|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.28.27.191
...and measurements to back it up.With some DACS claiming (and apparently getting) -140db, there was still work to be done.
About 10 years ago, high-bit converter chips began to disappear. The Burr-Brown 1704, Burr PCM63K, Analog Devices AD1862N, Ultra Analog D20400. With these gone, co. were trying to push the noise-floor down, with lesser, high-noise chips.
Stereo-2 chan., upsampling, volume control, clock oscillators -all these were forced on the Delta/Sigmas that flooded the market.
So here's what they did:
'Off-loading' -co.'s 'stripped the chip' of functions (more discrete): Auralic, EMM, Bricasti and I can imagine Ayre (and more, but these are a few).
New, ground-up R2Rs: MSB, Totaldac, Metrum, Mola Mola. Some incorporate FPGAs.
'Latch type' w/FPGAs for DSP: Playback Designs, PS Audio (and Nagra ? for a price)
Then, the AK 4490 -a new chip from the orient. Hegel's HD30 DAC uses it. Got reviewed twice by the U.K., once by 'Soundstage' -and it could be a breakthrough.
But probably not for long - DACs are red-hot right now !!
Edits: 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/05/16 04/06/16 04/06/16 04/06/16Follow Ups:
My buddy's Harmon Kardon HD 7500 still sounds great. I had one until it gave up the ghost. As others have said, the implementation matters including the analog stages.
Then, the AK 4490 -a new chip from the orient. Hegel's HD30 DAC uses it. Got reviewed twice by the U.K., once by 'Soundstage' -and it could be a breakthrough.
IMHO we don't need a "breakthrough" DAC chip. We need DAC designers who know how to 'voice' their product around any decent modern DAC chip. There are cheap DACs today that sound wonderful and very expensive ones that sound like crap. Why? This is not attributed to the DAC chip itself but what the designer does with it and in the analog stages that follow. Just my 2-cents worth. ;-)
"...and measurements to back it up."
Digital audio always measured well.....
"With some DACS claiming (and apparently getting) -140db, there was still work to be done."
That's about 50 dB better than the noise floor in a quiet room..... Or in other words, the bits running below -90 dB are all toggling randomly on noise...... So if anything, this *adds* noise.
"About 10 years ago, high-bit converter chips began to disappear. The Burr-Brown 1704, Burr PCM63K, Analog Devices AD1862N, Ultra Analog D20400. With these gone, co. were trying to push the noise-floor down, with lesser, high-noise chips."
The noise is already inaudible on paper..... And the noise we hear is almost all background noise during recording..... But it's encoded onto recording. No "quiet" DAC can fix this.
"Stereo-2 chan., upsampling, volume control, clock oscillators -all these were forced on the Delta/Sigmas that flooded the market."
OK......
"So here's what they did:
"'Off-loading' -co.'s 'stripped the chip' of functions (more discrete): Auralic, EMM, Bricasti and I can imagine Ayre (and more, but these are a few)."
You can make a chip "behave" discrete, but it isn't really discrete. Same goes for any integrated circuit.
"New, ground-up R2Rs: MSB, Totaldac, Metrum, Mola Mola. Some incorporate FPGAs."
And most are hideously expensive......
"'Latch type' w/FPGAs for DSP: Playback Designs, PS Audio (and Nagra ? for a price)"
Mostly re-inventing the wheel.....
"Then, the AK 4490 -a new chip from the orient. Hegel's HD30 DAC uses it. Got reviewed twice by the U.K., once by 'Soundstage' -and it could be a breakthrough."
If a DAC chip were to come forth that has significantly lower RFI emissions, that would be the only real "breakthrough".... But nobody is really addressing this.
"But probably not for long - DACs are red-hot right now !!"
I still like the DAC chip in my (modified) early-1990s-vintage Philips CDC-935 CD changer.
Some are saying that upsampling to double, quad and 8 times DSD via HQplayer (apparently the best sounding upsampling software) and a compatible great DAC will give sound that is better than most any other way of listening to PCM. Even the big DCS Vivaldi system (over $100K) is suppose to sound best using DSD upsampling. If you look, you will see this is the coming thing. If this is the revolution then we would not even need a DAC that would decode PCM at all. All it would have to do is decode upsampled DSD. Meitner's new DAC upsamples all formats internally to 16 times DSD! Of course, it is a mere $25K....however, there are much cheaper ways to scale the mountain of DSD upsampling.
"Upsampling" has been a design gimmick from the get-go..... Especially if it's asynchronous...... (Most synchronous upsampling is exactly the same thing as 4x/8x/16x oversampling that preceded it.) It took a while before designers came to the realization that it didn't really do anything beneficial, and the "technology" slowly faded out of existence in more recent products.
Ah, the same story. You need to get out more.This is not Asychronous nor is it high bit. You are converting all PCM directly to DSD (one bit) but at much higher frequencies than standard SACD. This pushes the noise way out there and allows more precise upsampling. The decoding is then done using a simple one bit decoder that has no other digital filtering. This is the revolution. You guys need to read what others are doing out there. They say this technology makes regular CDs sound as good as their turntables....maybe better. This is not just regular 10-15 year ago upsampling.....this is a revolution....right now the revolution requires HQplayer to do the upsampling.
"The revolution will not be televised" It will and can be heard....right now. But you have to look for it. You have to read other forums and magazines and reviews.....you will not find it on this forum yet. Uh Oh....did I start the revolution here? Have fun!......it has only just begun.
Edits: 04/07/16 04/07/16
I've been listening in DSD 128 almost exclusively since I got my Korg DS Dac 100,
on my Desktop System.
I now also have Sony HAPz1, Mytec 192/DSD Dac, and Lampiztaor Euphoria DSD- only Dac, also with HQ. ( Not the Sony, of course).
Not the multiple DSD conversion you are talking about, but love it none the less.
I had an Exasound E22 for a while but sent it back. The Coax input didn't work with my Satellite Box, only got break-up.
Also, it was a bit bright for me, and caused my ears to ring afterwards, which doesn't happen on the other Dacs or the Sony.
Naturally I'd like something even better, but cost becomes an issue,( prices from $3500-$6000, and beyond...) esp since I couldn't sell what I already have for much.
But higher-res DSD is the direction I've gone.
The word is the higher you upsample the better the sound. All upsampling sounds different. The HQplayer is suppose to sound the best of all the upsampling software (at this very moment....soon? to be superceded). Every time you double the sampling frequency your jaw drops closer to the floor. The revolution continues! 512X is the killer frequency....and now Meitner is doing 1024X inside their DAC.....whoa baby!
/
Right -Charlie Hansen has been saying this for years. Simply inserting another filter in the signal path. And you can't create something that was not recorded.And 'DSD' has been high-bit for years -16, 20 or more. So it "needs" an equivalent DAC to decode. Unless it's compressed in production - but I don't think it is.
Edits: 04/07/16 04/07/16 04/07/16
In your first post you were praising the trend of "off-loading" DSP from the DAC chip by preceding it with a FPGA or other programmable DSP. And now you're agreeing with Todd that upsampling is bogus. But these are one in the same thing. Those FPGA-based DSPs that are en vogue today are custom upsamplers.
An upsampler is just a Redbook reconstruction filter. Without it, you get whatever reconstruction filter is implemented in the DAC, which might be a generic short length 8x oversampling brick wall type. With an external DSP you can implement a more computationally intensive but possibly better reconstruction filter than what is in the DAC.
Well, I didn't 'praise' DSP. The idea was to list projects that are off-setting the (low-bit) commercial-DAC problem. In history, high-bit DACs were better than low.If you're saying that 'off-loading' DSP is no big advance, then I agree.
What I was shocked at was Ric Shultz's report of HQPlayer. This could void all the things I listed. Now a 1-bit DAC makes sense - as they might decode a 45 mega-hertz signal !!
Edits: 04/08/16 04/08/16 04/08/16 04/08/16
Off loading the DSP has been common since the beginning. The early DAC chips didn't even have an onboard digital filter.
All the top CD players 15 years ago were using fancy DSPs in front of standard DAC chips to implement the reconstruction filter. Wadia had their DigiMaster + BB 1704. Simaudio and several others were using the Pacific Microsonics PM100/200 with a BB 1704. Audio Aero were using an Anagram algorithm on a SHARC DSP with an Analog Devices DAC. And the gold standard of digital playback for most of the last decade was EMM Labs, which upsampled Redbook to DSD. So the technology hasn't been changing much. Things have been getting better via refinements to existing approaches, no revolutions.
And 45 MHz is nothing special. Most parts support MCLK of 1024fs which is 42 MHz just for 44.1k Redbook. Some ESS DACs can take over a 100 MHz clock if you disable the internal oversampling and it seems that most people working with them are using MCLK in the 80-100 MHz range. No reason you couldn't go even higher, but you just increase the RF noise on your audio board.
EMM has been the standard -but they only sample at 5.6 Mhz.
Before their (brand new) DA-2, who else sampled at 45-100Mhz ?
36.864 MHz was a common maximum MCLK frequency a decade ago, which allows operating up to 768fs for a 48 KHz PCM input, 384fs for a 96 KHz input, etc. Some current gen Cirrus and AKM DACs have maximums in the 40-50 MHz range. For example, the CS4398 can operate at 256fs at 96 and 512fs at 192, which is 49.152 MHz. ESS is the only one I know of that goes beyond that.
Assuming you want to achieve the maximum SNR in the pass band, there is a tradeoff involved with increasing the frequency multiplier. Using a higher frequency multiplier allows for better noise shaping. As you know from SACD, 64fs requires very aggressive noise shaping to achieve good performance up to 20 KHz. When operating at 256fs, the noise shaping can be a lot more gentle and push the noise further away from the audio band. But as you increase the system clock frequency, it's harder to control the analog noise from the clock and harder to control jitter. For this reason, the best Redbook performance is often achieved at 192fs or 256fs.
..ok, but DACs -capable- is one thing, actual DSP-work is another.Besides Meitner and PS Audio, I found (Antelope) Zodiac Platinum. All upsample to Quad DSD (11.2Mhz).
It's hard to find DACs that do this, let alone one that processes to 45Mhz.
Edits: 04/11/16 04/11/16
Wadia was doing 64fs with a digital filter implemented by FPGA way back in 1988! That's the same rate as DSD. Krell and some others were doing the same by 1990.
By the time SACD launched in 1999, it was common to convert 44.1 KHz PCM at 256-512fs, even with R2R ladder DACs. That's 11.3 to 22.6 MHz, 4-8x higher than DSD at the time of SACD's introduction. A popular high end combination was the Burr-Brown DF1704 filter and PCM1704 DAC, which was released in 1998 and worked at 512fs for Redbook and 256fs for 96 KHz PCM.
And a decade ago, 768fs was becoming common, that's 33.87 MHz.
As released at 64fs in the SACD format, DSD was way behind the state of the art. It was limited by the storage inefficiency of the 1-bit format and trying to fit 8 channels (5.1 plus stereo) on DVD-based media. Quad DSD would have been more in line with the state of the art circa 2000, but it would have been stereo only.
These days, quad-DSD playback is still operating at a slower clock rate than modern PCM playback, but we're at a point of diminishing returns where increasing the clock rate further doesn't necessarily improve the filtering but may increase noise.
Dave thanks for the post. I found the whole thing incredibly informative.
Does the very high sampling/oversampling rate in PCM R2R dacs allow for a low order gentle filter like DSD?
Do you think there is anything intrinsically more linear about one format versus the other? I mean in reproducing a waveform, or does it just come down to the sound of filters?
Thanks,
------------------------------------------------------------
Big speakers and little amps blew my mind!
Does the very high sampling/oversampling rate in PCM R2R dacs allow for a low order gentle filter like DSD?
No for the digital filter, yes for the analog filter.
There is a gentle low order analog filter on the output of an oversampling DAC. It can be gentle because it's not the reconstruction filter. All it's doing is filtering ultrasonic noise. With some R2R DACs, you might not even need it.
But the digital reconstruction filter will still be a brick wall or close to it. Oversampling at higher rates and with longer filters allows for a more ideal filter response. It also allows for better noise shaping. But you're still stuck trying to fit a steep cut-off between 20-22 KHz regardless of how fast you oversample. Such is the nature of Redbook unfortunately, and there is no way around using brick wall filters without accepting aliasing distortion or rolling off below 20 KHz.
Same thing if you're converting 44.1 KHz PCM to DSD, the conversion will involve a brick wall filter.
If you're sourcing DSD from DSD or analog or high sample rate PCM, then you can avoid it.
Do you think there is anything intrinsically more linear about one format versus the other? I mean in reproducing a waveform, or does it just come down to the sound of filters?
Are you comparing typical oversampled Redbook playback vs. Redbook upsampled to DSD? Or R2R vs. delta-sigma? Or more generally PCM vs. DSD?
Thanks for the response.
Hi res PCM vs DSD.
One tenth of an octave (20 to 22kHz) would require one hell of a brickwall filter. JA mentioned in another post that DSD provides an improvement over PCM in time domain errors. I suppose that could be called a form of linearity. Nelson Pass published a paper that has some graphs showing how steep filters for speakers obliterate the phase coherence of a speaker. My limited knowledge leads me to think that phase and time domain response are probably closely tied? Your opinion? I have used both gentle and extreme slopes with speakers, and neither is perfect or perfectly terrible. Just different with different tradeoffs.
Thanks again for answering my basic questions.
------------------------------------------------------
Big speakers and little amps blew my mind!
I suppose he is referring to the ringing impulse responses of the anti-aliasing and reconstruction filters, which is particularly bad for Redbook due to the need for steep filter responses. For higher sample rate PCM, 88.2 KHz and above, the filters do not need to be steep. And even if the filters are made to be steep, there isn't much musical content up at 44 KHz or above to incite ringing, and the ringing would be ultrasonic anyway. So in my opinion, it is impossible to make Redbook anti-aliasing and reconstructions filters truly transparent, but it should be easy to do that for PCM at 88.2 KHz and above.
I enjoy both DSD and hi-res PCM, but I think DSD 64 (the SACD format) is a little bit compromised relative to hi-res PCM. Going forward, I could be happy with either hi-res PCM or higher rate DSD. I think that a DSD-only DAC operating at 256fs could be superior to hi-res PCM due to simpler implementation, but at the expense of storage space and download bandwidth.
I found the following Audiostream interview with Mr. T. Of special interest was the following;
"If we convert from 24-Bit at 352.8kHz (DXD-PCM) to 1-bit at 2.822MHz (DSD) - we need to throw away around 99.96% of the amplitude information the PCM format is capable of, while we are only having 12.5% of the time domain information that the DSD system is capable of. If we convert to DSD from DXD, that is 1-bit at 2.822MHz to 24-Bit at 352.8kHz - we need to throw away 87.5% of the time domain information of DSD, though we can theoretically remap all of this into the amplitude domain. So in effect we get the worst of both formats, rather than the best of one."
Maybe higher rate DSD would even things some, but I find it interesting. Also very few albums for sale at 352kHz. I would bet Sony thinks they can remap into the amplitude domain, or they wouldn't have put it forward as an archival format, but then again maybe they would. I have read on a forum for recording engineers where they say DSD was meant to be converted to PCM, but PCM was not meant to be converted to DSD. They certainly seem to do it though.
-------------------------------------------------------------------
Big speakers and little amps blew my mind!
Thorsten's numbers are misleading because he's equating timing resolution with sample rate and amplitude resolution with sample size. I believe he knows better but is just trying to oversimplify for a non-technical audience.
Ultimately, amplitude resolution and time resolution are characteristics of the encoded waveform, not its encoding format. Amplitude and time resolution are limited by the digital noise spectra. So for example, 1-bit DSD has better amplitude resolution than 16-bit PCM, and 352.4 KHz 24-bit PCM certainly has better timing resolution than DSD (even 192 KHz PCM probably does).
Using delta-sigma modulation with appropriate noise shaping, one can trade sample size vs. sample rate without loss of resolution. That's how delta-sigma DACs work.
Very informative reading here.
nt
I didn't know this. But in recent years, co. were not running as high. Even DCS upsampling-devices were processing at DSD (2.8 Mhz).I guess these earlier rates were overkill. And so much for 'HQPlayer'- it looked like a breakthrough.
In your view, what was the biggest bottle-neck of Red Book ? The whole act of conversion ? Is Equibit the way to go ? (Still has a clock, but now controlling the output stage). Is this a better way to clock ?
Edits: 04/11/16 04/11/16
DCS is sort of an odd duck in that they use a proprietary 5-bit discrete ring DAC operating at 64fs, which is quite slow by modern standards.
As I was saying before, higher clock rates are pretty much standard fare. At the present time, 768fs (33.9 MHz) is typical for those using current CS, TI/BB, AKM, and Wolfson delta-sigma DACs. ESS is 80 MHz.
For older equipment, here is a link that has a nice summary of clock rates:
http://www.trichordresearch.co.uk/cd-player-list/
In my opinion, the biggest problem with Redbook is the sample rate is too low, which means the anti-aliasing and reconstruction filters will never be fully sonically transparent. Worse yet, the anti-aliasing and reconstruction filter responses overlap and interact because they're both occupying the same frequency space, which may be why some filters sound better with certain recordings but not others. The low sample rate guarantees that there will never be any consensus "best" reconstruction filter. As long as the format is around, I think we're doomed to go on iterating and tweaking different filter designs.
Upsampling is not a gimmick. Because you, or someone else you've read, may not have liked the sbjective sound of the upsampling players you've heard doesn't make the technique a gimmick. Upsampling/oversampling results simply from the operation of a digital reconstruction filter. Such filters have long been in standard use everywhere else in DSP.
Accurate objective recovery of the original waveform requires filtering of the output image products, which can appear strongly as low as 24KHz. Analog reconstruction filters simply not as effective as digital filters at separating a 24KHz image from the 20KHz signal which created it. None of which necessarily means that the human ear likes the sound of brickwall bandlimiting, whether analog or digital.
Another factor is that most digital reconstruction filters are implemented as what's called half-band. This means that they consume less silicon area while also, as you might imagine, being the most performance compromised FIR implementation. More of the subjective performance of digital audio seems to lay in implementation than in the requirements set forth in the sampling theorem.
_
Ken Newton
Many engineers would disagree with you.We're way-past 'upsampling'. It's the performance of the converter, voltage regulation, output stage, power supply.
More precise calculations won't hurt. But the point of this thread was to show what designers are doing, to off-set the problems of low-quality chips.
If you're right, it's only because there was a 'need' to keep re-calculating.
Edits: 04/07/16 04/07/16 04/07/16
I think you would be hardpressed to find an engineer who would disagree on objective technical grounds. Subjective sound is, however, a different matter.
_
Ken Newton
Your lack of knowledge is pretty astounding for someone positioning them selves as an internet pseudo authority.
I'm identified as an 'audiophile' - so, no 'authority'.I'm explaining what engineers are doing (or saying), that's all...
Edits: 04/07/16
Hi,
> With some DACS claiming (and apparently getting) -140db,
> there was still work to be done.
Forgive me for injecting any reality into this.
Given that the quietests microphones known to man have around 9dB(A) self noise and many recording microphones exceed 20dB(A) and given that a classical orchestra in a full tutti in the front rows rarely exceeds 105dB SPL, what is the use of very large SNR figures (I mean except for advertising)?
I would suggest that anything that betters around -110dB SNR is good enough to not represent a limitation on the basis of noise levels for real recorded music. It does no harm to have a few more dB, but I would not class it as a valid engineering goal to pursue SNR levels much past 110dB.
The issues are elsewhere, in the digital filter charateristics, the noiseshapers used (generally i find higher order noise shapers not very nice) and supersonic noise signature of DAC's.
> About 10 years ago, the Burr-Brown 1704 (converter-chip)
> disappeared.
About a year ago it remained in production, though it now has a pricetag of the "if you have to ask..." variety. It has since been discontinued because TI closed a FAB that did this sort of Job.
Lucky their hybrid DAC line was not affected. Too bad for their competitors that TI still has valid and non-expired patents on that tech. The nice thing is, with 6 Bits in multibit any noise from noiseshaping of the lower bits is pushed down 36dB at all frequencies. No Delta Sigma system can match that.
> Then, the AK 4490 - a new chip from the orient. ... -and
> it could be a breakthrough.
Don't see how.
AKM and Cirrus Logic used to share technology closely. Both their DAC lines continue to use switched capacitor systems which makes them dirt cheap but not particulary distinguished. The AK4490 is a decent DAC chip at a keen price, but not really materially better than the competition from Cirrus Logic.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Quality versions of the 1704 have been out of print since at least 2008. There was a "U" version, a "U-J" and "U-K". Reviewers even said this, in product reviews. One from 2009.But the point was high-bit DACS in general - so I updated the piece to include these.
The point about mikes is quite different from playback. If listeners can hear a difference between 105db and 140db then something must be going on.
It looks like you're back to all-theory, like your responses last year (here) on DSD. Real world, son, real world...
Edits: 04/06/16 04/06/16
I wouldn't worry about the PCM1704 because the one you really want to find is the even older PCM63, which actually sounds better despite being "only" 20bit.
I have tried back to back the Monarchy M24 and NM24, which only differ in the DAC chip and probably the I/V resistor and the M24 (with the PCM63) was significantly better sounding than the newer NM24 (with the PCM1704).
Go figure.
The only off-the-shelf DACs that I have heard that might be better than the PCM63 are the AD1865 and the UltraAnalog D24000. I have a DAC with the UltraAnalog and it is killer, particularly after I bypassed the SS output stage and coupled it to a Lampizator kit tube output stage. I don't have an AD1865 based DAC but there are some guys out there still swearing by this particular chip...especially when paralleled up!
I agree - but many said the 1704 was superior to all others. Lessloss co. was one I remember. And the new DAC from ex-Goldman people (CH Precision) uses 1704.
But, maybe the PCM63 can't be obtained. I don't know how (CH) found 1704s !
Could be the case that they are unobtainium, which is why I value my Monarchy DAC that has them...it is a very musical and bold sounding machine!
you could buy all the PCM1704U-K you wanted from Mouser, albeit about $60 each in quantities.
Gone now. :-(
I'm guessing that most of the manufacturers with products using said chip have loaded up.
Hi,
I missed answering that tidbit:
> Quality versions of the 1704 have been out of print since
> at least 2008. There was a "U" version, a "U-J" and "U-K".
K-Grade was still listed last time I saw the PCM1704 active from TI. It was of course priced at several times it's weight in gold and the PCM1704 was marked "NRND" for ages, but they kept making and selling it until they shuttered the Fab.
I was surprised to note that TI still lists the PCM56 as active as of today. Of course it's only 16 Bits...
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Hi,
> If listeners can hear a difference between 105db
> and 140db then something must be going on.
Well, there is a logical flaw here. You would first have to demonstrate that the difference heard between different DAC's i down to the different SNR's.
BTW, 105dB SNR means that if 0dBFS peaks are set to produce 105dB SPL at the listening position and if the rest of the system has no noise (and the recording microphones had no noise), the systems self noise would be essentially 0dB or at the limit of audibility.
By demanding at least 110dB SNR we are already past that.
Anyway, my point was that if you hear differences between a > 110dB SNR DAC and one with (say) 140dB, you are not hearing the differences in SNR.
As to theory and practice, don't worry, I practice. A lot...
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
I don't disagree, offhand.But with the Benchmark amp (using a British-designed feed-forward circuit), they're claiming -130db. Or "20db quieter than reference amps".
I, for one, can hear less noise. And there are 5 reviews of this unit that supports my view.
Edits: 04/06/16 04/06/16
Are you referring to the noise floor when no music is present?
ET
...when music is present. The only way, as reviewers have noted...
Hi,
> ...when music is present. The only way, as reviewers have noted...
Then what you hear is not the "self noise", but the interaction of the noise contained in the recording (e.g. the hall noise, microphone noise, recording electronic noise etc.) with the processing in the devices used for replay (see earlier posts).
Given the complexity of that, all bets are off...
TL:DR
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Hi,> I, for one, can hear less noise. And there are 5 reviews of this
> unit that supports my view.If you cannot hear the noise at all (because it is so low that it is inaudible) how can you hear even less noise?
Again, you may be mistaking other effects.
What you would need to do, in effect, is to find a way to add the noise of your old amp's back to the new Amp so you can switch the addition on and off and see if you actually can hear the difference.
I would also note that specific amp under debate has comparably low gain (which in my view is a good idea BTW) and so in a complete system with a preamp would quite substantially lower system-wide noise.
FWIW, I am using a Tube Amp with around 96dB SNR at 2.83V (measured on AP2) and speakers with 90dB/2.83V sensitivity. The Amp also has intentionally fairly low gain and a similar spectrum but much lower levels of HD than the speakers have at 2.83V input. The DAC that drives this Amplifier has 114dB(A) SNR. And yup, this DAC's is also inaudible with the volume set to max.
Even sticking my ear into the speakers I hear no noise whatsoever, no hum, no hiss, nothing. Lowering the noise of this amp would pointless, unless I switch to much higher sensitivity speakers. The HD is also while not very low, lower than most speakers I have ever come across at 90dB SPL and thus also inaudible.
But I am sure if I switched it for a 112dB SNR @ 2.83V Amplifier that is all transistor I would hear a difference. Yet I believe you would agree that it would be foolish to suggest that the difference I would be hearing would be explained by the 16dB difference in SNR (or the difference in HD FWIW), given that no noise is audible regardless which amplifier is in use.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Edits: 04/06/16
> "no noise is audible"...
Well, if it's not noise, what else is it ??!!
Hi,
> Well, if it's not noise, what else is it ??!!
Here is the crux of the matter. Actually, head over to Audiostream.com, it has a good article by Jason Stoddart (from that company with a name I will not take into my mouth) on this whole subjectivist/objectivist kerfuffle.
What I'd like to add, is that common audio measurements have never been shown to have any positive correlation with perceived audio quality, as long as certain minimum performance levels are met.
Indeed, there is no single instance where we have a formal test that produces single number performance indicators that - if made better - reliably produce better sound.
It is my experience that it is fairly trivial to design audio gear where the frequency response is adequately flat to leave the room as biggest problem, with noise that is so low, even super quiet acoustically treated and isolated rooms have more and HD that is reliably lower than the Quad ESL63, which at least in the mid-range is probably still the lowest distortion loudspeaker bar non.
I am also quite aware of the limitations of the human auditory system. It is so much not like a microphone or speaker, what is amazing that we can even get any credible illusion of listening to music from the current mechanisms (including recording).
It is actually a learned process, I remember reading a story where a TV crew filmed the Bushmen in Namibia. When they tried to show them the recordings on the TV monitors the Bushmen could not see the images, all they saw were colored dots. You might say they could not see the picture for the pixels. Of the story may be apocryphal.
Anyway, unlike the "everything sounds the same" brigade I however am not deluded enough to think just because Frequency response, noise and distortion are below anything that can be heard/perceived in normal or even extreme systems we have eliminated all audible differences.
Let me take an example that is brutal. If you have a 120dB/20Hz tone, you are unlikely to be able to hear around 10 - 20% 2nd harmonics, but you likely will be able to hear 0.0001% if the 100th harmonic. Now no-one measures this high harmonics and in a single number like THD the high harmonics are usually totally swamped by lower order HD.
There are many facets in audio performance which are currently not covered by generally agreed, standardised measurements.
As I said before, in my experience such aspects as the precise noise-shaping algorithms in Delta Sigma system (which in theory at least are all supra-sonic in nature), Dither noise (possibly shaped), digital filter algorithms, the precise nature and spectrum of harmonic distortion (actually the intermodulation distortion products generally are more pernicious in terms of audibility), general time-domain performance (something most "flat response"speakers totally mangle) etc. all impact more than the difference between -110dB SNR and -140dB SNR (in isolation) or the difference between 0.005% THD and 0.0025% THD.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
"What I'd like to add, is that common audio measurements have never been shown to have any positive correlation with perceived audio quality, as long as certain minimum performance levels are met."
Thorsten: Can you suggest what those minima might be in re (1) stereo separation, and (2) permissible phase shift from either phono cartridge or tape head though to speaker input? I ask because I'd claim to have heard too little of the former and/or too much of the latter more than once, reliably (subsequently confirmed by measurement). There must have been some standards suggested by Bell labs or some such, decades ago, but if so, their importance seems to have been somewhat forgotten.
Jeremy
Well, I don't know how much dither-noise we can hear, in a playback device. But you could be right. Filter algorithms seem almost impossible to hear.But -I'm glad you said 'audio tests are limited'. This, because they want them to be -or the equipment isn't there (yet), to dig deep into the human auditory-system.
It was said at a presentation at RMAF (show) last year, that measurements were created for 3 things: design criteria, production consistency and to sell audio products.
You're right-in-line with these statements...
Edits: 04/06/16 04/06/16 04/06/16 04/06/16
Hi,
> Well, I don't know how much dither-noise we can hear,
If we are talking 16 Bit Data and we use dither only in the LSB (which may not be enough to be effective) we are in effect down to 81dB "analogue audio equivalent" SNR, which is marginally better than dual half track master Tape at 15IPS, or not...
I use "analogue audio equivalent" here to indicate a measurement and calculation process that is equal to an analogue device, like say a Studer Professional Tape master machine.
I would certainly class broadband noise at -81dB down from maximum as audible in the right circumstances.
My old Sony Boodokahn Walkman (my several Pro ones too) managed noise better than -60dB with Dolby (and had 12dB headroom - so in modern parlance that would count as -72dB) and that noise was extremely intrusive on headphones - but in the analogue days that is what you got, live with it.
> Filter algorithms seem almost impossible to hear.
As I have designed no commercial digital device ever that did not offer alternative filter options, you might want to seek them out for a listen and try for yourself. I find the differences between "FIR Brickwall" and "Slow "Rolloff Apodising" gross and I have a firm favorite too (hint, neither of the above).
> It was said at a presentation at RMAF (show) last year, that
> measurements were created for 3 things: design criteria,
> production consistency and to sell audio products.
>
> You're right-in-line with these statements...
And I totally agree with the last two.
The first, I strongly disagree. If we want evidence based design goals it is useless.
If you let incomplete measurements that carry no proof whatsoever of correlating with valid design goals dictate your design - you belong into a loonie bin.
That is not engineering, that is not science. That is not even rank empiricism (which is unjustly denigrated these days).
It is what Richard Feynman called "cargo cult science".
Except the guys who practiced cargo cults (or those who read the future out the entrails, coffee ground or stars) can be excused on the principle they had no way of knowing better...
With what we now call the scientific method dating back to ancient Greece and most people these days being not only literate (it helps) but also schooled in the scientific method and critical thinking (Southern USA excluded), never mind people with higher education, there is no more justification of the worship of the "THD Meter Needle" (or "SNR meter Needle", AP2, Prism D-Scope etc. et al.) than there is for caring about the "E-Meter needle" (SciTo).
Or possibly much less.
The E-Meter is a primitive lie detector that reacts reliably to skin resistance which together with the so-called "Audit Process" is in fact a pretty reliable detector if you are getting close to stuff that matters to the subject, unless (s)he is a pathological liar.
That does not make SciTo good science (on the contrary), but it does put THD Meter Needle worshippers (and their related ilk) into a place where even the E-Meter readers go - ROTFLMFAO...
A "measurement" gives a precise answer to the precise question. In fact, the more precise your question is, the more useful the resulting measurement (that is paraphrasing Shannon/Weaver's definition of information).
Precision is of course limited by the test gear. Sometimes even a 15K AP2 is "not good enough", some times a top of the line LeCroy 5GHz Analyser (think a nice house in a nice 'hood, like Kensington) is not good enough.
But if you formulate the question right, the answer will precise within the limitations of the test gear. And the answer will make sense. If you get it wrong, the answer will still be entierly accurate, but it will make no sense whatsoever.
It matter Jack how precise or true the answer is, it is not even the wrong answer, it is what your teacher would mark your essay as "failed, missed topic", unless you got the rights question first.
Now if you are like, totally out of the ballpark, totally off the reservation, way beyond the pale... Should you use THAT as the basis of setting design goals?
Elsewhere (PM me - I'll send you the link) I expounded my time meeting a senior designer of DAC and ADC chips for one of the market leaders. We met in the queue at the immigration in Peking in 2008. He was even a bit (not much) of an audiophile...
But when we talked business we were in different galaxies.
He described how he would get his bonus if he designed a chip that measured slightly better than the competition and cost slightly less (I like a fat Bonus, share options etc as much as the next guy).
At the time he was really happy, because his company's product (and his design) had been chose by Apple to replace British Wolfson Micro in all Apple products. we are talking about a huge sale.
When I asked him (as a semi audiophile with a fairly serious system and even a degree of love for music) if he ever evaluated his ADC/DAC chips by listening - I got an even blanker stare than I got from that working girl in Amsterdam when I asked her if hopping on the good foot and doing the bad thing did anything for her...
So, bottom line. Failed.
An "AA+"for effort, a "FF-"for achievement.
Me? I don't give a pair of fetid dingo kidneys about measurements, except as a way to confirm that things work the way I intended them too, but I'd never use them as predictor of quality.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
nt
I meant noise generated (by) the algorithms...In any case, the person who broke-down what measurement/specs are good for works for a co. that makes the highest-precision audio analyzers in the industry.
Edits: 04/06/16 04/06/16
.
[...and measurements to back it up...With some DACS claiming (and apparently getting) -140db, there was still work to be done.]
Outstanding measurements haven't been a problem for digital audio pretty much from the beginning. A few remaining questionable performance parameters, such as low level linearity and jitter, were made excellent more than a decade ago. I don't believe that reducing converter quantization noise from, say, -120dB down to -140dB will subjectively improve digital replay anymore than would reducing amplifier noise by the same degree.
I, among others, do not believe we've yet identified all of the key factors/parameters responsible for the too often disappointing sound of practical digital audio implementation. I think the reason the market for DAC products remains red hot is simply because subjectively satisfying digital audio playback (especially at affordable proces) remains a work in progress. It seems clear that, ultimately, the answer will not be found along the major established parameters such sample rate, SNR, bit depth, etc.
_
Ken Newton
Agree, we don't understand how we perceive sound well enough to get to the fine details. Digital artifacting appears to be far more audible than once thought, and the way we hear transient information isn't the same way that we perceive tonality. And there is also the matter of perceving sound through the skin, too.
All weird stuff to be sure, but probably important if you want to make sound recording and reproduction as realistic as possible.
=Signature=================
As audiophiles, we take what's obsolete, make it beautiful, and keep it forever.
Hey! I have a blog now: http://mancave-stereo.blogspot.com or "like" us at https://www.facebook.com/mancave.ster
I read years ago that audio products "measured close to perfection". Yet digital clearly improved. So much for measurements.
And some would disagree with the noise drop. Benchmark's amp claims lower measured noise (than any other in history) and seems to deliver. If we believe the subjective-listening reports.
I'll take one, OK, make that 16 each Burr Brown PCM-1704U-K..
How much does this perfection cost?
No idea but it's WAY out of my price range.
That said, if I were to hit the lottery and became a multi-gazillionare that is what I would likely have as a DAC.
You do have to wonder, there is a point where you could just save yourself the hassle, and go to the symphony. Just sayin'
=Signature=================
As audiophiles, we take what's obsolete, make it beautiful, and keep it forever.
Hey! I have a blog now: http://mancave-stereo.blogspot.com or "like" us at https://www.facebook.com/mancave.ster
Probably the best DAC ever made. The cost killed it..
But inmate 'mikel' has one and loves it.
Me?
I gotta settle for only 8 ea PCM1704U-K in may Audio-GD Master 7 and worse, just 4 ea. in my Audio-GD Master 11. :-(
Ford in blue purple color
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: