Guest Eggcup The Daft Posted December 16, 2018 Share Posted December 16, 2018 12 hours ago, davewantsmoore said: By not damaging the audio .... rather than "delivering high frequencies". We've argued about the high frequency/impulse response feature and what it actually is already, so I know you understand that it is there and why it is there. It is an integral part of MQA. I don't see why you are suddenly saying that MQA is not about that. A number of aspects of MQA are about improving/changing the audio from the studio master - impulse response, treating for the effect of the ADC - and changing the music in the process - dropping bits, using multi bit dithering to improve the noise floor to compensate. And so we come full circle, because dropping bits and reducing bit depth is precisely what Archimago is pointing at in his criticism of MQA. In other words, MQA is, in one area, quite clearly damaging the audio. You say the light comes on to show that the audio is not damaged, when it is. And it is doing it exactly to selectively deliver high frequencies. Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 16, 2018 Share Posted December 16, 2018 12 hours ago, legend said: we can discriminate down to 10 uS in the time domain and the latter requires a bandwidth of 100 kHz so a sampling frequency of 200 kHz (assuming Nyquist/Shannon). Inter-channel.... sure. .... but within one channel?! Hopefully they're right, but I can't hear it. Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 16, 2018 Share Posted December 16, 2018 1 minute ago, Eggcup The Daft said: We've argued about the high frequency/impulse response feature and what it actually is already, so I know you understand that it is there and why it is there. It is an integral part of MQA. I don't see why you are suddenly saying that MQA is not about that. Someone is confused then.... 1 minute ago, Eggcup The Daft said: A number of aspects of MQA are about improving/changing the audio from the studio master - impulse response, treating for the effect of the ADC - and changing the music in the process - dropping bits, using multi bit dithering to improve the noise floor to compensate. I don't know what you're trying to say here. MQA is simply (not that I'm attempting to defend them any) trying to not damage the audio. 1 minute ago, Eggcup The Daft said: And so we come full circle, because dropping bits and reducing bit depth is precisely what Archimago is pointing at in his criticism of MQA. Indeed, he is. Here is what he says about that in the link from the original post of this thread: Quote 2. Technical concerns: a. It reduces the actual bitdepth to the aforementioned "typically 15.85 bits" and up to "17-bits" resolution when decoded. These numbers are from Bob Stuart. I know they want us to ignore this and focus on the analogue output rather than care about the digital technical values. We can argue about this of course; but the point is that a resolution limit has been imposed which is lower than standard 24-bit digital audio. His technical concern about reducing the bit rate, is that the bit rate is being reduced. The term "resolution limit" is quite concerning honestly. He understands this stuff much better, than to use such misleading language. 1 minute ago, Eggcup The Daft said: In other words, MQA is, in one area, quite clearly damaging the audio. Reduction of bit rate? No. It is not. 1 minute ago, Eggcup The Daft said: You say the light comes on to show that the audio is not damaged, when it is. And it is doing it exactly to selectively deliver high frequencies. Reducing the bit rate isn't doing any damage ..... and they're not trying to deliver high frequencies. For all MQA care there could be zero content at > 20khz What they're trying to do, by using the higher sampling rate information, is to create filters (eg. resampling filters, or the filters which operate inside your DAC) which do no damage the audio. Link to comment Share on other sites More sharing options...
Guest Eggcup The Daft Posted December 16, 2018 Share Posted December 16, 2018 @MLXXX Welcome to the joys of the MQA dispute. Your test can effectively be done now, by taking the highest resolution MQA files on the 2L website and comparing to the 16/44 files there, as well as unprocessed files. I've not heard much difference with the 2L files on Tidal. Link to comment Share on other sites More sharing options...
MLXXX Posted December 16, 2018 Share Posted December 16, 2018 23 minutes ago, davewantsmoore said: Audibility is very difficult and complex. I think it's a poor yardstick, for the most part. Although it is hard to be sure about in borderline cases, it is a very important test to satisfy if you wish to succeed in selling a new technology to consumers. If you cannot demonstrate that a different audio format definitely sounds better to real people listening to real recordings of music, you are in trouble. As an example of that, SACD stereo, after all these years, hangs by a thread. Link to comment Share on other sites More sharing options...
Volunteer sir sanders zingmore Posted December 16, 2018 Volunteer Share Posted December 16, 2018 2 hours ago, MLXXX said: It can offer objectively better performance in the time domain, Perhaps not https://archimago.blogspot.com/2018/02/musingsmeasurements-on-blurring-and-why.html?m=1 Link to comment Share on other sites More sharing options...
MLXXX Posted December 16, 2018 Share Posted December 16, 2018 15 minutes ago, Eggcup The Daft said: @MLXXX Welcome to the joys of the MQA dispute. Your test can effectively be done now, by taking the highest resolution MQA files on the 2L website and comparing to the 16/44 files there, as well as unprocessed files. I've not heard much difference with the 2L files on Tidal. That might have to wait till after Christmas, when I have access to my "good" headphones. Link to comment Share on other sites More sharing options...
eltech Posted December 16, 2018 Share Posted December 16, 2018 3 hours ago, davewantsmoore said: You want to play back what you have recorded without damaging it. Can you refresh my memory about how audio is damaged, by recording and playing back at high sample rates(for example)? Link to comment Share on other sites More sharing options...
Guest Eggcup The Daft Posted December 16, 2018 Share Posted December 16, 2018 7 minutes ago, davewantsmoore said: Reduction of bit rate? No. It is not. Did you not say that you wanted the best transfer of audio information, regardless of whether the changes are audible? The reduction of bit rate may be inaudible, but it is a reduction in the information available. As such, it fails in your terms. Is audibility the test (I guess it should be) or not? I think we're both trying to have it both ways here. If audibility is the test, then you're correct. If the best transfer of audio information regardless of whether the changes are audible is the test, then no, it's damage. I presume we can at least agree that there is a change to the audio in MQA processing. Playing the original master doesn't put on the "master quality" light. 20 minutes ago, davewantsmoore said: and they're not trying to deliver high frequencies. Except that they are, as explained by Bob Stuart in the link I gave MLXXX earlier. Remember the "origami" triangle? It stretches out to 48kHz. They aren't necessarily delivering high frequencies for the sake of delivering high frequencies, but are preserving selected information so other aspects of their processing works. Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 16, 2018 Share Posted December 16, 2018 2 minutes ago, eltech said: Can you refresh my memory about how audio is damaged, by recording and playing back at high sample rates(for example)? Resampling. 1 minute ago, Eggcup The Daft said: Did you not say that you wanted the best transfer of audio information, regardless of whether the changes are audible? The reduction of bit rate may be inaudible, but it is a reduction in the information available. No it's not. Let's say I have some audio, which has a noisefloor that is 36dB below the peak. I need 6bits to encode the audio above the noise .... perhaps I'd like to encode the noisefloor too ..... let's add a few more bits, let's just add another 10 (!!!!). So we use 16bits. If we're in a 24bit container, we can drop bits without losing any information..... assuming some other corner case caveats are respected. 1 minute ago, Eggcup The Daft said: As such, it fails in your terms. >_< 1 minute ago, Eggcup The Daft said: If the best transfer of audio information regardless of whether the changes are audible is the test, then no, it's damage. The "best transfer" .... doesn't mean "all the information". 'cos some of the information doesn't matter. It doesn't even necessarily mean transmitting all the information "unchanged" from the original master.... because it's what comes out the DAC which matters (and that is what you have to compare). 1 minute ago, Eggcup The Daft said: Except that they are, as explained by Bob Stuart in the link I gave MLXXX earlier. Remember the "origami" triangle? Link to comment Share on other sites More sharing options...
eltech Posted December 16, 2018 Share Posted December 16, 2018 Just now, davewantsmoore said: Resampling You've misunderstood. I mean recording and playing back at the same frequency. (No resampling involved) Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 16, 2018 Share Posted December 16, 2018 Just now, eltech said: No resampling involved Most DACs and ADCs resample internally....... but otherwise, if you record, distribute and playback all in one rate, then you could avoid it. Link to comment Share on other sites More sharing options...
eltech Posted December 16, 2018 Share Posted December 16, 2018 (edited) 7 minutes ago, davewantsmoore said: Most DACs and ADCs resample internally....... That's a fair point. They do. They oversample. Up to 128 times for some. What's the problem with that? How is that damaging audio? (Edit) there would be lots of DAC chip designers who'd disagree. I bet they'd say it improves the audio. Edited December 16, 2018 by eltech Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 16, 2018 Share Posted December 16, 2018 3 minutes ago, eltech said: That's a fair point. They do. They oversample. Up to 128 times for some. What's the problem with that? How is that damaging audio? The filters used are not necessarily very good. The idea is that if you pre-process the audio into the higher rate (using filters which are higher quality), then you can get better performance. This is already a popular things to do with player software, such as Audrivana, HQplayer, XXhighend and others.... with especially the latter two focusing on specific filters for specific DACs. (XXhighend make their own "dumb" DAC, specifically to use software filters on). The real "long answer" to all of this, is naturally quite a bit more complex. Link to comment Share on other sites More sharing options...
eltech Posted December 16, 2018 Share Posted December 16, 2018 (edited) 5 minutes ago, davewantsmoore said: The filters used are not necessarily very good Ok, so the bugbear is with the DAC chip hardware, not digital audio recording or playback per se? So if, hypothetically, the Delta Sigma DAC used "superior" digital filters all would be ok? Perhaps this will happen over the coming years anyway? So, no need to fart around with obscure new formats? Edited December 16, 2018 by eltech Link to comment Share on other sites More sharing options...
eltech Posted December 16, 2018 Share Posted December 16, 2018 (edited) A huge number of people can't tell an MP3 320 from a .wav. https://www.avhub.com.au/news/sound-image/did-sony8217s-chief-sound-architect-just-tell-me-they-can8217t-hear-or-measure-the-difference-between-upsampled-248k-files-and-true-high-res-audio-wow-437891 High sample rates eliminate "the ? problem" So, it seems that no matter how good things are people like to complain... A lot! I don't need to hear dark side of the Moon for the 10 billionth time on the twenty-seventy-th format. I'm sure the tapes were "damaged" during the first overdub. Just saying. Edited December 16, 2018 by eltech Link to comment Share on other sites More sharing options...
MLXXX Posted December 16, 2018 Share Posted December 16, 2018 (edited) 4 hours ago, Sir Sanders Zingmore said: 6 hours ago, MLXXX said: It can offer objectively better performance in the time domain, Perhaps not https://archimago.blogspot.com/2018/02/musingsmeasurements-on-blurring-and-why.html?m=1 This is material I had seen before and which had made me sceptical about MQA. There do seem to be a wide variety of doubts cast. However I suspect a technical jury would be unable to arrive at a verdict on the "time domain performance" question, without more evidence/explanation. Edited December 16, 2018 by MLXXX Link to comment Share on other sites More sharing options...
davewantsmoore Posted December 17, 2018 Share Posted December 17, 2018 11 hours ago, eltech said: can't tell an MP3 320 from a .wav It's not really surprising. Lots of work went into developing MP3 and similar codecs, with the goal of being perceptually transparent. If you could easily/always tell the difference, when using the most expensive (320kbps) settings of MP3, then they would have completely failed in their goal. MP3 got a bad name in the 2000s, when it was poorly used - suboptimal encoder settings or insufficient data rates - resulting in clearly audible artefacts (vs the original). 11 hours ago, eltech said: https://www.avhub.com.au/news/sound-image/did-sony8217s-chief-sound-architect-just-tell-me-they-can8217t-hear-or-measure-the-difference-between-upsampled-248k-files-and-true-high-res-audio-wow-437891 Sure.... If you start with audio which is (almost totally) perceptually transparent from the original .... then you can do things to it when you play it back, to "limit the damage" you do to it. ie. to make what comes out the analogue output, "better". This "damage limitation" can be done whether the audio you started with (the "almost totally perceptually transparent from the original" audio) .... is stored in a "lossy" or "lossless" format. Audiophiles need to let go of this notion of: If you "changed" the audio, then you damaged it. That is not necessarily true... and it is definitely not a "yes/no" question. 11 hours ago, eltech said: High sample rates eliminate "the ? problem" The filters that they use on the audio run at higher sampling rates. Here's an analogy. We're designing a new (fast) car. We do our research, on all the different engine types, and we conclude that to achieve the needed level of (extremely high) performance we must use a v8 petrol engine. Do all cars in the world which have v8 petrol engines have "extremely high performance"? No. Some are garbage. You can encode complete garbage with high sample rates..... but if you are making a filter which is going to try to best render the audio (even audio with low sampling rates) .... then you will probably want to use a filter which runs at a high(er) rate. 11 hours ago, eltech said: I don't need to hear dark side of the Moon for the 10 billionth time on the twenty-seventy-th format. I'm sure the tapes were "damaged" during the first overdub. Just saying. It's fair point ..... and I think people can understand how/why this happens (the focus on the technology, as opposed to how the technology is used). 1 Link to comment Share on other sites More sharing options...
legend Posted December 17, 2018 Share Posted December 17, 2018 12 hours ago, eltech said: I don't need to hear dark side of the Moon for the 10 billionth time on the twenty-seventy-th format. I'm sure the tapes were "damaged" during the first overdub. Just saying. I don't think it was just the overdubs that damaged the DSOTM master analog tapes - it was also the pretty ordinary op-amps that were only available at the time for the mixing desks etc. I can hear it as a certain 'greyness' in the sound - that only became more obvious at I went from LP (Japanese pressing), to CD to SACD! I guess there is little if anything that can be done about any technical 'faults' in the original analog master-tapes but as Dave-wants-more says one should try to reduce or correct if possible any further damage in the digital domain - and then of course in the rest of your hifi/sound reproduction system! Link to comment Share on other sites More sharing options...
Guest Eggcup The Daft Posted December 17, 2018 Share Posted December 17, 2018 (edited) 4 hours ago, davewantsmoore said: Audiophiles need to let go of this notion of: If you "changed" the audio, then you damaged it. I think this is the bit of your argument I'm not getting. For me, the best transmission of the audio (the words you used, my emphasis) is the one that changes it the least. I'm not concerned with someone else's definition of "damage", or someone else's definition of "improvement", in the transmission of the audio - and nor should you be. I'm sure we could have an MQA like process contained entirely in a DAC, albeit one that does a lot more work, with a standard file and some metadata for the ADC "damage" resolution. This is largely a philosophical difference but it's not necessarily a trivial one. After all, we were only discussing DRM a few pages ago... Edited December 17, 2018 by Eggcup The Daft Link to comment Share on other sites More sharing options...
eltech Posted December 18, 2018 Share Posted December 18, 2018 I don't think it was just the overdubs that damaged the DSOTM master analog tapes - it was also the pretty ordinary op-amps that were only available at the time for the mixing desks etc. I can hear it as a certain 'greyness' in the sound - that only became more obvious at I went from LP (Japanese pressing), to CD to SACD! I guess there is little if anything that can be done about any technical 'faults' in the original analog master-tapes but as Dave-wants-more says one should try to reduce or correct if possible any further damage in the digital domain - and then of course in the rest of your hifi/sound reproduction system!I was only making an example of DSOTM. The same applies to so many older albums that the record labels like to reissue Infinitely. Boring!When will people get over this reissue / remaster madness? Handing over hard earned money for the same thing in a slightly different package with slightly different sound is quite silly. How many remasters are a significant improvement, rather than just being different, or more compressed?Where I'm going with this is that we don't really need any more reissues of older albums. We just don't.Like you said, they're too old to be audiophile. The opamps and tape technology wasnt and isn't at contemporary standards.So what about early digital recordings? The same applies as above.(However, note: I don't have a problem whatsoever with them. I think you know my position on this)I don't think anything can be corrected here. Even if something can be, I'm fairly sure it will be a bodge job. It won't be the same as the original recording. It will just be a remaster of sorts. That leaves new recordings.Most people can't hear much if any difference between high sample rates and 44.1. Many people can't hear the difference between MP3 and lossless.What we presently have with existing digital audio is a ubiquitous technology that has been adopted worldwide and is open source. We ought to consider the serious benefit to humanity of having such a great free system of recording and playback. Introducing another format that attempts to denegrate an existing ubiquitous open source format with implications that the open source format is inferior, by using the presence or absence of a blue light to indicate "good" and "bad" could be a huge disadvantage on many levels. Acceptance or rejection of this model could have ramifications for video formats too.However the ubiquity of PCM recording will probably mean it's cemented as the format of choice for many years to come. When the perceptual differences are at the far edges of what can be perceived, we ought to seriously consider and weigh up what we'd lose, vs. what might be gained especially when gains if any are at the far edges of perception and may only be noticed by a very small percentage of people. The real cost of data is only in the size of HDDs and the electricity to run them and routine maintenance. Ultimately, this cost i guess is significantly less to the environment than producing and transporting vinyl records? Especially considering that thousands of hours of music can be stored on a tiny HDD in a server and streamed around the world.So, for purists, or the anxious, 192 or 384 kHz recordings should be the natural choice. They're open source and address the theoretical issues of the anxious audiophile, whereas mqa still leaves room for anxieties - the loosing of bit depth, the lossy encoding of ultra high frequencies, and the leaky digital filter. But for most people who can't hear a difference none of this matters, well except for if mqa is adopted as an industry standard.If that is the case, an open source format would be replaced with a format that lines the pockets of mqa. If even a single cent more is charged somewhere along the line someone is going to pay for it. If not the end user, then it will be the studio or the artist.I don't object to the creation of any new format because I accept that there probably is room in the world for things like this. MP3 has made many people happy, just as high res has made many people happy. I personally dispute the importance of the claims made by mqa, which are unproven. But should there be any truth to it, there will be a bunch of people working to create an open source version of it. Just because mqa files can be played back on regular equipment doesn't mean it's good. It's much like the upsell at fast food"restaurants", whereby if you spend an extra 50 cents you get the large coke.If there was really anything in it for you, the "restaurant" wouldn't do it. It all about them getting that extra 50 cents out of you! Your gut will thank you for avoiding the upsell.In the case of mqa, it's a downgrade for everyone who doesn't take the upsell of the mqa enabled DAC.Even though...it's all on the periphery of perception...Why bother? 1 Link to comment Share on other sites More sharing options...
eltech Posted December 18, 2018 Share Posted December 18, 2018 Continued...I mentioned open source.I think open source things are the future for this world.I'm sure we've all noticed that one of the biggest issues for humanity at the moment is the battle against neo liberal economics. It's not that people necessary have a problem with exchange of money for goods and services, but that people seek fairness and ethics in these transactions. People want value for money, and want everyone in the chain to be remunerated fairly. The potential for mqa to monopolize the audio distribution market and make massive profits is at odds with the open source concept.It's a form of rentseeking.Of course it's up to individuals to make their own conscious choices with what they support.To say politics and ethics isn't part of this discussion is simply not true. The way to take power back into our own hands is to contribute to the development of open source, not simply be consumers of the products of monopolized industries. For every cent they don't get, it's a cent that can go elsewhere, perhaps in this case, towards buying the album of a local artist? I realise this isn't something that's mutually exclusive, just an example.Mqa really is about DRM. But the major labels have a terrible history of giving artists a "take it or leave it" contract.Where most of the artists earnings come from live shows and merchandise, with the record labels getting most of the profits from album sales. The drm is about protecting record labels profits, not about protecting the artists. The artists often don't even own their own songs once they sign a contract. Is this fair? The narrative coming from the labels has always been about protecting the artists, but it's mostly about protecting their own revenue.When discussing intellectual property in terms of the recorded music industry we like to think it's about protecting the artists income, but they only get a pittance.It's terrible. 1 Link to comment Share on other sites More sharing options...
LHC Posted December 18, 2018 Share Posted December 18, 2018 6 hours ago, eltech said: To say politics and ethics isn't part of this discussion is simply not true. I have respect for all the things you wrote in your last two lengthy posts, and you put your case well. While there is a political and ethical dimension to MQA, I don't agree it is worthy of much discussion. Firstly we simply don't know enough about what is really going on behind the scene, and the company is being opaque and unhelpful with their marketing. Therefore much of the discussion on this dimension are just people guessing and conjecturing about what may happen and how it may play out. I just can't find any value in that. Secondly from our experience with the political and economic threads on SNA, those type of discussion mostly deteriorate into tribal arguments, lost of objectivity, personal attacks and closure of threads by the mods. There are good reasons why SNA no longer has a political sub-forum. I do not wish the same fate to befall the mainstream threads. 6 hours ago, eltech said: Is this fair? Again, these are subjective questions. There are no right or wrong answer, but simply difference in opinions. What is fair to some may come across as unfair to others. I may sound dismissive, but I don't see much objective value in these type of discussion. Link to comment Share on other sites More sharing options...
eltech Posted December 19, 2018 Share Posted December 19, 2018 I have respect for all the things you wrote in your last two lengthy posts, and you put your case well. While there is a political and ethical dimension to MQA, I don't agree it is worthy of much discussion. Firstly we simply don't know enough about what is really going on behind the scene, and the company is being opaque and unhelpful with their marketing. Therefore much of the discussion on this dimension are just people guessing and conjecturing about what may happen and how it may play out. I just can't find any value in that.Contemplating what the world will be like if things go this way or that has great merit, especially when we are at a critical juncture. People need to be aware of the power they have and the influence they can have on the world. Making informed decisions, and taking personal actions in line with desired outcomes is critically important and has great personal and public value. Again, these are subjective questions. There are no right or wrong answer, but simply difference in opinions. What is fair to some may come across as unfair to others. I may sound dismissive, but I don't see much objective value in these type of discussion. You're correct that there's no right or wrong answer to that question. I posed it as stimulus for pondering the ethics of the music industry. People will come to their own conclusions. Or perhaps they'll do further investigation. The value is in the thought provoking nature of the question. Link to comment Share on other sites More sharing options...
Guest rmpfyf Posted December 19, 2018 Share Posted December 19, 2018 On 18/12/2018 at 11:35 PM, eltech said: So, for purists, or the anxious, 192 or 384 kHz recordings should be the natural choice. They're open source and address the theoretical issues of the anxious audiophile, whereas mqa still leaves room for anxieties - the loosing of bit depth, the lossy encoding of ultra high frequencies, and the leaky digital filter. But for most people who can't hear a difference none of this matters, well except for if mqa is adopted as an industry standard. If that is the case, an open source format would be replaced with a format that lines the pockets of mqa. Yes.... 23 hours ago, eltech said: The potential for mqa to monopolize the audio distribution market and make massive profits is at odds with the open source concept. It's a form of rentseeking. Mqa really is about DRM. ...Yes!.... 23 hours ago, eltech said: When discussing intellectual property in terms of the recorded music industry we like to think it's about protecting the artists income, but they only get a pittance. It's terrible. ...1000x yes. Sorry @legend and other, Bob Stuart and co haven't convinced me and won't get a dime from me. Not least as it's extremely possible to decouple the technical and business objectives of what they're trying to do in a way that meets a broadly better, fairer intent. Truly rent-seeking of the worst kind. Great music material needs better access and lower barriers, not a pimp taking a dime at all ends of every transaction. Link to comment Share on other sites More sharing options...
Recommended Posts