaking the three favorites in the race (LDAC, AptX HD, AAC).

Try them, use if you like.

On practice
Everyone decides for himself whether he hears the difference or not.

Pretty much.

 

[Davesrose]

What do we have in the bottom line? Correct me if there is any inaccuracy.
In theory
LDAC 990kbps - better audio quality (lossless audio can be obtained), "lossless high resolution".
AptX HD, AAC - excellent sound quality, but lossy "high resolution".
UAT, SSC are dark horses capable of overtaking the three favorites in the race (LDAC, AptX HD, AAC).
On practice
Everyone decides for himself whether he hears the difference or not.

I was just interjecting if there’s theoretical advantages with lossy 24/96. Technically, LDAC 990kbps is capable of “lossless” (CD quality),

 

[gregorio]

Lossy wouldn’t be desired for re-encoding or editing vs lossless or uncompressed.

Sure but then we’re talking about BT going to headphones, so there won’t be any further re-encoding or editing.

You side with the RIAA (that does say the audio must also be lossless). Other groups define it as any format that is above 16/44.1(48).

TBH, I don’t side with any of them. Hi-res is just marketing, in practice the resolution within the audible band is the same for sample rates/bit depths greater than 16/44. But if we’re going down the road of technical fidelity, rather than actual audible fidelity, then lossless must be a requirement.

https://addictedtoaudio.com.au/blog...-the-bluetooth-codecs-which-performs-the-best
Some interesting reading. From the measurements taken here, it seems LDAC produces the cleanest signal and most complete reproduction of 24/96 lossless files by not brickwalling at 14kHz to 15kHz.

I notice you repeat this brickwall claim several times but I’m not sure where you got it from? The article itself states regarding SBC: “The response is almost dead flat up to 15kHz, at which point there’s a roll-off so that both are down around one decibel at 17kHz and hit 20kHz at between -4dB and -5dB.” - That’s for 16/44 and then: “the roll-offs are coming in at around 14kHz, and are closer to brick wall. The -3dB point is at 15.6kHz.” - For 24/96, so there’s no brickwall at 14kHz to 15kHz. Good luck hearing -1dB at 17kHz with 16/44 and even -3dB at 15.6kHz is going to be inaudible with music content for almost any adult at a reasonable listening level. Don’t make the classic mistake of comparing HF single/test tone audibility at high levels with actual audibility of HF music content at reasonable listening levels!

One final point, the methodology isn’t completely clear and standard measurements of fidelity comparison cannot be directly compared to audibility with lossy codecs because the differences are mostly according to perceptual models. In other words, relatively huge differences are typically measured with lossy codecs but are inaudible, due to hearing thresholds and most importantly, auditory masking.

Here's what I figured out of the Dali IO-12 drivers after a few months of ownership:
ownership:
- They are far, far better then than drivers ….

Did you really figure out those claims or are they just “impressions” from sighted testing?

G

 

[Davesrose]

Sure but then we’re talking about BT going to headphones, so there won’t be any further re-encoding or editing.

Sure, then that gets further into whether there are any academic reasons for "lossy hi-res"

TBH, I don’t side with any of them. Hi-res is just marketing, in practice the resolution within the audible band is the same for sample rates/bit depths greater than 16/44. But if we’re going down the road of technical fidelity, rather than actual audible fidelity, then lossless must be a requirement.

We've both been having to preface: this is theoretical and above auditory range for audible music reproduction. But it does seem from links that there are some measurements where lossy 24/96 can exceed 16/44.1 (and hence can start being considered "hi-res" even when lossy). I'm only bringing this up as a theoretical basis: as I take your point that these aren't really relevant in practice for stereo music.

Perhaps also as a content creator, you're more disparaging about these new reproduction marketing terms when CD is optimal for music. As someone involved with the visual side of video, I've seen how there's great advantages with lossy on that side. Most people know about being able to fit a higher resolution image with lossy. Now also we have lossy codecs that support high dynamic range. So much of the recent advancements in video codecs and display technologies is having higher contrast ranges (having finer gradients and being able to show more detail in a brighter to dark environment). There's also been a whole lot of confusion when it comes to technical nuance vs marketing. For example, when plasma TVs first came out, they were either EDTV (DVD quality) or HDTV (a native resolution that was probably a bit above 720P). It wouldn't be a few years later that there were "FullHD" TVs that had the native 1920x1080p.

 

[KinGensai]

@gregorio
The codecs are not reproducing the entire range evenly up to 20k or to the nyquist frequency of 44.1k or 96k, and the graphs indicate a brick wall starting at 14k to 15k and sharply decreasing after that, which looks like a brick wall filter to me. I did an ABX with recordings of different codecs and concluded there isn't really any significant difference in SQ (the test file used had a crackle that gave away which was SBC and AptX, so I controlled for that), so I concluded you were right about SBC being practically transparent given high quality settings.

 

[bigshot]

Apple uses AAC for Bluetooth which can get higher than 15kHz, can’t it? I haven’t found that Bluetooth has been a problem since its early years, but that might be because I’m Apple most things.

 

[Davesrose]

Apple uses AAC for Bluetooth which can get higher than 15kHz, can’t it? I haven’t found that Bluetooth has been a problem since its early years, but that might be because I’m Apple most things.

AAC itself can get up to 48kHz (sampling rate) in this scenario

 

[KinGensai]

Apple uses AAC for Bluetooth which can get higher than 15kHz, can’t it? I haven’t found that Bluetooth has been a problem since its early years, but that might be because I’m Apple most things.

According to that blog post, the codecs reproduce evenly up to around 17kHz with 16/44.1 and 15kHz at 24/96 unless it's LDAC. The differences show up in measurements, but in listening it doesn't really matter unless the bitrate drops below 2/3 the max bitrate. From blind testing done last year by kamedo2 on hydrogen.audio, AAC 192k seems to perform very well at being transparent, as does SBC 345k.

 

[bigshot]

I find that artifacting is more of an issue at low data rates than missing upper frequencies are.

 

[Dalmonegrig]

AptX HD, AAC - excellent sound quality, but lossy "high resolution".

I was just interjecting if there’s theoretical advantages with lossy 24/96. Technically, LDAC 990kbps is capable of “lossless” (CD quality),

You are indeed right. There is a contradiction here. Theoretically, the sound quality cannot be excellent if there are losses.
AptX HD, AAC - good sound quality, but with losses, will be more accurate.

LDAC 990kbps - better audio quality (lossless audio can be obtained), "lossless high resolution".

LDAC (990kbps) - is identical to a wired audio or an Audio-CD sound quality, Lossless audio transmission.

 

[bigshot]

Sound quality is how it sounds to your ears. Sound fidelity is how accurate it is. If a codec is audibly transparent, it may not be perfect fidelity, but it can have perfect sound quality.

 

[Dalmonegrig]

If a codec is audibly transparent, it may not be perfect fidelity, but it can have perfect sound quality.

Is this possible (theoretically and practically)?

 

[Dalmonegrig]

I have a question.
What minimum version of Bluetooth that a phone (player) should support, as well as headphones, so that LDAC (16 bits/44,1 kHz/ 990kbps) was able to transmit sound without loss?
Thank you!

 

[bigshot]

Is this possible (theoretically and practically)?

Yes. Controlled listening tests can determine the threshold of transparency. We can reproduce sound with more fidelity than our ears can hear. So if you reach the threshold of transparency, sound quality is the same to human ears, even if the file has higher fidelity.