Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Meridian Audio's new... sub-format called MQA. (Read 146873 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

Meridian Audio's new... sub-format called MQA.

Reply #175 –
Quote from: Porcus on
Do they really only deliver 16 bits including whatever extra information?

I think they're more flexible than that. The question is what their customers want, i.e. the content providers.

Meridian Audio's new... sub-format called MQA.

Reply #176 –
Quote from: greynol on
Anyway, I was answering the second portion of your post

Oh I see.
Well, I see no reason why someone shouldn't be able to use any filter they like to resample.

--

I don't think they only use 16 bits because the HF needs a couple of bits, so you would audibly degrade the freaking main audible content. It could only work with extremely loud and compressed music ... which kinda doesn't fit the purpose of the format.

So again, you cannot put this on a CD. I guess what they meant to say is that a limited receiver that can only operate at 44.1 or 48 kHz would probably accept 20 or 24 bit words but just throw away the lower bits. That's all the compatibility I see.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #177 –
Quote from: xnor on
I guess what they meant to say is that a limited receiver that can only operate at 44.1 or 48 kHz would probably accept 20 or 24 bit words but just throw away the lower bits. That's all the compatibility I


need.

Meridian Audio's new... sub-format called MQA.

Reply #178 –
No, it's not needed. And it's actually inefficient.

Would you rather transfer 24 bits to any consumer which will potentially throw away 1/3rd of that transferred information, or let a few audiophiles opt-in to receive the extra bits or higher sampling rate which most likely have hardware that can handle it?
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #179 –
Quote from: xnor on
No, it's not needed. And it's actually inefficient.

Would you rather transfer 24 bits to any consumer which will potentially throw away 1/3rd of that transferred information, or let a few audiophiles opt-in to receive the extra bits or higher sampling rate which most likely have hardware that can handle it?



That is S/PDIF? Wordlength: 32 bits. Header/subchannel etc.: 8 bits. Of the remaining 24 bits, 1/3rd is thrown away (4 through 12 are always zero with 16-bit audio).

Meridian Audio's new... sub-format called MQA.

Reply #180 –
Quote from: Porcus on
That is S/PDIF?

S/PDIF is a consumer digital audio interconnect. It has nothing to do with streaming audio over the Internet which costs bandwidth.

And show me an MQA ready DAC that uses S/PDIF and is limited to less than 192 kHz.
Because instead of being able to use your current DAC (192 kHz compatible ones have been produced for years) which may or may not use S/PDIF, you have to buy a new MQA DAC anyway! Otherwise you are just playing 16/44.1, throwing away a lot of information streamed over the Internet...

So it makes no sense.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #181 –
Quote from: xnor on
Because instead of being able to use your current DAC (192 kHz compatible ones have been produced for years) which may or may not use S/PDIF, you have to buy a new MQA DAC anyway! Otherwise you are just playing 16/44.1, throwing away a lot of information streamed over the Internet...


Do you know that the streaming part of it will transfer the full signal if there is no MQA-aware DAC identified? (I don't.)

Source?

Meridian Audio's new... sub-format called MQA.

Reply #182 –
No but if that were the case then it would make even less sense.

If you do it the non-dumb way and stream only the format compatible with the user's hardware (like I've explained above), then why would you need to package "hi-res" with probably some loss into a proprietary format with low sampling rate that costs license fees?
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #184 –
Well yes, it isn't lossless data compression by definition.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #185 –
Quote from: xnor on
Well yes, it isn't lossless data compression by definition.


I agree.  This is the key problem.  And what are the actual numbers for it's equivalent bitrate?  We know that Redbook Audio (44.1 kHz 16-bit linear stereo PCM) is 1411 kBps. 
So what kind of actual numbers does MQA actually have?  And I mean of AUDIO data, not meta-data?  They can't have it both ways.  All the bits are necessary from a hi-fidelity lossless
perspective.  You can't just suddenly re-appropriate the lower bits to some other purpose and still consider it lossless.  Either you are representing the waveform with those bits or you
aren't.  And if if they are represented by something else, how accurate is that? 

The format thus far brings more questions than answers.  And that's not transparency in a codec; it's convoluted or it's kind of a technical hoax. 
That's my educated guess anyways.  If it's perceptually encoded, it's lossy by definition.  They can't have it both ways.
Be a false negative of yourself!

Meridian Audio's new... sub-format called MQA.

Reply #187 –
If you look at the patent and just the first 16 bits for 96 kHz input the output has 13 bits LF + 3 bits lossy HF.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #188 –
Quote from: xnor on
If you look at the patent and just the first 16 bits for 96 kHz input the output has 13 bits LF + 3 bits lossy HF.

I can't remember, but I bet it'll have noise shaping too, which at 96k can get the noise floor below the equivalent of 44.1/16-bit flat or even noise shaped, even with "just" 13-bits.

Cheers,
David.

Meridian Audio's new... sub-format called MQA.

Reply #189 –
Btw, they say in the patent: "We can conclude that a 16-bit 96 kHz channel with appropriate noise shaping is entirely adequate as a distribution format, meeting audiophile requirements with some margin to spare."

FLAC compressed that would be roughly 1.2 Mbps. A 16/96 stream "compressed/packed" into a 24/48 PCM would be 2.3 Mbps.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #190 –
But how much would a MQA 24/48 PCM stream compress using FLAC (or another other lossless audio codec)?

EDIT: I mean, if a MQA "24/48" stream compressed to smaller than a real 16/96 stream, it would be of some value.

Meridian Audio's new... sub-format called MQA.

Reply #191 –
I have a few 24/48 FLACs and they are roughly 1.5 Mbps. This should get worse if the lower bits are randomized due to other compression.

"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #192 –
Quote from: xnor on
If you look at the patent and just the first 16 bits for 96 kHz input the output has 13 bits LF + 3 bits lossy HF.


Oh, thanks!

Then recalculating to lossless Red Book equivalent to compare to my collection, they need either a pass-through (that kills the HF completely), or compress down to 1146-ish kb/s. Of my rock/metal collection encoded to FLAC -8, the latter leads to a loss in 1 out of 300 tracks, but when streaming variable bitrate with less than a full track of buffering, one should exceed 1146 more often than that.
 

Meridian Audio's new... sub-format called MQA.

Reply #193 –
Quote from: xnor on
I have a few 24/48 FLACs and they are roughly 1.5 Mbps. This should get worse if the lower bits are randomized due to other compression.

When are the lower bits of 24 bit material ever NOT random?

Meridian Audio's new... sub-format called MQA.

Reply #194 –
Quote from: xnor on
If you look at the patent and just the first 16 bits for 96 kHz input the output has 13 bits LF + 3 bits lossy HF.


Yes, and the patent tells all. IMO a considerably different story than the one that might be discerned from the high end audio press:

Google Patents Link

"
An encoder for digital audio signals at a higher sample rate creates a stream for consumer distribution at a lower sampling rate, with compatibility for standard PCM players without a decoder. In conjunction with a suitable decoder, two enhanced playback options are supported, the first option allowing full lossless reconstruction of a noise-shaped higher sampling rate signal, the second option allowing lossy bandwidth extension even if an intervening transmission chain has truncated the least-significant-bits of the encoder's output signal.
"

Link To MQA Encoding Graphic

Meridian Audio's new... sub-format called MQA.

Reply #195 –
Quote from: pdq on
When are the lower bits of 24 bit material ever NOT random?


There are varying degrees of randomness. The more lower bits you scramble, the higher the entropy, the lower the later compression ratio.
"I hear it when I see it."

Meridian Audio's new... sub-format called MQA.

Reply #196 –
Quote from: pdq on
Quote from: xnor on
I have a few 24/48 FLACs and they are roughly 1.5 Mbps. This should get worse if the lower bits are randomized due to other compression.

When are the lower bits of 24 bit material ever NOT random?


Depends on the program material. Also depends on which lower bits that we are talking about.

Take a 24 bit file with power supply or deterministic HVAC noise that is 96 dB or more down, for example.  If it is HVAC noise, it is usually less than 96 dB down.  Eithe way, not uncommon in my investigations.

Meridian Audio's new... sub-format called MQA.

Reply #197 –
Quote from: xnor on
If you look at the patent and just the first 16 bits for 96 kHz input the output has 13 bits LF + 3 bits lossy HF.

Thanks for replying with some science instead of just telling me "it's lossless" over and over again.  Seriously, thanks.  I will maybe take a look at the patent later too.  Good idea.

About 24-bit audio... not all audio is recorded from microphones or through consoles.  There's a heck of a lot of electronic music that is generated and recorded entirely inside of computer software.  Every bit of every byte of that is ostensibly audio.  There are no "free" bits at the LSB end if the end result is rendered to a high-resolution audio file.  The audio is really pristine for some genres such as ambient. 

Personally, i wouldn't want just 13 bits for the bulk (lower end) of my audio signals.  That could start sounding like bitcrusher effects, which I just don't like the sound of. 
They need to be upfront about this format... MQA is LOSSY.  If they can't be honest and upfront about that, what other aspects of their business are they hiding?
Be a false negative of yourself!

Meridian Audio's new... sub-format called MQA.

Reply #198 –
Quote from: Nystagmus on
Personally, i wouldn't want just 13 bits for the bulk (lower end) of my audio signals.  That could start sounding like bitcrusher effects, which I just don't like the sound of.
13-bits, dithered and noise shaped, does not sound like intentionally lo-fi audio. The noise floor goes up a bit.

I wouldn't assume that the product is the same as the patent. I wouldn't assume that most/much MQA will be distributed as 44.1/16 anyway, hence the worry about 13-bits may be misplaced.

Cheers,
David.

Meridian Audio's new... sub-format called MQA.

Reply #199 –
In the patent they seem to compress 96/24 into 48/24, but their real product is supposed to compress 192/24 to the same output (?) ... so I don't see how it would be better. If anything, it is more "lossy".

And raising the noise floor is lossy by definition. Else lossyWAV could be renamed to losslessWAV.
"I hear it when I see it."