True. As I have said before, sound that is too loud is deafening. There is some optimal volume level for discrimination of distortion artifacts. Also, it seems to matter to our very nonlinear detection system whether or not a very low level sound is strongly correlated with the loud sound we are listening to. Especially so if the loud sound is one we are very familiar with, say, such as vocals. A subtle change in the perceived texture of vocal harmonies can be quite detectable at some optimal SPL. Uncorrelated sounds, such as pin dropping during a loud vocal performance, is much less likely to come to the attention of conscious awareness. All the more so if attention is deeply focused on the vocals.
Changing the subject for a moment, I am familiar with the sound of a number of Sabre dacs, and I have listened to DAC_Lite. That includes for piano music. I understand with @The Well Audio is talking about. IME its not BS, there is some truth to it. However, middle of next week there will be a listening session here. We will see what opinions come out of that. Could be there will be more to say about the pros and cons of various approaches to dac design.
When a dacs own distortion is higher than a speaker driver that it is driving and then yu are adding amplification which may also contain distortions of that order and your DSP is attempting to correct for distortions via FIR. i'm not sure I would want my DSP to identify the DAC as a source of error. no, sorry I mean you wouldnt want your DSP to lump that distortion in with the driver/room, but be unable to remove that part. I would think that would compound in some way. I dunno, just gives me the heeby jeebies
That could certainly do it. However, there have been complaints that dacs do not fully reproduce the reverb tails already recorded to master tapes. IMHO there must be something else wrong. IMHO using compression at playback to compensate would not really be a proper fix.
As an additional observation, besides master tapes, natural reverb tail ambience can also be heard on direct-to-disk-lathe vinyl recordings.
Actually not. You jump to conclusions is all. I have been asked to reproduce reverb tails for people that want and like natural ambience. As I think about the accurate digital reproduction question, hard not to notice how some classic audio circuits work to sound the way they do.
Last edited:
Why, because it's inaudible?
Could it possibly be due to different listening spots cause different listening experience, a.k.a. "room mode"?
Should we wait till he brings up Yamaha piano?
Many home audio enthusiasts ignore or are unaware of the importance of room acoustics quality. When they aren't satisfied with their purchase, they bark up the wrong tree (cables, amps, DACs, oscillator... etc.) and the money pit persists.
I thought I mentioned "accuracy of output compared to input." Is it not showing on your screen?
What about audio reproducing equipment?
Do you mean our hearing sense? If so, yes, it is non-linear per frequency level as well as age. Sound detection tests should be handled by those with young ears.
Is that the one which you invited andrea_mori to?
CD fails to reproduce anything above 22.05 kHz and CD players typically clip on intersample overshoots. In what sense is that accurate? You just have different inaccuracies than with a vinyl record.
Were wires dangling something like below?
The CD player is a Marantz CD-94 that used a single TDA1541A converted to NOS. When the wiring was pulled apart in just the right manner the sound was spectacular. Before you ask the listening session consisted of a personal evaluation with conclusions being drawn in the first three seconds, along with measurements speculated to be spectacular as well. What was particularly noteworthy as extraordinary was its ability to acoustically render the plant life and other artifacts in my front yard.
View attachment 1063246
Many of us stopped using CD players a while ago Marcel
Evenharmonics brought them up, I didn't. In any case, 44.1 kHz sample rate recordings are still quite common as far as I know.
Hard clipping hampers audio reproduction, and some believe steep filtering of ultrasonics does the same.
That said, I personally hear no difference whatsoever between the apodizing and steep settings of my DAC and the 15 kHz ... 16 kHz bandwidth limitation of FM radio doesn't bother me at all (unlike the gross dynamic range compression of most FM stations).
yes, but easy to attenuate by a few dB to help with intersample over with a dac
What I called BS was the claim that "In the real world the Steinway is always recognizable not depending on the location." That has nothing to do with any DACs.
I suspect there is more to masking artifacts being in variant relationship to the amplitude of harmonics that affects our perceptions. A picture of harmonics in a distortion analysis shows noise values at the baseline. Although this is indicative of the energy of such noise, as to suggest reflection the degree of masking, that energy is band limited by the sweep. As you are no doubt aware the baseline is variant by the bandwidth selected in the test sweep. In other words a 1/10th octave sweep has a much higher baseline than a 1/100th octave sweep. Some might believe that the baseline indicated in a distortion analysis relates directly to the degree of masking suggested by that baseline. A question becomes as to the manner the ear is dealing with masking noise as such relates to the bandwidths.
I previously compared the SMSL Su-9n Sabre based dac to the original version of the SMSL M100 that used an AKM device. This was using RCA connections, as the M100 only has RCA's. In that case the SMSL M100 appeared sonically superior to the Su-9n, as notwithstanding the poorer performances indicated by ASR in the testing. As the M100 is powered from an independent USB port, this was also tried with a Lithium ion battery having a USB output port. This magnified the differences. IME one of the more significant indicators of improved sonics relate to the blackness of the background, this as supporting clarity into the longer decays of the resonance of instruments.
The cause for continuing in using the SMSL Su-9n related to a desire to investigate into its balanced XLR capability. The DAC was connected to a pair of Electrovoice Q66 MKII professional power amplifiers having provision for only XLR inputs. As only Channel A was connected in each amp the power supply was relieved of providing current to a second channel. (As an aside, being connected as mono-blocks improved its sonic refinement and increased a sense of ease as related to dynamics). As the amps have input gain controls the levels were set to approx. -20dB. This means that the preamp output levels are x10 greater for a specific SPL. Given the remarkable distortion characteristics of such DAC's such performance can be traded off for higher signal to noise ratios by increasing levels in relation to existing noise. The reduction of amplifier gain, or having selections to do so, are becoming increasingly common, seemingly for that reason.
The result was that the sonic improvements gained in the use of XLR's seemed at least to make up for the losses in the differences in the DAC's using alternative RCA connections. This improved with the use of the SMSL AO100.
Were wires dangling something like below?
The CD player is a Marantz CD-94 that used a single TDA1541A converted to NOS. When the wiring was pulled apart in just the right manner the sound was spectacular. Before you ask the listening session consisted of a personal evaluation with conclusions being drawn in the first three seconds, along with measurements speculated to be spectacular as well. What was particularly noteworthy as extraordinary was its ability to acoustically render the plant life and other artifacts in my front yard.
Sorry pic was lost...
I agree that inter-overs is a problem but also that the time spent on a CD where levels are in a risk territory is quite short - no?
So general SQ is not really constantly effected - only at forte
//
This is technical nonsense - 16/44 is superior to tape in every aspect - every time. Do you really believe there is someting magic in "reverb tails" that dont stick to digital ;-)
The thing is that when you have linear level reproduction, you just don't hear them as there is no compression. This is trivial - you need to stop repeating this over and over. There is just noting in it. Please, be scientific instead of mystic.
//
- Status
- Not open for further replies.