View Full Version : Meridian Quality Audio Launched
http://www.expertreviews.co.uk/tvs-entertainment/media-streamers/1402178/meridian-reveals-mqa-studio-quality-music-streaming
Er.. that would be Meridian:D
They are doing a lot better on the Wam.
This is what it is, in a nutshell:
That means streaming high quality audio is much easier. MQA is able to stream music at 1Mbps - roughly the same bandwidth required for a CD-quality file but much smaller than a hi-resolution, 24-bit/96KHz audio stream which typically uses 4.6Mbps
So if you don't stream music and/or you understand that having more bits and higher sampling rate than red book CD makes no difference at all to what you hear then it is of no interest. probably the same reason we don't seem to have a thread about Pono.
Hm I'm not sure Bob Stuart is doing his supposed rep much good with this little ditty. It seems way to open to attack from the Nyquist is good enough/the research on timing and the human ear is well dodgy brigade. But hey.
Good post Martin.
I suppose if you do stream music and you want what is potentially a better mastering than the CD issue it may be of interest.
I really don't get it with 'streaming' - what is it anyway? Or, more important, why bother? Back in my day* all you needed was a PC or laptop connected to a DAC in turn connected to your amp. Of course you needed material to play, but that is trivial: HDtracks, Linn, Channel Classics, B&W, not to mention Usenet, Torrent and your own rips, etc etc.
*a few months ago
Gordon Steadman
07-12-2014, 13:01
I really don't get it with 'streaming' - what is it anyway? Or, more important, why bother? Back in my day* all you needed was a PC or laptop connected to a DAC in turn connected to your amp. Of course you needed material to play, but that is trivial: HDtracks, Linn, Channel Classics, B&W, not to mention Usenet, Torrent and your own rips, etc etc.
*a few months ago
Streaming can be useful. My Mac is in the snug at the front of the house but the hi-fi is in a different room. We are going to get a streaming device because it saves us having to have yet another computer attached to the main h-fi.
The cheapest streamer is the UK invented Raspberry with a DAC stuck onto it. I would really like to hear a "high-end" multi hundred (thousand) $$ streamer compared to the humble Raspi.
Sorry slightly :offtopic:
I really don't get it with 'streaming' - what is it anyway? Or, more important, why bother? Back in my day* all you needed was a PC or laptop connected to a DAC in turn connected to your amp. Of course you needed material to play, but that is trivial: HDtracks, Linn, Channel Classics, B&W, not to mention Usenet, Torrent and your own rips, etc etc.
*a few months ago
Well if you could sign up to TIDAL with 25 million tracks at CD quality the tip of your search fingers... that's a pretty good reason for streaming.
Those who think there is some sort of SQ compromise gong on with it I think are wrong. TIDAL is as good as my CD rips.
As Martin says MQA could give you pseudo hi-rez where if you have low internet bandwidth you wouldn't otherwise have access to on a live stream.
But that is it, really. It won't better the high res sources, and high res doesn't give you much in the way of improved sound quality in the audible range over CD anyway.
If you want to know about just how good CD quality really is just play with bit depth and up/down sampling in JRiver. Take a 96KHz 24 bit recording and reduce its bit depth and output sampling rate. I know of no other better way to practically hear through the BS.
Reduce the high-res to CD quality. Then reduce the bit depth etc on CD quality and you'll be amazed how low the bit depth needs to go before you can reliably blindfoldedly tell the difference. There's a dither button there too.
Hi-res can seem deceptive as Martin indicates because they may have received better mastering or have been fiddled with. I can certainly think of a few examples that definitely sound better in "high res".
Hi-res can seem deceptive as Martin indicates because they may have received better mastering or have been fiddled with. I can certainly think of a few examples that definitely sound better in "high res".
I'm still prepared to change my mind on it. First time I heard SACD (friend's system) I was very impressed. Bought some hybrids but then discovered the SACD layer no different sounding from the red book one. Compare to the original CD issue though, and they are noticeably better. Then read the copius sleeve notes about how much trouble they went to for the remaster and ah! light dawns. Got Sinatra Live at the Sands on DVD=A. Oh! prefer the old issue CD mix by a margin. Gave DVD=A away. But if someone wants to demonstrate that hi-rez is worth while I'm all ears providing we are not comparing ducks with geese.
I don't know how computer savvy you are Martin but do with JRiver as I suggested if you can.
The maths applied in dropping bit depth and sampling rate should be pretty much the same as studio conversions I'd guess.
Actually doesn't look like you are set up to do it looking at your kit list. No matter:)
Actually doesn't look like you are set up to do it looking at your kit list. No matter:)
No I'm not. I still consider CD to new-fangled. :) And I was once computer savvy but that was back in 1993. Still, the lack of interest in this topic is interesting in itself...
walpurgis
08-12-2014, 00:02
I still consider CD to new-fangled.
It is isn't it? :scratch: :)
Well if you could sign up to TIDAL with 25 million tracks at CD quality the tip of your search fingers... that's a pretty good reason for streaming.
I already have over half a terabyte of flac files on my PC, and about another half terabyte of music videos. So that is one reason for not streaming. Another reason is that I usually prefer hi-rez. The difference in quality is there. Go for Tea for the Tillerman or Hans Zimmer's soundtrack to The Dark Knight Rises and it will be obvious. They are both 24bit 192 khz and well produced. I put the claims of little or no difference down to two things: one is that if the system is not sufficiently revealing then you cannot hear it. Especially DACs This planet has so many crappy DACs, but all other components too. More frequent is that we are (now, as always) being fobbed off with crap, such as oversampling a CD and selling it as hi-rez. Also, and I think Martin may have alluded to this, a different format such as SACD should probably be specifically mastered for that format. When CDs came out I bought many and was uncomprehending as to why they sounded less dynamic - now I know it was because it was an easier buck to make to just transfer the same thing onto CD with virtually no work.
I already have over half a terabyte of flac files on my PC, and about another half terabyte of music videos. So that is one reason for not streaming. Another reason is that I usually prefer hi-rez. The difference in quality is there. Go for Tea for the Tillerman or Hans Zimmer's soundtrack to The Dark Knight Rises and it will be obvious. They are both 24bit 192 khz and well produced. .
Trouble is there is no reason why 24/192 would give any difference in quality. Do we need to hear anything over 22KHz? no. (There is the hypothesis that although sounds over 20KHZ cannot be heard by any human they do produce sounds lower in the frequency rage that we can hear but there is no evidence for this)
Do we need a dynamic range greater than the 100+ dB that CD already provides? No. So what is 24/192 doing to give extra quality? Was Teas for the Tilllerman re-mastered for the hi-rez release? I think that is where you will find the answer. Put the same re-master on a red book CD it will sound the same as the hi-rez, its what I am saying.
I've done the comparison with the Stomes' SACD hybrid releases (same painstaking re-master in SACD and redbook on one disc) on a number of different systems. Also compared the DVD-A Gaucho 24/192 with the re-mastered red book CD. There is nothing in it that I can hear. Now if someone wants to sit me down with their system that they say can show up the differences and play the same master at red book and hi-rez and if I hear that extra quality then I will say so and accept I am wrong, even though there is no explanation for it.
Trouble is there is no reason why 24/192 would give any difference in quality. Do we need to hear anything over 22KHz? no. (There is the hypothesis that although sounds over 20KHZ cannot be heard by any human they do produce sounds lower in the frequency rage that we can hear but there is no evidence for this)
Do we need a dynamic range greater than the 100+ dB that CD already provides? No. So what is 24/192 doing to give extra quality? Was Teas for the Tilllerman re-mastered for the hi-rez release? I think that is where you will find the answer. Put the same re-master on a red book CD it will sound the same as the hi-rez, its what I am saying.
I've done the comparison with the Stomes' SACD hybrid releases (same painstaking re-master in SACD and redbook on one disc) on a number of different systems. Also compared the DVD-A Gaucho 24/192 with the re-mastered red book CD. There is nothing in it that I can hear. Now if someone wants to sit me down with their system that they say can show up the differences and play the same master at red book and hi-rez and if I hear that extra quality then I will say so and accept I am wrong, even though there is no explanation for it.
Well I am glad you have an open mind about it.
Subjectively I have no doubt that 24/192 is an improvement when it is done right, as I have heard it. Why would 24/192 give a improvement in sound quality? First up is the higher frequency. While those frequencies are not audible in themselves the waveform that creates the sound is made from a base wave plus various overtones. It is the overtones that shape the soundwave. Far more important though is that the higher bitrate affects the sound quality of all frequencies. Take one second of music in, say, the midrange. Having that one second of music defined with four times the amount of samples means that the resultant analogue wave has to be more accurate. A simple analogy is a digital photo print. If your 8 by 5 print is at 300 dpi it will not and cannot be as accurate as 600 dpi - providing the source is 600 dpi.
As for the bit depth: the extra headroom is more than is needed and this is a function that something like 18bit would be pointless. The digital data is held in bytes, so to improve from two bytes (16 bits) one can only realistically go to three bytes (24 bits). However that extra headroom is not entirely wasted. First is that there is the potential (not often used due to poor mastering) of avoiding compression which is the unfortunate result of the 'loudness wars'. Second, and I do not understand this properly, is that apparently the extra headroom enables the noise to be put below the threshold of hearing. Most of my CDs are 'silent' but overall I would say the SACDs are more so.
It is interesting that you could not hear any diff with the DVD-A of Gaucho. I will keep a look out for it. The DVD-A format wont have any DSD 'shenanigans' so should be a very fair test.
Far more important though is that the higher bitrate affects the sound quality of all frequencies. Take one second of music in, say, the midrange. Having that one second of music defined with four times the amount of samples means that the resultant analogue wave has to be more accurate. A simple analogy is a digital photo print. If your 8 by 5 print is at 300 dpi it will not and cannot be as accurate as 600 dpi - providing the source is 600 dpi.
.
Well not really. More bits increases the dynamic range available and higher sampling frequency means you can reproduce higher frequencies but it does not mean that the analogue waveform resulting is 'more accurate', it is in fact no different. There is no real comparison with digital photography:
Sampling rate[edit]When it is necessary to capture audio covering the entire 20–20,000 Hz range of human hearing,[4] such as when recording music or many types of acoustic events, audio waveforms are typically sampled at 44.1 kHz (CD), 48 kHz, 88.2 kHz, or 96 kHz.[5] The approximately double-rate requirement is a consequence of the Nyquist theorem. Sampling rates higher than about 50 kHz to 60 kHz cannot supply more usable information for human listeners. Early professional audio equipment manufacturers chose sampling rates in the region of 50 kHz for this reason.
'There has been an industry trend towards sampling rates well beyond the basic requirements: such as 96 kHz and even 192 kHz[6] This is in contrast with laboratory experiments, which have failed to show that ultrasonic frequencies are audible to human observers; however in some cases ultrasonic sounds do interact with and modulate the audible part of the frequency spectrum (intermodulation distortion).[7] It is noteworthy that intermodulation distortion is not present in the live audio and so it represents an artificial coloration to the live sound.[8] One advantage of higher sampling rates is that they can relax the low-pass filter design requirements for ADCs and DACs, but with modern oversampling sigma-delta converters this advantage is less important.' - wikipedia
Well not really. More bits increases the dynamic range available and higher sampling frequency means you can reproduce higher frequencies but it does not mean that the analogue waveform resulting is 'more accurate', it is in fact no different. There is no real comparison with digital photography
Why not?
If one were going to accept this then what do you say to sampling rates of 20 khz or 6 khz or 1 khz or 100 hz? If you accept there is a difference with 100hz and 44.1 khz then do you say that difference is not because the resulting analogue waveform is not more accurate?
Firebottle
11-12-2014, 07:24
No it's not more accurate, it's more extended in frequency.
A 100Hz sampling rate would give you an accurate representation up to about 40Hz bandwidth.
:cool: Alan
The spec for red book is designed to give more than sufficient frequency response and dynamic range. Reduce it further you lose dynamic range and lower the maximum frequency you can reproduce. Go to far and it will become audible.
sorry, go too far and what will be audible?
StanleyB
11-12-2014, 09:59
sorry, go too far and what will be audible?
Sampling errors. Voices will start to sound metallic/digital.
if done properly why would there be sampling errors?
sorry, go too far and what will be audible?
lets say you reduce the bits - you lose dynamic range i.e the difference in dB between the quitest sound on the recording and the loudest.
Reduce the sampling frequency and you reduce the maximum frequency that you can reproduce. At some point you will get to a situation where you can no longer reproduce signals that are in the audible range. How much that effects the sound will depend on your hearing acuity and the recording.
Studios like to record at 24 bit because it gives them a lot of dynamic range to play with in the recording process, making their life easier. You don't want to push a digital recording into the red.
Now if you have ever heard an unmastered red book cd recording taken straight from an recording session you will realise a) why CDs are mastered and dynamic compression is added and b) red Book CD has more than enough dynamic range from 16 bits for any domestic (and probabaly any pro) situation (as you hear the bass response trying to pull the loudspeakers to bits).
No it's not more accurate, it's more extended in frequency.
A 100Hz sampling rate would give you an accurate representation up to about 40Hz bandwidth.
:cool: Alan
Can we put the frequency range to one side for a moment, and consider this: If one is playing a musical note - lets say 260 hz. Recorded and played back in mono at 22.5 khz. This would be CD quality. If the same note is recorded and played back at 96 khz is that going to sound the same?
I would suggest if you isolate a single note, at a defined frequency and that frequency is no more than half the sample rate it will sound identical purely as function of sampling frequency
however looking at a piece of music as a whole and assuming your speakers can produce said frequency a higher sample rate will give you more of the original sound.
that sound may arguably be above human hearing range, however you're then looking at how high an individuals hearing can go? and do we fully understand the impact of frequencies we technically can't hear on the experience. That's a less simple question
Personally I'm open to the possibility that we sense frequencies above the audible level in some way that affects our enjoyment. Note I'm not saying its fact but I'm open minded on the point
StanleyB
11-12-2014, 17:56
however looking at a piece of music as a whole and assuming your speakers can produce said frequency a higher sample rate will give you more of the original sound.
First of all, there are two types of sampling: up and over. Which one do you mean?
Secondly, a higher sampling rate at the playback end won't give you more of the original sound if the audio information wasn't there in the first place or was missed out when the music was sampled at the recording stage.
agreed
I'm assuming the spec of the recording is at least as good as the playback or the information won't even be there
I see no point whatever in upsampling red book to supposed high def
Personally I'm open to the possibility that we sense frequencies above the audible level in some way that affects our enjoyment. Note I'm not saying its fact but I'm open minded on the point
The ultrasonic frequencies will excite the air in which they move. The audible frequencies move through that same air so I suppose the possibility exists for it to make a difference. These exotic supertweeters seem to have enough buyers to justify production. They can't all be fools can they? Another thing I read a long time ago when .mp3s and .oggs were newish was a study that had a result that the test group were better able to determine 'direction' when the control group could not. The experiment was done whereby the control group's music samples were cut off at some frequency point. I think it was at the limit of what mp3s could do, while the test group were given the same music but the frequencies were not cut off.
that sound may arguably be above human hearing range, however you're then looking at how high an individuals hearing can go? and do we fully understand the impact of frequencies we technically can't hear on the experience. That's a less simple question
Personally I'm open to the possibility that we sense frequencies above the audible level in some way that affects our enjoyment. Note I'm not saying its fact but I'm open minded on the point
That pretty much sums up where I am with this as well. We know what frequency humans can hear up to, that is easily tested. But can higher frequencies than that have an effect on how we hear lower frequencies? A lot of people would say definitely no. I'm keeping an open mind, although 'better mastering' still seems the more obvious ands logical alternative for the improvements reported.
As regards super tweeters, they pretty much all operate within audible range as well as inaudible. It would seem to me that you wouldn't consider super tweeters unless you were already unhappy with your systems HF performance. And if you are then just sticking on more tweeters is probably not the best solution. So I don't think the fact that some folk use super-tweeters is indicative of an ability to 'hear' what, in theory, we shouldn't be able to.
I've been busy reading the white papers released through AES on this and they've made for some absolutely fascinating reading. I'm not sure I understand even 30% and I'm happy to be corrected on this, but in a nutshell what I understand is that the people behind MQA are saying that we've been doing this sampling malarkey all wrong. We've been concentrating on representing frequencies accurately, when the way the human brain picks up on sound, the frequencies themselves are very much secondary to timing information which allows us to position the origins of sounds more accurately (stands to reason from human evolution). The measurements suggest that the average human is sensitive down to 10us of timing accuracy with more sensitive listeners managing down to around 5us. Red book CD, if I've got my calculations correct is accurate down to around 23us which suggests that timing accuracy is somewhere below what we can perceive, so what MQA is doing is to re-introduce higher timing resolution into a losslessly compressed stream.
I went into the Meridian shop in Chelsea last week to have a listen for myself. Sadly, there was no way to compare MQA with a standard PCM recording of the same music, but what I heard completely blew me away. A selection of songs, at anything between 96/24 and 384/24. Every single one of them sounding natural, not a hint of grain or harshness, incredible power and authority. Breathing, natural music.
Tellingly all of the recordings were quite old - Sinatra, Roberta Flack, Dave Brubeck, some early Metallica, Steely Dan, etc. so clearly all analogue recordings - So does this mean that anything recorded digitally using PCM cannot be remastered into MQA, presumably because the information is already lost...
Interesting. I've seem the timing accuracy figures disputed before, though. What level of discrimination we are capable of perceiving seems to be open to debate, like much of this.
You would think they would have a PCM recording there for comparison, it is a little suspicous that they don't. It would seem prima face to indicate that there is very little or no difference when compared back to back.
After the blatent false marketing of SACD and DVD-A I'm more than a little suspicious of formats with 'improved sound quality' claims nowdays.
I don't think there was any intention to be misleading during the demo and the people there were just genuinely keen to play you as much MQA as they could get their hands on. I'm not sure it would have helped me that much not being familiar with the equipment or room if there had been standard PCM for comparison, but the only real way of telling whether there is truly a difference will be to play with it for yourself. I ended up buying one of the little Explorer2 DACs, which I am mightily impressed with in standard PCM mode (as I wanted a good quality DAC/HP amp for my travels anyway), but I'm keen to get my hands on some MQA downloads for an evaluation in more familiar territory. It looks as if the first wave of music is set to hit after CES.
I don't think there was any intention to be misleading during the demo and the people there were just genuinely keen to play you as much MQA as they could get their hands on. .
Maybe not but given the artists you mention are hardly obscure (SteelyDan, Brubeck etc) it would not be in any way difficult to have a comparison available:
'Here it is on Red Book and now here it is on MQA: what a difference!'
If they are claiming improved sound quality it is a bit odd not to have a comparison there to demonstrate it unless a comparison would be undesirable.
It's pretty much certain that whatever you had in a pre-existing format would be a different mastering, so you'd be comparing apples with oranges anyway. As MQA files are going to be backwards compatible then playing the same file on an MQA and then a non-MQA enabled device is going to be a better comparison. I'm guessing there is no MQA off button on the Meridian kit...
Are they re-mastering everything again for MQA then? Surely they are just using whatever 'hi res' version exist currently and running it through their software. That was what I understood from what I read.
The details are all rather sketchy at the moment. I couldn't get much information due to the fact that the companies involved, be it record companies or Meridian themselves are all under NDAs at the moment. I guess we'll know a lot more come the CES announcements.
I understand that one of the major improvements made possible is some form of profiling from the original A2D conversion. Mainly affecting timing issues as AD's of age of course also has got their own "signatures".
Whether any of this is revolutionary i must say is questionable.
There are still a lot of questions to be properly answered, especially how it's possible to store audio material in the band 24KHz-96Khz within the boundaries of a 48Khz sampling frequency. (Using the excessive dynamic headroom of the 24bit wordlength.)
However, anything that closes in on a pure analog reproduction system is a good thing! ;)
-Mike
My understanding is MQA sets out to do two things: 1) Offer true high res audio for streaming but at data rates comparable or less than standard Red-Book CD 2) Offer better sound quality when the original master is encoded with MQA and then decoded with an MQA capable DAC.
Point 2 is the critical gotcha, you need to encode first and have an MQA compatible DAC which would mean (I presume) a license agreement with Meridian….that might be a show stopper right there.
Good info here: http://www.stereophile.com/content/ive-heard-future-streaming-meridians-mqa
eisenach
03-01-2015, 13:37
Good info here: http://www.stereophile.com/content/ive-heard-future-streaming-meridians-mqa
Thank you for the link. It's the best explanation I've seen; I almost understand it !
It all sounds very interesting and one to keep an eye on.
The measurements suggest that the average human is sensitive down to 10us of timing accuracy
Sound at STP travels about 3mm in 10us, I would doubt is we can keep our head that still unless one used a clamp.
tizer2000uk
02-05-2015, 11:57
I know this thread is a little old now but I had the opportunity to hear MQA at a demo and managed to A/B Dave Brubecks Take Five which pretty much left my jaw hanging!
There is most definitely a night and day difference between red book and MQA which isn't just about an increase in resolution, MQA is about looking at all of the original components used in the recording and correcting for that as well as corrections for timing information. It sounded like I was there with the band, and in between quick exchanges of dialogue I kept finding myself looking at the instruments being played. The cd version still had the instruments in the same places just without the feeling we were all in the same room. Absolutely incredible, even more so for a recording made in 1959!
Now most people haven't followed Meridians purist approach (everything in the digital domain until it hits the amps with no analogue crossovers) to the reproduction of sound so how this will all sound on the sort of systems most of us own remains to be heard.
As for licensing the technology, well Meridian are doing this for free! Best way of ensuring wide acceptance.
Admittedly the system cost in the region of 20k so is currently out of my price range but at least I know now what to aim for.
Rothchild
03-05-2015, 07:31
I know this thread is a little old now but I had the opportunity to hear MQA at a demo and managed to A/B Dave Brubecks Take Five which pretty much left my jaw hanging!
There is most definitely a night and day difference between red book and MQA which isn't just about an increase in resolution, MQA is about looking at all of the original components used in the recording and correcting for that as well as corrections for timing information. It sounded like I was there with the band, and in between quick exchanges of dialogue I kept finding myself looking at the instruments being played. The cd version still had the instruments in the same places just without the feeling we were all in the same room. Absolutely incredible, even more so for a recording made in 1959!
Now most people haven't followed Meridians purist approach (everything in the digital domain until it hits the amps with no analogue crossovers) to the reproduction of sound so how this will all sound on the sort of systems most of us own remains to be heard.
As for licensing the technology, well Meridian are doing this for free! Best way of ensuring wide acceptance.
Admittedly the system cost in the region of 20k so is currently out of my price range but at least I know now what to aim for.
What was the transcription process from the 1959 master to red book / MQA so we can be assured that the only difference was MQA?
Given that Meridian are a software licencing company (by their own account) I suspect it's unlikely that the whole codec will be made free (and it certainly would only be 'free beer' free as opposed to 'free speech' free) odds on they free only one end, most likely the decoder, and will then sell the encoder to soundcard manufacturers to build in to their hardware so that there's MQA validated audio coming in to the stream of stuff available for listeners.
It is a massive feat to pull off for Meridian. The cynic in me says fail but they might manage some penetration.
Out of curiosity was it an all Meridian system?
tizer2000uk
03-05-2015, 21:59
Not sure what else occurred in the process but given they correct for the sound signatures of the original equipment, mic's, mixing equipment etc it can't be said it has followed the same route that the cd version followed although my logical mind says that they have followed the correct line in delivering what the artist wanted you to hear. The really clever bit is compatibility with a lot of existing digital audio players, on equipment lacking an MQA decoder you just get to hear the cd quality version, with the right kit however the MQA stream is delivered. As I understand, Tidal are going to be streaming titles in MQA which given their recent announcements means that MQA will be a success.
The free bit is the rights to the decoders so hardware manufacturers can include them in their products, I would imagine there is a nominal fee for titles sold with MQA.
The demo I heard, only had Meridian kit. I have since heard a number of other high end systems using some of the tracks from the demo and the only speakers to come close to what Meridian offer and by close I mean still trailing behind were a pair of Linn Akurate's.
If anyone wants to sell a kidney to help fund a Meridian system... ��
Rothchild
04-05-2015, 07:28
they correct for the sound signatures of the original equipment, mic's, mixing equipment etc
What does this mean? A substantial part of the sound of a recording is that which is imparted by the tools and methods used to make it - if the engineer spent their lifetime of experience selecting the 'right' mic for the thing being recorded (in the context of the song/music) how would removing its signature improve things? (and what do you imagine we'd actually be left with).
'There's a Riot Goin' on' 'corrected' would be an interesting project but I'd argue that it wouldn't still be Sly's record....
The cheapest streamer is the UK invented Raspberry with a DAC stuck onto it. I would really like to hear a "high-end" multi hundred (thousand) $$ streamer compared to the humble Raspi.
Me too! Why pay for a 'fancy badge' if the product doesn't sound any better?
Marco.
tizer2000uk
04-05-2015, 19:48
I believe it concerns adjusting for the playback equipment such that you get as close to the original recording as possible. Would be great if the amp or player could be fed a list of components in the system and the player then re equalize for them.
I highly recommend getting to a dealer with a demo setup and hear for yourselves!
Rothchild
05-05-2015, 07:09
I just went back to the link in the OP, I'd be interested to see where you're getting this info about MQA working as a de-emphasis curve to filter out the recording?
What I'm seeing is that it's actually a lossless compression codec with a (not to loudly spoken about) digital watermark (copyright protection) built in.
Essentially they've realised that some folk will pay extra for 24 / 96 even thought the amount of real data above 20khz is miniscule (if there's any there at all) so they're using that space to help pack the data more efficiently for transfer across a network ('fold it in on itself' as they put it).
There's some interesting phrasing in the OP, I like:
"Daft Punk's Get Lucky sounded almost exactly like the 24bit/96KHz high resolution FLAC"
"...difference between the MQA-encoded download of Metallica's Enter Sandman and our own MP3-encoded version."
Well, what do you know, a lossless encoding of a record sounds better than a lossy encoding of a record :stalks:
As for the concept of 'getting closer to the original recording' (whatever that actually means), for Take 5 you'd want to find a first press vinyl straight from the master tape, everything else is necessarily degraded.
MQA is a benefit to the record and distribution industry because it makes files smaller (and therefore 'cheaper' to store and distribute) and it adds a digital watermark to files that will enable to a copy protection system which will generate revenues back to Meridian. I can see very little benefit back to music lovers (especially ones who value quality over convenience).
That is all it is unless this 'corrected timing' thing actually does make an improvement to the sound.
Trouble is if you go to an MQA demo you have no way of being certain that you are listening to the same master. Or if they are doing the comparison against a lossy MP3. The same trick they used when DVD-A and SACD were launched.
Rothchild
05-05-2015, 09:05
IMO the timing thing is a mis-direction, there's no 'correction' going on - unless they're remaking the whole record with a better master clock on the converter. The key to it is in here:
"Essentially data that would normally be stripped by a traditional MP3 encode is instead folded in on itself to occupy the frequency spectrum otherwise inaudible to the human ear."
MP3 works on a psychoacoustic principle whereby certain frequencies at certain levels will 'mask' other frequencies at lower levels, therefore the frequencies being 'masked' can be removed without the listener noticing.
What it appears they are doing is looking for places in the files where there is sequentially zero data, for most people that's the gaping hole above 20kHz that is neither caught by the recording equipment nor heard (in anything but an audiophile sense) by the listener - you could then write a single bit flag that says 'nothing for X period' rather than writing X number of 0s in the file. Upon decoding the decoder then just writes in the appropriate number of nothing bits rather than carting the around all those empty bits.
I think 'hiding' data in bits of signal you can't hear is an old trick, it used to be used for things like quadrophonic and the like?
I also have to state the love I have for this statement :mental::
"basing their research on neuroscience and psychoacoustics rather than frequency graphs and oscilloscopes."
Clearly not all science is equal, well known and commonly understood engineering principals will have to take a back seat to more complex and less well understood research science because otherwise we might understand how the emperor's clothes were constructed. Also the notion that psychoacoustic research could be conducted without use of frequency graphs or an understanding of the properties of sound, in a physics sense, is laughable.
Indeed how else was "the frequency spectrum otherwise inaudible to the human ear." understood without a someone mapping and graphing it at some point (be they a neuroscientist or psycho acoustician)?
MQA is not for music lovers, it's for the corporate machine that drives the commercial record industry.
Nobody cares about the file size so if it does not sound better as well then it has little chance of catching on.
From the Stereophile article linked to earlier:
...the old question is why do we need to preserve and reproduce frequencies above the limit of human hearing, even if we can do it? Bob spent some time discussing this in his presentation and it comes down to the fact that the ear-brain doesn't just operate as a frequency analyzer. Evolution has fine-tuned the system to be able to detect temporal differences that are equivalent to a bandwidth considerably greater than 20kHz and that the anti-aliasing filters in A/D converters and reconstruction filters in D/A converters introduce temporal smearing that it is considerably greater than what our ear-brains are tuned to expect from natural sounds: this smearing is, I believe, responsible for so-called "digital" sound.
The MQA encoder and decoder together have been designed to have a transient response of the same form and order as that of the temporal sensitivity of the ear-brain. And if at the MQA-encoding stage, the temporal effect of the A/D converter can be compensated for, the complete system offers a transparent window into the original musical event. Meridian describes this as "taking an original master further, toward the original performance, in an analogous way to the processes expert antique picture restorers use to clean the grime and discolored varnish from an Old Master to reveal the original color and vibrancy of the work."
One question I have is regarding the material used at the demonstration i.e jazz recordings from the 1950s. Exactly how much musical information above 20KH will be on these recordings given the limitations of the mics, amps and tape recorders of that era? My guess would be none.
Rothchild
05-05-2015, 13:09
I missed that article on my second skim through the thread, thanks for the prompt.
Timing is still a bit of a red herring (when presented as a counterpoint to frequency especially) because frequency is timing - For reference the peak to peak duration of a 20kHz wave is 50 micro seconds, the 10 micro seconds referred to elsewhere equates to 100kHz - which, according to Wolfram at least, is audible to bats, mice and beluga whales but not to dogs, elephants or horses ;-).
When they talk about 'the temporal effect of the original encoder' they really can only mean 'the removal of frequencies above Nyquist' - It's wishful thinking to believe that anyone can put that data back in once it's been discarded by the anti-aliasing filter.
What the stereophile thing expounds is that the very small amount of information above 20kHz has a very low dynamic range - so subsequently requires a far lower bit depth in order to be able to capture it properly, much like VBR Mp3 encoding.
Fig one of that article is interesting insofar as calling the red line 'musical' information seems a bit of a stretch, to my judgement it looks like a statistical error against the calculation of the noise floor!
.
Fig one of that article is interesting insofar as calling the red line 'musical' information seems a bit of a stretch, to my judgement it looks like a statistical error against the calculation of the noise floor!
That was sort of my point - it is a 1950's analogue recording how can there be musical content at that frequency? The mics couldn't pick it up and even if they did the tape deck could not have recorded it.
tizer2000uk
06-05-2015, 19:34
[/QUOTE]
MQA is not for music lovers, it's for the corporate machine that drives the commercial record industry.[/QUOTE]
If the format bests every other digital format out there, it's for music lovers, maybe not for the green pen/tin foil hat wearers that insist music sounds better if the cables are free range and given weekly massages.
Given that Bob Stuart et al have been instrumental in the birth and subsequent improvements/pioneers of digital audio, anything they do to further the industry is only going to be a good thing, regardless of whether they get rich doing so.
Suggesting that neuroscience is not science is somewhat ill informed, all they have said is that just looking at frequency graphs is not enough to truly understand how we best package music for optimum quality.
As for the quality of the recording, well it had to be pretty good, I have the original cd which sounds as good in terms of its frequency range as the MQA version. As the most popular Jazz tune of all time there will likely be a copy up on YouTube to listen to, have a listen and see what you think.
Rothchild
07-05-2015, 09:34
I don't think it's besting anything, it's basically 24/192 in a smaller file (wrapped in a proprietor layer).
tizer2000uk
07-05-2015, 17:42
I don't think it's besting anything, it's basically 24/192 in a smaller file (wrapped in a proprietor layer).
It sounds better than any format out there, is backward compatible with non MQA devices so providers now need only have one file instead of two and did I mention it sounds better?
Rothchild
07-05-2015, 18:52
did I mention it sounds better?
Yes, you also mentioned a lot of bobbins about 'correcting' timing and compensating for original recording equipment which turns out on further investigation not to be the case (MQA is just a clever way of losslessly compressing a high resolution 'wav' file).
I think I'll let the green pen, free range cable gang rush in with their money first, but thanks for the heads-up.
I'm with you Marc I'm definitely in the sceptical camp;)
Of course they sound good! The files are truer to the original source due to the enhanced techniques for correcting the original AD-process...
But does that make the format better? No, not to my ears and brain...
I'm all for better mixing and mastering, but give me a proper 24/48 PCM file and ill be satisfied!
Rothchild
10-05-2015, 18:07
I can't see any decent reference to 'correction' all I can find is this from the stereophile article:
"The MQA encoder and decoder together have been designed to have a transient response of the same form and order as that of the temporal sensitivity of the ear-brain."
As discussed 'temporal sensitivity' = frequency (as in how frequently something (a pressure wave when we're talking about sound) happens)
"And if at the MQA-encoding stage, the temporal effect of the A/D converter can be compensated for, the complete system offers a transparent window into the original musical event."
My highlight because it's a big if and imo it isn't and doesn't. Clearly it is not possible to put back something that isn't there, if the source was recorded digitally it only contains frequency (timing) information up to half the sample rate (and the likelyhood that a 1959 tape has much more is likely an esoteric belief), if it's adding stuff that's not in the master then it clearly isn't:
"... "taking an original master further, toward the original performance, in an analogous way to the processes expert antique picture restorers use to clean the grime and discolored varnish from an Old Master to reveal the original color and vibrancy of the work."
Powered by vBulletin® Version 4.2.3 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.