"Breaking Silence" is an audiophile favourite from way back and the recording sure sounds pretty amazing. I believe they plugged the mic pre's straight into the (A80) analog multitrack tape machine.
Having said that I think it is also pretty boring musically speaking which could also be said for a lot of the old audiophile classics, some incredible recordings but most of the songs or performances just don't grab me.
TCD
I got, and read a copy of Sony’s CD white paper in 1980. At the time, taking a graduate course in sampling theory, i suggested that the time resolution would need to increase at least 4 x before it started to compete with analog.
The first CD players sucked big time. I did not get my own until near 2k. I am actually quite amazed at how far they have managed to be improved.
dave
The first CD players sucked big time. I did not get my own until near 2k. I am actually quite amazed at how far they have managed to be improved.
dave
Not necessarily for enjoyment. A few good recordings can be selected for use in listening comparisons when discrimination is sought rather than preference. It means one has to learn a recording well enough to discriminate when something very specific about the sound changes even if its only very briefly. A good recording with a good combination of reproduction attributes can provide a lot of information about system performance in one listening. OTOH using random playlists may be more enjoyable, but it may be better suited for preference comparisons rather than for discrimination of specific differences.
The WELL Audio site shows photos of the TWSDAC-1862, and TWSDA-1541 which appear to show transformer coupled line outputs. Perhaps, these might be photos of newer versions of those two DACs? Meanwhile, I agree that the uniform presence of NOS operation among the top 5 DACs doesn’t by itself prove the significance of NOS to their subjective performance, but it does legitimately raise the strong suspicion.
Last edited:
Sure... I had the first Magnavox copy of the Philips top loader.
I suspect the recorders were highly problematic as well. With only 16 bit headroom I suspect setting the clipping limits in predicting what transients could occur seems likely to have restricted bit depth or dynamic range to far less than 16 bits. With the advent of 24/32 bit ADC's and ultra low distortions there is limited need to adjust levels to prevent overload. Hence a 16 bit recording can "windowed" from the 24/32 bit recording to preserve maximum dynamic range being restricted to 16 bits if desired.
What is interesting is that streaming in Tidal is still 16 bits /44.1KHz in "HiRez". Yet the outcome can be drastically superior to stuff produced 40 years ago... or not.
I hope you realize a difference between sound reproducing equipment and sound reproducing hobby.
You do realise that getting under 5% THD out a vinyl replay system is good going. The cartridge generates a whole host of pleasing harmonics.
That depends more on the replay system...
I suspect the CBS DisComputer did something pretty nasty, but no idea how digital delay lines were done in the late 70s.
True. In my experience the weighting, suspension and balance are also critical to performance, perhaps with isolation from ground being the most critical. It is fortunate that the mechanical harmonic spectrum becomes relatively attenuated as corresponding to points on the RIAA curve, falling off at about a 6dB/octave. Unfortunately the intermodulation products can also be relatively amplified at the low end.
Is my claim an extraordinary one? Also, do you now acknowledge that Markw4's post was a claim and not a "statement of opinion"?
How do you know that?
When did ordinary vs extraordinary claim discussion become the veracity of claim discussion?
Level matched double blind listening tests of DACs reported to be audibly indistinguishable despite the price differences or market status (high-end vs run of the mill spec grade) and those have been ordinary occurrences. If someone did distinguish them in the same type of test, that would be extraordinary.
When did ordinary vs extraordinary claim discussion become the universal truth discussion? I don't remember reading Markw4's post saying so.
Indeed. Was using Sumiko, now using Korf ceramic. Listening differences were obvious enough. Only keeping the Sumiko as backup in case I ever want to use it to try a test cartridge.
Notwithstanding previous posts, there is a distinction between "statements of opinion" and "statements of fact". A claim can be a "statement of fact" or a "statement of opinion". "A fact is a statement that can be proven true or false. An opinion is an expression of a person's feelings that cannot be proven". The "claim" by Markw4 appears a statement of opinion, yours is presented aa a statement of fact. At issue is if harmonic distortions and other artifacts, as visually depicted in measurements, can be heard. Secondly, if the testing encompasses all manner of noise (as not signal), as being identifiable or otherwise, as to support the conclusion that Markw4's "statement of opinion" is false.
I recently compared an SMSL Su-9n DAC to the SMSL AO100 one. Both have harmonic distortion artifacts below 140dB (as tested by AudioScienceReview ("ASR")). This suggests that if the full scale dynamic range was set to reproduce an actual jet engine sound at 120dB SPL a test subject would need to resolve harmonic distortion artifacts at 20 dB below the threshold of human hearing, or at -20dB. Of note is that the threshold of human hearing is determined in an anechoic environment, whereupon any background noise corrupts that resolution. This suggests that for physical detection of harmonics to occur under such circumstances, the threshold of human hearing is being enhanced by 20dB (to -20dB) as a function of the presence of a jet engine being reproduced at 120dB SPL.
The conclusion is that a claim presented as "statements of fact", specifically that artifacts heard are caused by such low levels of harmonic distortion, are too far fetched to imagine. Hence any double blind, or otherwise, as could still result in some form of detectable outcome, can't be attributed to their detection. It can only be something else.
With that said the SMSL A0100 was found sonically superior to the SMSL Su-9n. Of note is that the SMSL Su-9n has technical performance data nearly identical to the Gustard X-16 as indicated in reviews by ASR. In turn the Gustard X-16 DAC was compared by the Absolute Sound in their review of the Topping D90SE, being found sonically indistinguishable. All DAC's use variant ESS Sabre DAC's and have XLR outputs. The costs and the D/A converters employed are as listed below:
Topping D90SE $1169 Canadian with MQA (ES9038Pro)
Gustard X-16 $649 Canadian with MQA. (2x ES9068AS)
SMSL Su-9n $520. Canadian no MQA. (ES9038Pro)
SMSL DO100 $299. Canadian no MQA. (2x ES9038Q2M)
Despite evidence presented herein being of questionable interpretation, and that AKM D/A's are perceived at least marginally superior to ESS devices in previous experiences, the SMSL DO100 is considered at least as good as the Topping D90SE, making the SMSL a substantial bargain in this grouping of DAC's.
Feel its necessary to comment on some of the above reasoning, just in case anyone reading isn't familiar with some related issues:
Thresholds of Hearing are not absolute limits. They are estimates of an average value for a population. In this case, 'average' refers to the concept that 50% of the population can't hear below the limit, and the other 50% of the population can hear below the limit. Also, such thresholds are measured with sine wave test signals. Modern research into human hearing is now mostly conducted in the field of "Auditory Scene Analysis." https://www.frontiersin.org/article...is (ASA) refers,sound waves reaching the ears.
It is becoming increasingly apparent that human hearing is much linear and time-invariant than had been previously assumed by some people. It means sine wave derived measurements cannot be assumed to accurately apply to more complex waveforms. Nor can dB SPL numbers necessarily be accurately added and subtracted from each other as though they were taken from measurements of a much more linear and time-invariant system.
IOW, the arguments about jet engine SPL verses Threshold of Hearing SPL numbers cannot be assumed to represent reasonable modeling. Quite the opposite. We all know from experience that loud noise is deafening.
IMHO what we measure is HD with simple sine wave test signals, since the results are easy to interpret. Some people assume those numbers represent actual distortion levels possible during music playback.
OTOH, what we hear when we listen to reproduction of a 100-piece orchestra playing dynamically changing volume level and including harmonics from each instrument, must amount to 10s of thousands of frequencies present at once. What is the total level of IMD we should then expect? Below the 'Threshold of Hearing,' whatever it may be for the actual complex time-domain waveform?
All the above having been said, we still know there is more to delta-sigma dac artifacts than we would expect from a simple analog linear amplifier. Assumptions about linearity and time-invariance as being no more that 'weak' effects in such dacs would seem to involve an unjustified leap of faith.
Too late to edit. Should refer to "no more than weakly non-linear and non-time invariant behavior"
IOW, forgot to invert the logic.
Looking at ASR measurements SMSL SU-9n suffers from similar issues as Topping D90 which were already discussed earlier in this thread. The noise skirt is much bigger and wider than in SMSL DO100. The cause for this is probably Vref noise and clocking issues. Whether or not that is the reason for any audible differences is another thing altogether.
- Status
- Not open for further replies.