IMO MQA (My Quest for Annuity) is/was a giant fraud thrust upon gullible audiofools and they got exactly what they deserved - absolutely nothing for their money but a blue light to facilitate their confirmation bias/egos. Now lets start arguing about whether FLAC is lossless or not. I had one nitwit (who is very highly regarded in Sydney audio circles) and who insists on ripping his CD collection to uncompressed FLAC ask me "if you compress it and JRiver expands it for playback, where does it get the zeros from?". I put the phone down!
MQA should have been dismissed before it even took off. The "inventor" is IMO insane and greedy, but had just enough sense to market it at a select/stupid target audience.
MQA should have been dismissed before it even took off. The "inventor" is IMO insane and greedy, but had just enough sense to market it at a select/stupid target audience.
As far as I understand it, MQA is/was a way to create a 24-bit, 44.1 kHz file that contained the same information as the original 24-bit, 192 kHz studio recording contained. They called it "folding" and "musical origami" (both marketing terms, though "folding" sorta smelled like convolution - an actual math operation). Supposedly, they were able to do this without corrupting the 16-bit MSB of data so they could offer CD quality (16-bit, 44.1 kHz) to those without MQA players and offer the full studio quality if you had an MQA-compatible player, all from the same file.
It seems brilliant on the surface, though I'm pretty sure both Shanon and Nyquist would be rotating in their respective graves if they heard of the marketing claims. But hey... What do I know?!
In reality, the original 16-bit data were altered, including with compression and in some cases the addition of spurious tones. The folks at GoldenSound put the MQA marketing claims to the test. Have a look here:
MQA went into receivership (British English for Chapter 11 bankruptcy) a few years back and was snatched up by Lenbrook. What they'll do with it is anybody's guess.
FLAC, on the other hand, is an open-source, lossless audio compression format that is enjoying widespread support.
I might stay on Tidal now that they've dumped MQA. Good riddance!
Tom
It seems brilliant on the surface, though I'm pretty sure both Shanon and Nyquist would be rotating in their respective graves if they heard of the marketing claims. But hey... What do I know?!
In reality, the original 16-bit data were altered, including with compression and in some cases the addition of spurious tones. The folks at GoldenSound put the MQA marketing claims to the test. Have a look here:
MQA went into receivership (British English for Chapter 11 bankruptcy) a few years back and was snatched up by Lenbrook. What they'll do with it is anybody's guess.
FLAC, on the other hand, is an open-source, lossless audio compression format that is enjoying widespread support.
I might stay on Tidal now that they've dumped MQA. Good riddance!
Tom
The most interesting part of that video is the response from MQA that is shown for a second or so, much too short to read it. Pause the video and you can read what GoldenSound did wrong.
Basically the automatic MQA CODEC that was used chooses filters based on the amount of ultrasonics in the programme, as part of the attempt to keep the overall impulse response short. Suddenly change from no ultrasonics to huge ultrasonics in a short part of the recording, like in this video, and the automatic CODEC gets into trouble. You need a CODEC with manual filter settings for that.
It's a pity that the lossy data reduction and the system to keep the overall impulse response short were combined into a package deal.
Basically the automatic MQA CODEC that was used chooses filters based on the amount of ultrasonics in the programme, as part of the attempt to keep the overall impulse response short. Suddenly change from no ultrasonics to huge ultrasonics in a short part of the recording, like in this video, and the automatic CODEC gets into trouble. You need a CODEC with manual filter settings for that.
It's a pity that the lossy data reduction and the system to keep the overall impulse response short were combined into a package deal.
Last edited:
FLAC is a codec. It is lossless. What goes in is what comes out. MQA is not lossless; what comes out is different from what goes in.
Jan
I got the same e-mail from Tidal. Not feeling sorry for the MQA quite the contrary. As I use the Tidal, and then transmit it to amp via Bluetooth, I do have another "loss" along the way, but that's another matter...
I definitely am not mourning the loss. I dropped Tidal years ago specifically b/c of MQA. I was an early(ish) adopter when it became available in the US. I still have a few DACs that will do the 'final' unfolding, and Roon provided support for the first unfold after a while.
Since the Warner Bros. catalog was originally a decent chunk of the 'Master' list, I tried various tracks against my CDs / CD rips. To me... in some cases... the 'Masters' were vastly inferior. I recall a few tracks being almost 'unlistenable'. Again... all subjective and from a guy that definitely doesn't have golden ears, so ... YMMV ... grain of salt etc.
I wondered if they were truly made off of different masters or what was going on.
After asking people far smarter than I including (Scott?) the inventor of some sort of digital watermarking... they seemed to think that it was some sort of interaction between the MQA CODEC and the digital watermarking. It's not MQA's fault that the studio(s) didn't check each track. There's no way a human or two could listen to the 100's of 1000s of tracks after they did the digital butchering. They chuck it into the software, and the software spits out the result. However, when I compared tracks between Qobuz and Tidal; the tracks from Qobuz (that I believe were required to carry the same watermark) were fine. When I played the tracks through Tidal at 16/44.1, I recall them being fine. So, I'm not trying to assign 'blame' to MQA, but it was an unfortunate interaction it seems. I still don't know if it was ever resolved or if more users reported issues.
Either way... from the marketing BS to the cost... good riddance.
Since the Warner Bros. catalog was originally a decent chunk of the 'Master' list, I tried various tracks against my CDs / CD rips. To me... in some cases... the 'Masters' were vastly inferior. I recall a few tracks being almost 'unlistenable'. Again... all subjective and from a guy that definitely doesn't have golden ears, so ... YMMV ... grain of salt etc.
I wondered if they were truly made off of different masters or what was going on.
After asking people far smarter than I including (Scott?) the inventor of some sort of digital watermarking... they seemed to think that it was some sort of interaction between the MQA CODEC and the digital watermarking. It's not MQA's fault that the studio(s) didn't check each track. There's no way a human or two could listen to the 100's of 1000s of tracks after they did the digital butchering. They chuck it into the software, and the software spits out the result. However, when I compared tracks between Qobuz and Tidal; the tracks from Qobuz (that I believe were required to carry the same watermark) were fine. When I played the tracks through Tidal at 16/44.1, I recall them being fine. So, I'm not trying to assign 'blame' to MQA, but it was an unfortunate interaction it seems. I still don't know if it was ever resolved or if more users reported issues.
Either way... from the marketing BS to the cost... good riddance.
As I understand it, BT is a transmission system, not a codec. Just like WiFi is a transmission system.
Both WiFi and BT are lossless as long as the link stays up, as far as I know.
Jan
^ You are correct.
My interpretation of their statement was closer (rephrased) to -
'I'm not aware of any "CD Quality" lossless audio transmission CODEC over the typically available bandwidth of even the most current Bluetooth version'.
My interpretation of their statement was closer (rephrased) to -
'I'm not aware of any "CD Quality" lossless audio transmission CODEC over the typically available bandwidth of even the most current Bluetooth version'.
Last edited:
Bluetooth uses a lossy upper layer for audio transmission. This is needed as Bluetooth has less bandwidth than WiFi.
Last edited:
ChatGPT writes much faster then I do, so I asked her to help.
Bluetooth, as a wireless technology, has limitations when it comes to audio transmission, particularly in maintaining lossless audio quality. Traditional Bluetooth codecs, such as SBC (Subband Coding) and AAC (Advanced Audio Codec), are designed to compress audio data to fit within the bandwidth constraints of Bluetooth, resulting in a loss of audio quality. However, there are some developments and codecs that attempt to achieve higher quality audio transmission over Bluetooth:
1. aptX and aptX HD: Qualcomm's aptX codec offers higher quality audio transmission compared to SBC. aptX HD further improves the quality, supporting 24-bit audio, which is closer to lossless but still involves some compression.
2. LDAC: Developed by Sony, LDAC supports higher bitrates up to 990 kbps and can transmit 24-bit/96kHz audio, offering a near-lossless experience. However, the actual quality depends on the connection stability and the device's implementation of the codec.
3. LHDC (Low Latency and High-Definition Audio Codec): This codec, developed by HWA (Hi-Res Wireless Audio), supports high-resolution audio up to 24-bit/96kHz and aims to provide lossless audio quality.
Despite these advancements, achieving true lossless audio transmission over Bluetooth is still challenging due to the inherent bandwidth and latency limitations of the technology. For truly lossless audio, wired connections or Wi-Fi-based solutions like Apple's AirPlay or Wi-Fi Direct are generally more reliable.
Bluetooth, as a wireless technology, has limitations when it comes to audio transmission, particularly in maintaining lossless audio quality. Traditional Bluetooth codecs, such as SBC (Subband Coding) and AAC (Advanced Audio Codec), are designed to compress audio data to fit within the bandwidth constraints of Bluetooth, resulting in a loss of audio quality. However, there are some developments and codecs that attempt to achieve higher quality audio transmission over Bluetooth:
1. aptX and aptX HD: Qualcomm's aptX codec offers higher quality audio transmission compared to SBC. aptX HD further improves the quality, supporting 24-bit audio, which is closer to lossless but still involves some compression.
2. LDAC: Developed by Sony, LDAC supports higher bitrates up to 990 kbps and can transmit 24-bit/96kHz audio, offering a near-lossless experience. However, the actual quality depends on the connection stability and the device's implementation of the codec.
3. LHDC (Low Latency and High-Definition Audio Codec): This codec, developed by HWA (Hi-Res Wireless Audio), supports high-resolution audio up to 24-bit/96kHz and aims to provide lossless audio quality.
Despite these advancements, achieving true lossless audio transmission over Bluetooth is still challenging due to the inherent bandwidth and latency limitations of the technology. For truly lossless audio, wired connections or Wi-Fi-based solutions like Apple's AirPlay or Wi-Fi Direct are generally more reliable.
I had the same experience. If I wanted the sound of 64 kbit/s MP3 I'd just go to Spotify.
Tom
I've had Tidal in use with Roon for about 4 years and note that MQA wasn't that common a codec over that period of time. I am agnostic on MQA sound quality which with Roon doing the first unfold seemed just fine in most cases. I also have a Topping D90 SE with MQA decoder in my headphone system and on most material did not detect a significant loss in sound quality. Maybe I am just deaf.
Tidal was very upfront about why they are dropping MQA and it is entirely cost driven, the licensing fee is just another cost against their bottom line, there is no such fee for FLAC obviously being open source. The MQA files will be replaced with FLAC files with no guarantee that the replacements will be high res.
Obviously the huge reduction in monthly subscription fees has to be paid for; the previous layoffs, and dropping proprietary (licensed) codecs is a bid to reduce operating costs wrt the substantial reduction in revenues.
Tidal was very upfront about why they are dropping MQA and it is entirely cost driven, the licensing fee is just another cost against their bottom line, there is no such fee for FLAC obviously being open source. The MQA files will be replaced with FLAC files with no guarantee that the replacements will be high res.
Obviously the huge reduction in monthly subscription fees has to be paid for; the previous layoffs, and dropping proprietary (licensed) codecs is a bid to reduce operating costs wrt the substantial reduction in revenues.
- Status
- Not open for further replies.