Ability to play 24bit/96 files (like the competition: slimdevices transporter)
Page 23 / 41
As a (chartered) electronics engineer I agree with Keith's view! I thought about what you said, and while "utter rubbish" may be a bit heavy, nevertheless it's a fact! Why not just concentrate on enjoying the wonderful music experience that the Sonos system facilitates?
How does that work then?
I read that a while back, I still don't see how something that's inaudible can be detrimental, particularly as the people that hold that article up are usually the same people that will say that frequencies above 20kHz can have no benefit to audio reproduction, because they're inaudible. If they can't have any benefit then they can't be detrimental either.
Distortion from the ultrasonics appearing in the audible range, as explained at the top of the section I linked.
Neither audio transducers nor power amplifiers are free of distortion, and distortion tends to increase rapidly at the lowest and highest frequencies. If the same transducer reproduces ultrasonics along with audible content, any nonlinearity will shift some of the ultrasonic content down into the audible range as an uncontrolled spray of intermodulation distortion products covering the entire audible spectrum. Nonlinearity in a power amplifier will produce the same effect. The effect is very slight, but listening tests have confirmed that both effects can be audible.
They are inaudible to humans, but when electronics try to reproduce frequencies they are not meant to reproduce, it causes intermodulation distortion in the audible spectrum, which can be heard, and is detrimental to the sound. In layman's terms, straining the electronics in order to play ultrasonic noise causes distortion you can hear.
Right, so it's not inaudible then is it?
The frequency itself is inaudible, the affect on the electronics is not. In other words, you cannot hear a pure 30 kHz tone, such as a dog whistle. Blow a dog whistle, you will hear nothing. But record a dog whistle and play it back, you may then hear distortion caused by the strain on the transducers and/or amplifiers. This distortion is purely an artifact of the electronics, the recording only has the inaudible 30 kHz tone.
So that would imply that so-called "super-tweeters" that reproduce frequencies far above what we can hear could be a good thing by removing that "strain" from the tweeter's transducers right?
No, because you can't hear them. The answer is not more and more esoteric snakeoil, the answer is not to sample frequencies that 500 years of hard science has proven nobody can hear. Golden ears do not exist. Period.
Firstly it's not just the speakers: it's every electrical/electronics component in the system. That means your amplifier as well.
I guess it would be possible to make an audio amplifier which handled ultrasonic noise with less intermodulation than conventional amplifiers, but it would be a lot of effort and you would end up with a very expensive piece of equipment. As far as I know no one makes such a beast.
Secondly, even if you can totally remove the audible artefacts that could be caused by these ultrasonic frequencies, the fact that the amp and speakers has to handle them in the first place puts an additional load on the system: you are burning off some of your system's power on something that, at best, you will never hear.
Ultimately what would be the point? The problem is artificial and is self-created (by using hires files). It seems a bit daft to create such a problem and then spend time and money trying to find Engineering solutions to cure it.
The best way to reduce the impact of ultrasonic noise is to filter it out. A good way to do this is to downsample to 16/44.1 or 16/48.
Cheers,
Keith
Are there other effect (maybe implication is a better word) to the sampling rate?
(Not an electrical engineer here, mostly curious)
I am thinking that higher sampling rate would give you more "interpolation" points in between the discrete therefore possibly a better resolution overall resolution. But maybe it is not a factor or it is not worth the extra points.
Thanks
(Not an electrical engineer here, mostly curious)
I am thinking that higher sampling rate would give you more "interpolation" points in between the discrete therefore possibly a better resolution overall resolution. But maybe it is not a factor or it is not worth the extra points.
Thanks
(Not an electrical engineer here, mostly curious)
I am thinking that higher sampling rate would give you more "interpolation" points in between the discrete therefore possibly a better resolution overall resolution. But maybe it is not a factor or it is not worth the extra points.
Thanks
Mathematically, there is no need for a "more interpolation points" in order to recreate the entire original signal. Per the Nyquist–Shannon sampling theorem, a perfect reconstruction of a signal is possible when the sampling frequency is greater than twice the maximum frequency of the signal being sampled. Therefore, for all audible frequencies (up to 22 kHz), a 44.1 kHz sampling rate is more than enough to create a perfect copy of the original audio signal.
I am thinking that higher sampling rate would give you more "interpolation" points in between the discrete therefore possibly a better resolution overall resolution. But maybe it is not a factor or it is not worth the extra points.
I'll echo what jgatie said.
Quote often when sampling is being discussed, diagrams of sine waves with "jaggies" are used. The trouble is these are highly misleading and naturally leads to the intuitive conclusion that a higher resolution is better because it will create a smoother curve, and that must be better somehow. Unfortunately intuition is wrong here.
The mechanisms and mathematics of sampling show that it is possible to recreate a perfect sine wave (the purest possible tone, representing a single frequency) with just two samples. You can increase the sampling frequency and, therefore, the resolution to give you a billion data points, and you will not make that sine wave any better.
Cheers,
Keith
The idea that two ultrasonic signals can collide with audible results is called "intermodulation". Even though nether a 40kHz or a 42kHz signal can be directly heard by humans, if they pass through a nonlinear system, a 2kHz tone will be clearly audible. This is a case where two "out of band" signals will interact and cause the creation of an "in band" product.
Further, since the 2kHz intermodulation product is in band, it will be subject to the usual distortions that create harmonics and further intermodulation products with the program material.
In the end, these two inaudible signals create a blizzard of in band, very audible problems. And, they are easily preventable by blocking the ultrasonics.
Further, since the 2kHz intermodulation product is in band, it will be subject to the usual distortions that create harmonics and further intermodulation products with the program material.
In the end, these two inaudible signals create a blizzard of in band, very audible problems. And, they are easily preventable by blocking the ultrasonics.
Actually the SACD has a CD layer that would be sourced from the same master. You can listen to either one on most universal players. (An egregious exception is the SACD of "Dark Side of the Moon" The CD layer on that was deliberately compressed to make the SACD version sound better.)
Actually the SACD has a CD layer that would be sourced from the same master. You can listen to either one on most universal players. (An egregious exception is the SACD of "Dark Side of the Moon" The CD layer on that was deliberately compressed to make the SACD version sound better.)
Read what I said: "You cannot compare a production SACD to a production CD." Meaning you cannot compare a SACD to it's regular CD version. I said nothing about comparing the SACD layer to the CD layer on the same SACD, for I have no data on whether they are the same mix or not. However, I do know for a fact there are SACD versions which are mastered differently than the standard CD version.
I am one of those who would prefer that Sonos offer high rez. The reason is that many who record or master digital music claim that it sounds better, and their opinions should not be discounted. I would like to find out for myself.
That being said, if you really want to improve sound, there are other things that will make much more dramatic improvements--improvements that you won't have to ask yourself if they are really there. First, buy the best speakers you can afford and take the time to set them up properly. Second, control vibration and EMI/RFI as best you can. Third, and IMO most important, treat your room. Strategically placed bass traps and other absorbers/diffusers made more of an improvement in my listening experience than any other change I have made over the years.
Finally, thank you Sonos for providing an incredibly good product that has given me great enjoyment over the past few years. And please give us hi rez.
That being said, if you really want to improve sound, there are other things that will make much more dramatic improvements--improvements that you won't have to ask yourself if they are really there. First, buy the best speakers you can afford and take the time to set them up properly. Second, control vibration and EMI/RFI as best you can. Third, and IMO most important, treat your room. Strategically placed bass traps and other absorbers/diffusers made more of an improvement in my listening experience than any other change I have made over the years.
Finally, thank you Sonos for providing an incredibly good product that has given me great enjoyment over the past few years. And please give us hi rez.
That being said, if you really want to improve sound, there are other things that will make much more dramatic improvements--improvements that you won't have to ask yourself if they are really there. First, buy the best speakers you can afford and take the time to set them up properly. Second, control vibration and EMI/RFI as best you can. Third, and IMO most important, treat your room. Strategically placed bass traps and other absorbers/diffusers made more of an improvement in my listening experience than any other change I have made over the years.
Finally, thank you Sonos for providing an incredibly good product that has given me great enjoyment over the past few years. And please give us hi rez.
The first part of the above is spot on - I would just add speaker positioning in the room to factors that have audible impact, and I guess that is what you allude to in your reference to proper set up. Once you have addressed these, hi rez won't make an ounce of audible difference. Too many people forget to address these issues that deliver 90-95% of the quality of the heard sound in these days of reliable solid state electronics upstream of the speakers.
Unless - as has been pointed out often here - the hi rez files have superior mastering. And in which case, the better sound is solely on account of this mastering.
Sonos has already indicated that the idea is not under consideration. For the short term, get used to it and down-convert (bit perfectly) your music to a format that Sonos can play. Longer-term, you can politic Sonos for a change...
And they would be wrong. Read the relevant information posted via links in this Thread. Science trumps the mythical "Golden Ear".
Sure they should. Especially when the facts (not opinions) are decidedly against the "high-res" cohort.
Good luck. Unfortunately, it will not be by using Sonos, at least anytime soon.
Best of Luck
I am used to it and do downconvert. I am very pleased with how things sound through the Sonos. I'm just curious.
And they would be wrong. Read the relevant information posted via links in this Thread. Science trumps the mythical "Golden Ear".
You're probably right, but I would like to know for sure. Science told us when CDs first came out that they were perfect. I listened for awhile and went back to vinyl until the CD production and CD players got their acts together. When the CDs got to the point where they sounded better in my system, I digitized my albums and went 100 percent digital. My point is that scientific orthodoxy sometimes changes over time. I wish I could be put into a blind test comparing 16/24 and 24/96 so I would know for sure if there is an audible difference.
Sure they should. Especially when the facts (not opinions) are decidedly against the "high-res" cohort.
Well, they're the guys who do this for a living (making great records), so I can't know for sure. If there is some infinitesimal improvement that one would have to strain to hear, it wouldn't be worth it. As I said before, you are probably right.
Good luck. Unfortunately, it will not be by using Sonos, at least anytime soon.
Agreed. I am sticking with Sonos because listening to music through it makes me happy.
Best of Luck
Same to you.
You are confusing recording and mastering with playback. Digital recording and mastering is done by oversampling (96 kHz) and with a larger dynamic range (24bits) so that they can move all noise to the extra space (or "padding") present above the audible spectrum. This is very useful for mastering, because after applying digital filters, you simply downsample to 16/44.1 and the noise is gone. This use of oversampling means nothing when it comes to playing the files, except for the fact it has led to the audio myth that because recordings and mastering is done in 24/96, it should follow that playback should do the same. In fact, the opposite is true.
The argument for higher sample rates has nothing to do with trying to capture and play back higher frequency sounds. Anyone who discounts higher sampling rates because humans can't hear above 20kHz simply doesn't understand how these things work.
When you convert a signal from analogue to digital (sampling) you must first remove all content in the signal which is at half the sampling frequency or above. This is done with a low-pass filter. For CD, we want to preserve content up to 20kHz while removing 'everything' above 22kHz. This requires an extremely steep filter. Such filters are very hard to engineer, and typically will suffer from defects such as pre-ringing and poor handling of transients.
When we convert from digital to analogue (playback), artefacts are again produced which are related to the sampling frequency. Again, a very steep low-pass filter is required and again this cannot be a perfect filter. After CD was launched, many of us quickly realised that the promise of 'perfect sound forever' had not been realised. Two of the causes were: jitter (not relevant to this discussion) and the extremely steep low-pass filters required during recording and playback.
The best engineering solution was to increase the sample rate. If we sample at 96kHz, then the low-pass filters can be much less steep.
It was easy to migrate the recording studios to 96kHz, which solves the sampling problem. But the CD sample rate was by then too well established, so we still have the problem of trying to playback a signal with content up to 20kHz from a digital format whose sampling rate is only just high enough in a perfect world with perfect filters.
Let me put it another way - when the engineers were deciding on the sample rate for CD, the thinking went something like this:
Well we know we can't go below about 40kHz, because sampling theory tells us that would be flat out impossible. But the guys working on the lasers and the read mechs and the discs and the DAC chips are really struggling with bandwidth and capacity. They say we have to find the absolute lowest sampling rate that we can get to work. Lets do some tests and see how close to 40kHz we can get and still make this work. I know we'll be stuck with some issues but we have no choice.
Its no accident that audio formats which have launched since CD have supported higher sampling rates. If the argument for 96kHz was based on capturing and playing back sounds above 20kHz, as some on here have claimed, then I don't think many would be gullible enough to buy into it.
When you convert a signal from analogue to digital (sampling) you must first remove all content in the signal which is at half the sampling frequency or above. This is done with a low-pass filter. For CD, we want to preserve content up to 20kHz while removing 'everything' above 22kHz. This requires an extremely steep filter. Such filters are very hard to engineer, and typically will suffer from defects such as pre-ringing and poor handling of transients.
When we convert from digital to analogue (playback), artefacts are again produced which are related to the sampling frequency. Again, a very steep low-pass filter is required and again this cannot be a perfect filter. After CD was launched, many of us quickly realised that the promise of 'perfect sound forever' had not been realised. Two of the causes were: jitter (not relevant to this discussion) and the extremely steep low-pass filters required during recording and playback.
The best engineering solution was to increase the sample rate. If we sample at 96kHz, then the low-pass filters can be much less steep.
It was easy to migrate the recording studios to 96kHz, which solves the sampling problem. But the CD sample rate was by then too well established, so we still have the problem of trying to playback a signal with content up to 20kHz from a digital format whose sampling rate is only just high enough in a perfect world with perfect filters.
Let me put it another way - when the engineers were deciding on the sample rate for CD, the thinking went something like this:
Well we know we can't go below about 40kHz, because sampling theory tells us that would be flat out impossible. But the guys working on the lasers and the read mechs and the discs and the DAC chips are really struggling with bandwidth and capacity. They say we have to find the absolute lowest sampling rate that we can get to work. Lets do some tests and see how close to 40kHz we can get and still make this work. I know we'll be stuck with some issues but we have no choice.
Its no accident that audio formats which have launched since CD have supported higher sampling rates. If the argument for 96kHz was based on capturing and playing back sounds above 20kHz, as some on here have claimed, then I don't think many would be gullible enough to buy into it.
I will point out that noone has discounted hires just for this reason. What has happened is we have discounted that as a reason for hires being beneficial when addressing the very real and common misconception that, somehow, signals above 20kHz are audible.
These myths include:
* We don't understand much about human hearing and perception WRONG - We understand and can measure a massive amount about human hearing and perception and it's limits and thresholds
* There are people with "golden ears" who can hear these ultrasonic frequencies If they exist, they are vanishingly rare and have yet to be discovered and measured. This is in contrast with those in the audiophile community who claim to have this ability, who are 10 a penny.
* There is something about ultrasonic signals which can, somehow, influence audible signals This is true, but contrary to belief these signals are detrimental to audio, not beneficial
* CDs are a compromised format due to limitations in technology of the day WRONG - CD quality is fully capable of supporting the audible range of humans. There are some technical challenges in doing so, but these are all understood and addressable
* They use hires in recording studios because it sounds better WRONG - They use higher res in recording studios because it gives them more data space to work with. Most importantly it gives space to push noise and other artefacts which can be generated by some audio processing techniques
These myths exist and people believe in them, and it's one of the primary reasons buy hires.
When you convert a signal from analogue to digital (sampling) you must first remove all content in the signal which is at half the sampling frequency or above. This is done with a low-pass filter. For CD, we want to preserve content up to 20kHz while removing 'everything' above 22kHz. This requires an extremely steep filter. Such filters are very hard to engineer, and typically will suffer from defects such as pre-ringing and poor handling of transients.
As it is, digital filter technology is extremely good these days, and it's really not that hard to make a good digital filter.
Even if that was not the case, it could be largely solved with oversampling techniques.
Most modern ADCs and DACs internally oversample. This may have been a problem 20 years ago, but it isn't today.
http://people.xiph.org/~xiphmont/demo/neil-young.html#toc_o
Its no accident that audio formats which have launched since CD have supported higher sampling rates.
No, but it's nothing to do with Engineering of audibility. It's to do with marketing.
The industry has successfully managed to market degraded lossy audio (MP3, etc.) at the same price as lossless files. Noone will buy lossless files at a premium. They need an excuse for increasing prices, and hires is it.
If the argument for 96kHz was based on capturing and playing back sounds above 20kHz, as some on here have claimed, then I don't think many would be gullible enough to buy into it.
I think you grossly overestimate most people.
Cheers,
Keith
When you convert a signal from analogue to digital (sampling) you must first remove all content in the signal which is at half the sampling frequency or above. This is done with a low-pass filter. For CD, we want to preserve content up to 20kHz while removing 'everything' above 22kHz. This requires an extremely steep filter. Such filters are very hard to engineer, and typically will suffer from defects such as pre-ringing and poor handling of transients.
When we convert from digital to analogue (playback), artefacts are again produced which are related to the sampling frequency. Again, a very steep low-pass filter is required and again this cannot be a perfect filter. After CD was launched, many of us quickly realised that the promise of 'perfect sound forever' had not been realised. Two of the causes were: jitter (not relevant to this discussion) and the extremely steep low-pass filters required during recording and playback.
The best engineering solution was to increase the sample rate. If we sample at 96kHz, then the low-pass filters can be much less steep.
It was easy to migrate the recording studios to 96kHz, which solves the sampling problem. But the CD sample rate was by then too well established, so we still have the problem of trying to playback a signal with content up to 20kHz from a digital format whose sampling rate is only just high enough in a perfect world with perfect filters.
Everything you say is completely true . . . for recording and mastering!! Unfortunately, it does not apply to playback. Matter of fact, ultrasonics present in high sample rates can cause intermodulation in the transducers and amplifiers used for playback, making 24/96 recordings sound worse instead of better.
The are many valid and proper reasons for using higher sampling rates and bit-depth when recording and mastering, which you have stated eloquently. This is why it has become industry standard to oversample for recording and mixing. The snakeoil comes in when you try to apply these valid and proper reasons to the playback of those recordings, where they simply do not apply (or even cause distortion which is not present in a comparable 16/44.1 file).
This point about recording/mastering vs. playback has been stated many times in the thread, and was also emphasized in the included links. You would do well to educate yourself by reading them.
Let me put it another way - when the engineers were deciding on the sample rate for CD, the thinking went something like this:
Well we know we can't go below about 40kHz, because sampling theory tells us that would be flat out impossible. But the guys working on the lasers and the read mechs and the discs and the DAC chips are really struggling with bandwidth and capacity. They say we have to find the absolute lowest sampling rate that we can get to work. Lets do some tests and see how close to 40kHz we can get and still make this work. I know we'll be stuck with some issues but we have no choice.
Actually, the sample rate was chosen by Sony because it could be recorded on modified video equipment running at either 25 frames per second (PAL) or 30 frame/s (using an NTSC monochrome video recorder) and cover the 20 kHz bandwidth thought necessary to match professional analog recording equipment of the time.
Its no accident that audio formats which have launched since CD have supported higher sampling rates. If the argument for 96kHz was based on capturing and playing back sounds above 20kHz, as some on here have claimed, then I don't think many would be gullible enough to buy into it
Given this is an industry which sells $10,000 speaker cables, $2000 Cocobolo wood component feet, and green magic markers, you greatly underestimate the gullibility of the average audiophile.
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.