[bigshot]

http://www.expertreviews.co.uk/tvs-entertainment/media-streamers/1402178/meridian-reveals-mqa-studio-quality-music-streaming
 
This article says that the compression scheme focuses on timing, not frequency response. How is that even possible? You can take a lossy file and sum it back with a lossless file and the timing is precisely the same. How does MP3 possibly alter timing? Is this some kind of scheme for addressing jitter introduced by streaming? If so, it isn't necessary. I stream through my house using AAC and Airports and the jitter is well below the threshold of audibility.

 

[RRod]

  http://www.expertreviews.co.uk/tvs-entertainment/media-streamers/1402178/meridian-reveals-mqa-studio-quality-music-streaming
 
This article says that the compression scheme focuses on timing, not frequency response. How is that even possible? You can take a lossy file and sum it back with a lossless file and the timing is precisely the same. How does MP3 possibly alter timing? Is this some kind of scheme for addressing jitter introduced by streaming? If so, it isn't necessary. I stream through my house using AAC and Airports and the jitter is well below the threshold of audibility.

 
I know a really svelte way of delivering 24/96 content in an audibly lossless way at about 256kbps, but I can't tell you b/c Meridian might nab it from me 

 
 
BTW, why lock this thread previously, mod? I'm sorry but when a graph like this is part of the pitch, the product is fair game for skepticism. 

 

[cer]

That is a genuinely lovely graph. Reel-to-reel and LP both much better quality than DVD-A/SACD. Streaming worse than cassette. Definitely not stupid or wrong or anything.

 

[Hudson]

The convenience plot is also odd, I wonder how they compile that.
 
I had a quick read of the article and filled it under meh, too much ambiguous gumph.
 
The article and the MQA website are pretty confusing, is this a just a new codec or streaming service or both?

 

[pwiles1968]

There is  a lot more info here with several articles - http://www.realhd-audio.com/
 
As far as I can tell is is not a new codec, it can be imbedded in existing codecs such as FLAC, it involves adding additional information embedded in the file, if you have an MAQ decoder you get the extra info, if not it plays as normal.
 
I have been reading a bit, appears to be that the reason there are acoustic benefits with 24/96k or higher files is not necessarily the additional bandwidth it is the ability to get better timing, we can not hear the frequency but we can hear the timing errors even if they are very small, MQA is able to improve the timing with smaller files, approx CD bandwidth that is equivalent to 24/192 or better ?
 
This what I have been able to gather so far.

 

[cjl]

CD already has fantastic timing precision though (much, much better than just the sample time).

 

[bigshot]

It talks about throwing out ultra high frequencies that are down at -70dB. It sounds like they are just going to extend some of the compression scemes of MP4/redbook to high bitrate/high sampling rate files.
 
It seems to me that no one who uses FLAC or high sampling rate files cares a bit about file sizes. They actually want the file size to be as large as possible to maintain the purity of every bit possible. The second they hear about throwing away inaudible frequencies, their OCD will kick in and they will demand that all the inaudible sound be put back in.
 
I'm betting that this will just be a less compact lossy format designed to preserve a little bit more of the inaudible... and ultimately audibly transparent, just like every other format above AAC192.
 
The timing thing may just be some sort of buffering during streaming to reduce jitter... which is inaudible anyway.
 
They also talk about "tailoring the compression to different parts of the distribution chain". That sounds to me like one file format that is converted to different levels of compression on the fly, depending on whether you are playing it on a home stereo system or streaming it to your phone.

 

[Roly1650]

It talks about throwing out ultra high frequencies that are down at -70dB. It sounds like they are just going to extend some of the compression scemes of MP4/redbook to high bitrate/high sampling rate files.

It seems to me that no one who uses FLAC or high sampling rate files cares a bit about file sizes. They actually want the file size to be as large as possible to maintain the purity of every bit possible. The second they hear about throwing away inaudible frequencies, their OCD will kick in and they will demand that all the inaudible sound be put back in.

I'm betting that this will just be a less compact lossy format designed to preserve a little bit more of the inaudible... and ultimately audibly transparent, just like every other format above AAC192.

The timing thing may just be some sort of buffering during streaming to reduce jitter... which is inaudible anyway.

They also talk about "tailoring the compression to different parts of the distribution chain". That sounds to me like one file format that is converted to different levels of compression on the fly, depending on whether you are playing it on a home stereo system or streaming it to your phone.

This article is a bit more informative : http://www.realhd-audio.com/?p=3857

Apparently the packaging is a little like HDCD if you can remember that format, the media contains a "flag" which signals the decoder for the higher resolution content, otherwise it plays back at standard resolution.

The timing issue is only obliquely related to jitter, it's based on new research claiming that human hearing is more sensitive in the time domain than it is in the frequency domain. Conveniently for Meridian 24/96 doesn't cut it, the sample rate is too slow for the time intervals we can apparently resolve, it has to be 24/192, but you knew that was coming, right?

The other thing discussed is that higher resolution doesn't necessarily increase the audible content, it makes the filtering easier, but no mention is made of how they get rid of the negative effects of this in the standard resolution files, once the additional content has been triggered and the whole issue seems like it's a well worn path anyway.

Its interesting that the Meridian graph has gone through at least two third party revisions including one by John Slau of Benchmark.

 

[stv014]

Another critically important factor (and one that I’ve failed to acknowledge in previous posts) is the resolution of human hearing in the time domain. There is new information that accuracy in the time domain is multiple times more important than the frequency domain. I’m not sure where the cutoff is for this level of accuracy but I’ve read and heard that 5 to 10 microseconds is the range. This requires a sampling rate of 192 kHz…according to this new information 96 kHz/24-bits is not sufficient.

It is actually not new research that inter-channel delay of as low as 10 us can be detected under best case conditions. Nor is the sampling theorem, according to which sampling does not limit the time resolution, only the bandwidth.

 

[Roly1650]

It is actually not new research that inter-channel delay of as low as 10 us can be detected under best case conditions. Nor is the sampling theorem, according to which sampling does not limit the time resolution, only the bandwidth.

But, I was only paraphrasing the article and Meridians claims, honestly I was officer.........

 

[stv014]

I knew it was from the article, and my reply was to the claims from the article.

 

[nick_charles]

show me some level matched double blind tests and I'll still ignore it ...

 

[ralphp@optonline]

The 5000 pound elephant in the room is that hardly anyone dares to compare, via a controlled double test, the sound of standard redbook audio and the sound of high resolution audio, particularly when the redbook audio is made by downsampling the high resolution audio. In the article cited by the OP the listeners were played high resolution files but not redbook files, which I content would have sounded equally good. In other words, more useless snake oil.

 

[Greenears]

  The 5000 pound elephant in the room is that hardly anyone dares to compare, via a controlled double test, the sound of standard redbook audio and the sound of high resolution audio, particularly when the redbook audio is made by downsampling the high resolution audio. In the article cited by the OP the listeners were played high resolution files but not redbook files, which I content would have sounded equally good. In other words, more useless snake oil.

I resemble that remark!  I'm running a whole series of ABX tests on 16 vs 24, where 16 is derived from 24 with SoX.  It's easy, and fun.  A few people have joined in. It's on the other thread 24 vs 16.  Pitch in we could use some help.  3 or 4 people have actually run something.  The rest are sniping at us. 4999 pound pachyderm? 

 

[bigshot]

It doesn't matter whether it's redbook. I'd love to see someone discern AAC 256 VBR in a test with "HD" audio!