Francis, thanks for posting those articles.
I know I am going to be flamed for saying this on a digital forum but sometimes I wonder -
All this messing around, trickery, noise shaping, interpolation, manipulation, makes me wonder if it is easier to take care of scratch from a analog turntable vinyl recording, or the wow and flutter from a reel tape than all these manipulations ?
I know I am going to be flamed for saying this on a digital forum but sometimes I wonder -
All this messing around, trickery, noise shaping, interpolation, manipulation, makes me wonder if it is easier to take care of scratch from a analog turntable vinyl recording, or the wow and flutter from a reel tape than all these manipulations ?
I think it important to realise the fundamental duality in the systems here. At the end of the day the crucial issue is the transfer of information. In an analog system or a digital system the information we want to move is essentially the same. Shannon essentially showed the duality here, there is no difference between the information encoded in either an analog channel or a digital channel - however we use different names. Where we have an advantage with digital encoding is our ability to perform manipulations on the encoded information in a lossless fashion. Interpolation is one such lossless transformation. In a properly implemented interpolator there will be no loss of information. Indeed we should be able to implement a decimator that returns to interpolated data stream back to the identical base stream.
Attempting to dismiss these manipulations by suggesting that the implementations are flawed is a straw man argument. I might suggest that there are likely at least as many flaws in analog designs that result in at least as great a degradation of information. The huge advantage of digital, as has been noted, is that you only have to get it right once. Further - the lookup tables used are not for the multiplication operations - they merely hold the coefficients of the taps of the FIR filters. A fixed point multiplier is really a very trivial thing to implement on silicon. Sure, it is possible to make a mistake in the coefficients, and indeed there have been a number of times where I have seen digital systems upgraded with new FIR tap values. Sometimes this is more a matter of deciding on a slightly different perceptual balance rather than redressing a fundamental flaw.
Noise shaping is however more interesting. Simple dither spreads the noise equally across the band. However if we shape the dither we can create real additional resolution where it matters, at the cost of losing it where we decide it doesn't matter. But this process is pretty similar to design tradeoffs made in the analog domain. An LP or a tape system is essentially optimised to sound the best taking into account the human perception of sound. Tape speed for instance is balanced against the geometry of the magnetic material on the tape, we trade off high end extension vs low end. Actual magnetic bias and recording levels balance signal to noise and frequency dependant distortion products to yield a system that sounds as good as it can. This manipulation is no different to noise shaping in concept. It is just vastly messier to implement, as it is constrained by the physics of the materials, whereas in the digital domain the only restraint is processing power - something we have an extraordinary abundance of now.
But if you go back to Max's essay, the crucial point he tries to make, and one that is so often overlooked. It is a synergistic relationship. There are engineering tradeoffs to be made about where the burdens are placed, but every stage of the process has to be got right. Modern digital processes simply give the engineer a greater range of tools to address the issues, whereas only a short while ago such processing was a mere dream and everything had to be handled in the analog domain.
Try this. http://sound.westhost.com/cd-sacd-dvda.htm.
And the interesting part here is that both measurements and subjective listening test conducted by Stereophile show that the SACD is inferior to the antiquated CD.
Do you have any other corporate propaganda to spread?
Not that it matters now that the SACD seems to go the way of the Betamax.
And the interesting part here is that both measurements and subjective listening test conducted by Stereophile show that the SACD is inferior to the antiquated CD.
Do you have any other corporate propaganda to spread?
Not that it matters now that the SACD seems to go the way of the Betamax.
With all due respect to Rod Elliot, his sources of information are wrong on this matter.
I won't go into this any more. The readers here are perfectly capable of finding credible information on the matter if they are truly interested in it. I'm sure that they aren't interesting in being subject yet another debate concerning DSD vs. PCM.
Konnichiwa,
The second statement (SACD has 120 dB SNR at 20 kHz) is patently untrue and repeatedly proven to be. Please observe the following reviews measurements:
Stereophile - Sony SCD-XA9000ES SACD player
SACD S/N @ 10KHz = -108db noisefloor
Stereophile - Krell SACD Standard multichannel SACD player
SACD S/N @ 10KHz = -103db noisefloor
Stereophile - Musical Fidelity Tri-Vista SACD player
SACD S/N @ 10KHz = -108db noisefloor
Stereophile - dCS Verdi SACD transport, Purcell D/D converter, Elgar Plus D/A converter
SACD S/N @ 10KHz = -99db noisefloor
I believe this applies TO YOU and not to your opponent. It seems you deliberly spread misinformation known to be untrue!
Sayonara
macboy said:
The second statement (SACD has 120 dB SNR at 20 kHz) is patently untrue and repeatedly proven to be. Please observe the following reviews measurements:
Stereophile - Sony SCD-XA9000ES SACD player
SACD S/N @ 10KHz = -108db noisefloor
Stereophile - Krell SACD Standard multichannel SACD player
SACD S/N @ 10KHz = -103db noisefloor
Stereophile - Musical Fidelity Tri-Vista SACD player
SACD S/N @ 10KHz = -108db noisefloor
Stereophile - dCS Verdi SACD transport, Purcell D/D converter, Elgar Plus D/A converter
SACD S/N @ 10KHz = -99db noisefloor
macboy said:
I believe this applies TO YOU and not to your opponent. It seems you deliberly spread misinformation known to be untrue!
Sayonara
I missed your edit.
This isn't about debating anything. My original post wasn't about which is better. The SACD, I'm sure, has its advantages just like the in many ways flawed vinyl. I made a point regarding DACs. It's you that try to create a debate, and an artificial one at that, by spitting facts in the face.
As stated earlier, this "debate" won't exist six months from now, now that Sony reportedly has given up on the SACD format. And it seems like the DVD-A is soon to follow. At least if I read this right, http://www.stereophile.com/news/013105dualdisc/.
This isn't about debating anything. My original post wasn't about which is better. The SACD, I'm sure, has its advantages just like the in many ways flawed vinyl. I made a point regarding DACs. It's you that try to create a debate, and an artificial one at that, by spitting facts in the face.
As stated earlier, this "debate" won't exist six months from now, now that Sony reportedly has given up on the SACD format. And it seems like the DVD-A is soon to follow. At least if I read this right, http://www.stereophile.com/news/013105dualdisc/.
With the fear of this thread derailing into a Analog-Digital (and also a SACD/DVD-A) war...although it'd be fun if we get educated about these in the process.
Maybe modern higher resolution formats may change the way I think about this, but so far from what I have seen/heard/read about digital, it really does not contribute 1-bit (pun intended
) towards improving sound quality from older analog formats. Wait, I will re-phrase, I'd say it gains in some areas(bandwidth, noise, longevity, convenience, portability) but looses in others(creating the exact analog image of the original analog signal). This disadvantage IMHO is not justified by its advantages. (at this point read the first 13 words of this paragraph again to calm ruffled feathers)
Sometimes I wonder if analog and vaccum tube had progressed today to a stage where a discman would mean a vinyl running through vacuum tubes off of two AA battries then would anyone buy a cd ?
I am ignorant about sacd and dvd-a so I can't comment on that.
Maybe modern higher resolution formats may change the way I think about this, but so far from what I have seen/heard/read about digital, it really does not contribute 1-bit (pun intended
) towards improving sound quality from older analog formats. Wait, I will re-phrase, I'd say it gains in some areas(bandwidth, noise, longevity, convenience, portability) but looses in others(creating the exact analog image of the original analog signal). This disadvantage IMHO is not justified by its advantages. (at this point read the first 13 words of this paragraph again to calm ruffled feathers)Sometimes I wonder if analog and vaccum tube had progressed today to a stage where a discman would mean a vinyl running through vacuum tubes off of two AA battries then would anyone buy a cd ?
I am ignorant about sacd and dvd-a so I can't comment on that.
OK, so the noise floor in these cases isn't as good as I said it was, but it's still a heck of lot better than phn implied it would be ("inferior to CD").
In any case, the noise floor is more than good enough. I don't care much about some stupid numbers anyway; I've listened to CD vs. SACD comparisions and the difference is far from subtle. The upper few octaves of SACD are immeasurably better sounding than CD. To my ears SACD sounds great. It's a great medium for audio reproduction, a fact which isn't diminished by the fact that some other medium measures better at ultra-sonic frequencies.
This article doesn't include any statement from Sony saying that they've abondoned anything. The author's opinion that Sony is "apparently abandoning SACD while pushing DualDisc" doesn't hold much water with me, since it seems to be based only on the fact that Sony BMG is releasing a few pop artists in this new dual format disc. I'd hardly call that "pushing". Sony has been making dual format CD/SACD discs for quite some time now, and continiues to do so today.
Of course I am as interested as you to see how things will actually play out. I've noticed that DVD-A is stronger in Pop , while SACD is stronger in Jazz and Classical. This announcement seems to follow that trend.
The only way I can interpret "a heck of a lot" here is that I somehow implied that the SACD was bad. I did not.
Well, as already stated, even Stereophile's subjective panel was in favor of the CD. Not that it means anything. After all, they can all be wrong.
It's obvious to me that you have already made up your mind.
I couldn't say if Sony is abandoning its format or not. Only Sony knows. Nor do I really care. But one thing I have learnt is that you often have to read between the lines when it comes to Stereophile. That's why I included, "At least if I read this right." But I can't for my life understand why you continue to chase ghosts. That is, creating a debate where there is none.
Konnichiwa,
Hmmm. Good enough is a relative term.
What would worry me much more is an excessive amount of supersopnic noise. In two of the reviews referenced earlier (dCS & Musical Fidelity) extended plots at "digital silence" are shown. They illustrate around -50db noise around 100KHz for Musical Fidelity and -40db around 100KHz for dCS.
You quoted them, why bother if they don't matter?
Hmmm. I have both very good CD and SACD Players here, I have a few cases where I have a "known good" CD and the matching SACD. I observe that SACD sounds marginally better than the best CD in the midrange, MUCH WORSE in the top octaves. Of course, I use really good CD Replay and I make sure that when i compare SACD to CD usually not to use the CD layer on Hybrid disks, which is in many cases severely crippled sonically and sounds much worse than the original CD as well.
I find that once I remove any pre-applied bias for SACD and ensure the playing field is level that there is not that much to choose between CD at it's best and what SACD offers. I have one SACD where the CD Layer is actually HDCD encoded (Best of Roxy Music). Comparing SACD to HDCD, level matched and all, leaves no doubt in my mind that the HDCD encoded version is significantly superior to SACD.
Comparing SACD to real HiRez (96/24 PCM) shows just how poor SACD really is, compared to "state of the art" PCM Digital.
Sayonara
macboy said:
Hmmm. Good enough is a relative term.
What would worry me much more is an excessive amount of supersopnic noise. In two of the reviews referenced earlier (dCS & Musical Fidelity) extended plots at "digital silence" are shown. They illustrate around -50db noise around 100KHz for Musical Fidelity and -40db around 100KHz for dCS.
macboy said:
You quoted them, why bother if they don't matter?
macboy said:
Hmmm. I have both very good CD and SACD Players here, I have a few cases where I have a "known good" CD and the matching SACD. I observe that SACD sounds marginally better than the best CD in the midrange, MUCH WORSE in the top octaves. Of course, I use really good CD Replay and I make sure that when i compare SACD to CD usually not to use the CD layer on Hybrid disks, which is in many cases severely crippled sonically and sounds much worse than the original CD as well.
I find that once I remove any pre-applied bias for SACD and ensure the playing field is level that there is not that much to choose between CD at it's best and what SACD offers. I have one SACD where the CD Layer is actually HDCD encoded (Best of Roxy Music). Comparing SACD to HDCD, level matched and all, leaves no doubt in my mind that the HDCD encoded version is significantly superior to SACD.
Comparing SACD to real HiRez (96/24 PCM) shows just how poor SACD really is, compared to "state of the art" PCM Digital.
Sayonara
Of course I have. I said I can hear a difference, and I like what I hear. That's enough to convince me. This isn't based on technical merits or any other factor, including opinions expressed by anybody at audio-snob-magazine-x or -y. I'll trust my ears over anyone else's. I hope others here do the same.
I'm done here.
Probably partly my fault that the spectre of analog vs digital crept in. Such is life.
But back on track. SACD is of course in some senses the ultimate in single bit encoding. Sony's claims for the format are certainly dramatic, but also contain a few weasel words and are a bit open to interpretation.
First up however, it should come as no surprise that the measured performance of a SACD implementation might fall below the 120 dB spec, since we must include the vagaries of the analog parts of the system. 120dB is a pretty tall ask. Seriously, if the player had the standard 1v p-p output, -120dB noise is the thermal noise from a 12k Ohm resistor at room temperature. Worrying that the SACD format is flawed because no player can meet this level of noise is a waste of time.
But to the intrinsic performance of the encoding. First up we should all agree that the conventional CD has an intrinsic dynamic range of 98dB. The formula is 1.7dB + 6 dB per bit. This is simple stuff. Now we have already looked at noise shaping. The interesting result is that the encoding of the information onto a CD is able to move some of the noise about, and actually create a situation whereby the dynamic range at one point in the passband is greater than 98dB, but at the cost of a lower range somewhere else. At the end of the day we must conserve the information content of the channel. This is simply Shannon again. The SACD is no different. We could, in principle, place a 10Hz reconstruction filter on the output, the same on the encoder, and gain a dynamic range that defies comprehension.
The SACD format however intrinsically does contain four times the information of CD. The sample rate is 2.8MHz, so the bit rate is also 2.8MHz, compared to the effective bit rate of CD of 0.7MHz. In principle, there is enough information in the channel to get SACD to have a dynamic range of 386dB, IF the sample rate were still 44.1kHz. But it isn't. By itself the encoding would be 8dB and a bandwidth of 1.4MHz. But of course in the encoding they perform noise shaping, and they push the noise out of the audio passband. However it should be noted that they do not do this evenly, there is still an emphasis on improving resolution at lower frequencies vs the higher. So the final answer is a little vague. We don't actually know where the upper limit on the sampled analog stream is set, although they do talk about frequency response up to 100kHz. We could assume it is set about here. Curiously that is about 4 times the upper limit for ordinary CD, so there would actually be enough information to create a conventional 16 bit sampled stream (i.e. 98dB S/N) with the same bandwidth (100kHz.). That rather outlines the tradeoff. There is trivially enough information in the channel to achieve the claimed performance. The exact S/N at a given frequency depends upon the precise nature of the noise shaping performed.
On the other hand, compare it to 96/24. The information rate is pretty close (2.3M bits/s). With no noise shaping at all we might hope for 146 dB of S/N. But a pass band of only 50kHz roughly. In reality these numbers as so vastly far away from any realisable technology, they must be regarded as unattainably good.
But back to SACD. The cute trick with SACD is that it is intrinsically just the output of the sigma delta encoder - and it is possible to mimic the nature of any AD converter using the data stream. A suitable FIR filter can be created that essentially takes the SACD stream and yields 96/24, 44.1/16, or whatever. And in principle it is exactly the same as having simply built an encoder of that design at the start.
Where does this leave us? Really I'm trying to emphasise that the nature of encoding is one where a huge range of flexibility is available - and done right the only real limit is the information transfer rate of the channel. I don't doubt that there are very real differences in the perceived quality of individual instantiations of different format players. But to blame the intrinsic nature of the channel - i.e. whether it be CD, SACD, or whatever is likely picking on the wrong culprit.
But back on track. SACD is of course in some senses the ultimate in single bit encoding. Sony's claims for the format are certainly dramatic, but also contain a few weasel words and are a bit open to interpretation.
First up however, it should come as no surprise that the measured performance of a SACD implementation might fall below the 120 dB spec, since we must include the vagaries of the analog parts of the system. 120dB is a pretty tall ask. Seriously, if the player had the standard 1v p-p output, -120dB noise is the thermal noise from a 12k Ohm resistor at room temperature. Worrying that the SACD format is flawed because no player can meet this level of noise is a waste of time.
But to the intrinsic performance of the encoding. First up we should all agree that the conventional CD has an intrinsic dynamic range of 98dB. The formula is 1.7dB + 6 dB per bit. This is simple stuff. Now we have already looked at noise shaping. The interesting result is that the encoding of the information onto a CD is able to move some of the noise about, and actually create a situation whereby the dynamic range at one point in the passband is greater than 98dB, but at the cost of a lower range somewhere else. At the end of the day we must conserve the information content of the channel. This is simply Shannon again. The SACD is no different. We could, in principle, place a 10Hz reconstruction filter on the output, the same on the encoder, and gain a dynamic range that defies comprehension.
The SACD format however intrinsically does contain four times the information of CD. The sample rate is 2.8MHz, so the bit rate is also 2.8MHz, compared to the effective bit rate of CD of 0.7MHz. In principle, there is enough information in the channel to get SACD to have a dynamic range of 386dB, IF the sample rate were still 44.1kHz. But it isn't. By itself the encoding would be 8dB and a bandwidth of 1.4MHz. But of course in the encoding they perform noise shaping, and they push the noise out of the audio passband. However it should be noted that they do not do this evenly, there is still an emphasis on improving resolution at lower frequencies vs the higher. So the final answer is a little vague. We don't actually know where the upper limit on the sampled analog stream is set, although they do talk about frequency response up to 100kHz. We could assume it is set about here. Curiously that is about 4 times the upper limit for ordinary CD, so there would actually be enough information to create a conventional 16 bit sampled stream (i.e. 98dB S/N) with the same bandwidth (100kHz.). That rather outlines the tradeoff. There is trivially enough information in the channel to achieve the claimed performance. The exact S/N at a given frequency depends upon the precise nature of the noise shaping performed.
On the other hand, compare it to 96/24. The information rate is pretty close (2.3M bits/s). With no noise shaping at all we might hope for 146 dB of S/N. But a pass band of only 50kHz roughly. In reality these numbers as so vastly far away from any realisable technology, they must be regarded as unattainably good.
But back to SACD. The cute trick with SACD is that it is intrinsically just the output of the sigma delta encoder - and it is possible to mimic the nature of any AD converter using the data stream. A suitable FIR filter can be created that essentially takes the SACD stream and yields 96/24, 44.1/16, or whatever. And in principle it is exactly the same as having simply built an encoder of that design at the start.
Where does this leave us? Really I'm trying to emphasise that the nature of encoding is one where a huge range of flexibility is available - and done right the only real limit is the information transfer rate of the channel. I don't doubt that there are very real differences in the perceived quality of individual instantiations of different format players. But to blame the intrinsic nature of the channel - i.e. whether it be CD, SACD, or whatever is likely picking on the wrong culprit.
Konnichiwa,
BUT, the typhical digital player does not have 1V P-P output but 5.6V P-P which is 15db higher.
The worrying thing is not the fact that the 120db may be marketing speak, but that DSD decoded outputs around -40....-50db steady state noise at around 100KHz, even with digital silence. This places a severe burden on the analogue section and following Amplification. At 100KHz non-linearities in conventional amplification may very well be more than sufficient to create broadband noise that folds back into the audible range.
Moreover, there are some highly abstract and technical arguments about the amount of dither essential to make the DSD Modulator work as it is claimed, which would be such that the dither would be in effect equal to full scale, taking up the channels bandwidth entierly (BTW, the same applies to other 1-Bit Systems) and leaving no space for the signal.
Yes, but should be sure to make clear that this figure is not the same dynamic range that would result from a typhical analogue S/N Measurement nor is it USABLE dynamic range. It a simple theoretical and meaningless number. In reality we are closer to around 90db usable dynamic range without headroom & footroom, or about the same as good analogue tape at 30ips, except bandwidth is narrower.
So, we should be carefull not confuse such numbers as you state with reality.
Yup. We can push resolution at low frequencies if we throw it out at high frequencies. Just as SACD does.
Hmm, that is a claim often made but does not account for the presence of (neccesary) dither in the format. The dither noise takes up some of the information space.
EXACTLY. Now we shift a lot of dither into the range between 5KHz and 1.4MHz returning to 0db dynamic range at 1.4MHz and hope to be able to filter out enough of this added noise to not hurt the signal. Except it does not work well enough at 2.8MHz.
Nope, they try and fail miserably, as a simple spectral plot of digital silence of PCM'ed DSD shows....
The reality appears to be however that if you convert DSD to PCM there is no material benefit converting to any higher than 88.2KHz sample rate. While there are plenty of Bits available after the conversion there does not appear to be anything like the resolution of a native double speed 18Bit recording.
Yes, os we have NATIVE theoretical abilities for the formats of 8db over a 1.4MHz bandwidth for DSD and of 146db over a 48KHz bandwidth for 96/24 PCM.
Yes, this was the main reason for Sony & Philips to choose DSD as their archival format. However, as said, in order to make this work in reality an excessive amount of noise if introduced in the ultrasonic range, which cannot be filtered out without affecting the signal (remember the 17th order analoge LPF's in early CD Players?). One may argue as to the audible results of this (the same problems are faced equally by Delta Sigma DAC's and ADC's), listening tests do tend to suggest that they do create problems.
The key issue is that information theory is information theory and that to describe the human auditory system according to it yields inconsistent results as the human auditory system includes many complex nonlinear mechanisms that lead to rather surprising results as to what is audible and what is not.
The same can be said of 96/24 PCM. It is easy to generate a DSD datastream from 96/24 PCM.
The same holds true of any Digital System.
That remains debatable. DSD encoding and the involved noiseshaping while not violating information theory do appear to significantly interfere with human perception in at the very least a significant minority of listeners.
Sayonara
Francis_Vaughan said:
BUT, the typhical digital player does not have 1V P-P output but 5.6V P-P which is 15db higher.
Francis_Vaughan said:
The worrying thing is not the fact that the 120db may be marketing speak, but that DSD decoded outputs around -40....-50db steady state noise at around 100KHz, even with digital silence. This places a severe burden on the analogue section and following Amplification. At 100KHz non-linearities in conventional amplification may very well be more than sufficient to create broadband noise that folds back into the audible range.
Moreover, there are some highly abstract and technical arguments about the amount of dither essential to make the DSD Modulator work as it is claimed, which would be such that the dither would be in effect equal to full scale, taking up the channels bandwidth entierly (BTW, the same applies to other 1-Bit Systems) and leaving no space for the signal.
Francis_Vaughan said:
Yes, but should be sure to make clear that this figure is not the same dynamic range that would result from a typhical analogue S/N Measurement nor is it USABLE dynamic range. It a simple theoretical and meaningless number. In reality we are closer to around 90db usable dynamic range without headroom & footroom, or about the same as good analogue tape at 30ips, except bandwidth is narrower.
So, we should be carefull not confuse such numbers as you state with reality.
Francis_Vaughan said:
Yup. We can push resolution at low frequencies if we throw it out at high frequencies. Just as SACD does.
Francis_Vaughan said:
Hmm, that is a claim often made but does not account for the presence of (neccesary) dither in the format. The dither noise takes up some of the information space.
Francis_Vaughan said:
EXACTLY. Now we shift a lot of dither into the range between 5KHz and 1.4MHz returning to 0db dynamic range at 1.4MHz and hope to be able to filter out enough of this added noise to not hurt the signal. Except it does not work well enough at 2.8MHz.
Francis_Vaughan said:
Nope, they try and fail miserably, as a simple spectral plot of digital silence of PCM'ed DSD shows....
Francis_Vaughan said:
The reality appears to be however that if you convert DSD to PCM there is no material benefit converting to any higher than 88.2KHz sample rate. While there are plenty of Bits available after the conversion there does not appear to be anything like the resolution of a native double speed 18Bit recording.
Francis_Vaughan said:
Yes, os we have NATIVE theoretical abilities for the formats of 8db over a 1.4MHz bandwidth for DSD and of 146db over a 48KHz bandwidth for 96/24 PCM.
Francis_Vaughan said:
Yes, this was the main reason for Sony & Philips to choose DSD as their archival format. However, as said, in order to make this work in reality an excessive amount of noise if introduced in the ultrasonic range, which cannot be filtered out without affecting the signal (remember the 17th order analoge LPF's in early CD Players?). One may argue as to the audible results of this (the same problems are faced equally by Delta Sigma DAC's and ADC's), listening tests do tend to suggest that they do create problems.
The key issue is that information theory is information theory and that to describe the human auditory system according to it yields inconsistent results as the human auditory system includes many complex nonlinear mechanisms that lead to rather surprising results as to what is audible and what is not.
Francis_Vaughan said:
The same can be said of 96/24 PCM. It is easy to generate a DSD datastream from 96/24 PCM.
Francis_Vaughan said:
The same holds true of any Digital System.
Francis_Vaughan said:
That remains debatable. DSD encoding and the involved noiseshaping while not violating information theory do appear to significantly interfere with human perception in at the very least a significant minority of listeners.
Sayonara
Ever heard of current integration?Kuei Yang Wang said:
Current-output DACs switch constant current. 50% duty cycle of constant current on true integrator will return to no midway between Vss and Vdd.
*Must* be considered case unproven??
Do you actually understand how delta-sigma works? I suspect you don't.
Again, I stress that there is more than 1 way to do 1-bit DAC. Can you accept that? Low-end buzzers are not all there is to 1-bit.
Well, apart from your arrogance you fail to illustrate that:
A) a DAC actually CAN'T work according to said principle
B) that for each doubling of DS sample rate you gain just 1 bit of resolution.
You claim A) and B) without any material references or derived math.
Yes, of course FR is linear within audio band. What was I thinking.. The effect occurs elsewhere.
Though dynamic range and resolution *is* dropping with rising audio frequency. Rising noise floor is a consequence, not the reason of resolution dropping.
Delta-sigma is not dithering, and its noise shaping is not addition of some colored noise. Any oversampling is noise shaping, btw. Delta modulation itself affects the already existing noise spectrum, the only reason its called noise shaping. Noise shaping is not dither there.
No, you miss the point. Dithered multibit is completely different game. That can be used besides noise shaping in delta-sigma too.
OS is much more difficult with multibit - you loose precision fast. Ever checked DAC specs? It always gets worse as you increase sample rate, not better. With 1-bit its opposite.
"Resolution enhancing trickery of all sorts" includes laser-trimmed R2R dacs, for eg. You can only go that far with this.
Noise shaping is unseparable intrinsic property of delta-sigma principle and is no worse trickery than any other. It has its limits and side effects, true. So does any other trickery.
Don't get me wrong, I'm not defending DS to support sacd, at all, I'm just trying to tell that 1-bit DAC is by far not as bad as you try to make it look. You do NOT need gigaherts bitrates, ever.
Though by many reasons I also do not like what SACD does and would prefer multibit any time, you just can't dismiss the idea behind it as a lie without actually understanding it.
I really think you hardly understand how delta-sigma works. Many people don't. Thats the crux of this 1-bit issue.
Anyway, I made a little LTSpice simulation of 2nd order delta-sigma modulator. There you can clearly see how increase of OS lowers noisefloor and increase dynamic range. going from 32xOS to 256xOS drops the noise floor by ~46db. Thats 7 bits for 8 times bitrate.
Thats my "illustration" of 2 bits for doubling OS. Also there you'll see what I meant by delta coding before dac and integration after.
Try 1024x OS there. 50MHz 1-bit dac.