TIDAL is NOT Worth it! Listening Test

[fas42]

How so? Seems like pretty normal stuff...what struck you as naughty?

That the quality of the electronics feeding the speakers which he is extremely familiar with, , made a huge difference to the subjective quality of the listening experience - ie, the chain prior to the speakers are more important than the speakers.

 

[Ken Newton]

An example of that is... what?

Mass market AVR receivers, for example. As complex and excellently engineered as most of them are, they are understandably designed to a set of specifications, as engineers require objective design targets. Which is common professional engineering practice. My issue is that common metrics don't necessarily equate to perceieved sound.



An example of something that has been (anything fully translated to accurately reflect the user's subjective experience) is... what?[/QUOTE]

 

[Blumlein 88]

Making a "big deal" is a relative notion. Actually, that's not the characterization which I had applied. Only that time and money resources were being needlessly spent in those areas. But, certainly not by everyone.

Not only is it not a big deal nor being pursued, like watchnerd it mostly is a side effect of better IC manufacturing. So it might be picked up a bit in marketing as those guys will use anything to push a product. It isn't a big driving factor in thinking to any real extent. You could make a better case that SNR is being pursued further than it needs to be. And yet again I am not so sure that isn't just a benefit of better IC design and manufacturing. Things are at the point it costs nothing to design those better so why not.

Where you could make something of a case is the pursuit of ever higher sample rates. Of course that is driven by the golden eared crowd. Or maybe the guys pushing the non-rining filters when it isn't clear ringing is audible (it surely isn't if you do use 96 khz rates). Yet the various non-ringing filters degrade aliasing performance and frequency response in pursuit of a myth. Again that is the golden eared crowd pushing that stuff.

 

[Blumlein 88]

Mass market AVR receivers, for example. As complex and excellently engineered as most of them are, they are understandably designed to a set of specifications, as engineers require objective design targets. Which is common professional engineering practice. My issue is that common metrics don't necessarily equate to perceieved sound.

You do realize this isn't really in support of what you were claiming. Yes complex AVR design requires designing to spec. The spec if achieved is something that isn't going to be heard. Where you do get some real differences is AVR's don't have super amplification and if used with difficult speakers they differ some in sound. That does not seem to be an example of what you were claiming.

 

[Ken Newton]

Again you're making this up. As asked by others give some examples.

Right, because you disagree, I must necessarily he making it up. Good logic.

While I've already given some simple examples, no doubt, and to no surprise, you'll find those faulty.

I think other than transducers all the problems are below audibility. So that part of the chain is a clear channel without humanly observable differences.

Another assertion. My experience is different, that doesn't invalidate your experience, and yours doesn't invalidate mine. Let's leave it at that.

 

[Blumlein 88]

Right, because you disagree, I must necessarily he making it up. Good logic.

While I've already given some simple examples, no doubt, and to no surprise, you'll find those faulty.



Another assertion. My experience is different, that doesn't invalidate your experience, and yours doesn't invalidate mine. Let's leave it at that.

Let's not leave it that way. You are intentionally misrepresenting me. You also have been coy and indirect about what you mean. You mentioned AVR's so I'll suggest a supposition about what you mean.

Designed to serve specification instead of serving the human ear. Director of ACME AVR says to his team, I need another 9.1 channel AVR. Make sure it can do 100 wpc with at least 95 db SNR and flat response to 40 khz while keeping THD, IMD etc to .1 % or less. Product is made, and some Superear somewhere or another listens to this one and another one made to about the same spec and declares either they aren't good or they sound different or both. It wasn't designed to serve the ear.

Now should they have said, in service to the ear, we know it doesn't like high feedback. So we should have built this to maintain all those specs without using feedback. Even low distortion is an irritant to the ear when done with high, high feedback.

Is this something like what you going on about here? If not clarify please.

The problem with that primacy of the ear hearing beyond spec is whenever we let Mr. Superear try to pick high feedback .1 % distortion from no feedback .1% distortion with any controls in place he fails to do so. His experience hearing the difference was real enough for him, but his ability to actually hear what he experienced in physical reality was not so good.

The reason designing to the spec of .1 % distortion is chosen in the first place is at or below that is where humans have been tested and found not to hear a difference. In my mind when such parameters are well chosen it is in fact the ear's performance envelope they have been chosen to serve transparently. If some other transfer function makes the ear think it is more real it might be because it adds a desirable character. Not because it is of superior fidelity. The ear is not a reliable chooser of the most real only what it thinks is the most real. But those are the differences in fidelity and preference.

 

[Ken Newton]

Let's not leave it that way. You are intentionally misrepresenting me. You also have been coy and indirect about what you mean.

Again, with the negative assertions of my intent. You have no way of knowing my intent. You would have been better off leaving it alone, as you're becoming increasingly reactionary and emotional. I hope to make this my final response to you on this subject.

You mentioned AVR's so I'll suggest a supposition about what you mean.

Designed to serve specification instead of serving the human ear. Director of ACME AVR says to his team, I need another 9.1 channel AVR. Make sure it can do 100 wpc with at least 95 db SNR and flat response to 40 khz while keeping THD, IMD etc to .1 % or less. Product is made, and some Superear somewhere or another listens to this one and another one made to about the same spec and declares either they aren't good or they sound different or both. It wasn't designed to serve the ear...Now should they have said, in service to the ear, we know it doesn't like high feedback. So we should have built this to maintain all those specs without using feedback. Even low distortion is an irritant to the ear when done with high, high feedback.

Is this something like what you going on about here? If not clarify please.

I've already attempted to explain multiple times that my concern is only that traditional specifications don't seem to tell the compete story of the resulting subjective human experience. I believe that because my subjective listening experience leads me to suspect that something psychoacoustic in nature is being missed in the translation of ear-brain response in to parametric design objectives.

For example, the CD standard was designed to a set of outstanding specifications, just as AVRs are. Most objectivists declared that if your ears didn't like the sound of CD, then your ears were necessarily in error. Later, it was realized that other technical parameters, which evidently, were not much considered when CD's spcs. were orignally set down, were audible factors in the subjective experience. Parameters such as clock jitter, just to name one, but there are others which are maybe still less well recognized.

No, I'm not picking on high-feedback, nor any other particular technique. I'm also not suggesting that I have some universally applicable secret formula to superior sound. I've been suggestng that there's more going on with the human subjective reaction to audio system playback than is strictly dictated by traditional audio design specifications. The reaction by many objectivists to that suggestion is much the same as it was by CD objectivists to subjective observation reports that something was lacking with CD. As I've siad earlier, I believe that if the sound is different then the signal must also be different. I simply suspect that there are still such difference left to find.

While the story of CD digital audio provides a relatively recent example, here's another example, an old one, Which is the discovery that another distortion parameter was important to the ear that once wasn't recognized as such. We learned some long time ago that a singular THD figure isn't the only distortion parameter important to the ear, but that it's spectral distribution can also affect the perception of that distortion.

The problem with that primacy of the ear hearing beyond spec is whenever we let Mr. Superear try to pick high feedback .1 % distortion from no feedback .1% distortion with any controls in place he fails to do so. His experience hearing the difference was real enough for him, but his ability to actually hear what he experienced in physical reality was not so good...The reason designing to the spec of .1 % distortion is chosen in the first place is at or below that is where humans have been tested and found not to hear a difference. In my mind when such parameters are well chosen it is in fact the ear's performance envelope they have been chosen to serve transparently. If some other transfer function makes the ear think it is more real it might be because it adds a desirable character. Not because it is of superior fidelity. The ear is not a reliable chooser of the most real only what it thinks is the most real. But those are the differences in fidelity and preference.

I'm not questioning whether 0.1% THD is a sufficiently low metric. I'm questioning whether a distortion test traditonally utilizing static test signals is sufficent to determining the subjective expeience of sound based on a dynamic signal, especially a spectrally complex and dynamic signal such as music. I'm not declaring that staic test signals are necessarily inadequate, I don't know that answer, but it seems a fair question. However, what I have observed is that there doesn't seem to be an accurate and 100% reliable correlation between traditional specifications and the subjective experience.

Lastly, I don't see why my offering personal observations provokes such an emotional response from you, including personal attacks on my motives, when I'd made none towards you. Enough said from me. You, no doubt, will feel compelled to respond. However, I expect to be finished with this as it's become a non-productive discussion with you at this point.

 

[Sal1950]

I hope to make this my final response to you on this subject.

We hope so too!

 

[Ken Newton]

We hope so too!

 

[oivavoi]

Again, with the negative assertions of my intent. You have no way of knowing my intent. You would have been better off leaving it alone, as you're becoming increasingly reactionary and emotional. I hope to make this my final response to you on this subject.



I've already attempted to explain multiple times that my concern is only that traditional specifications don't seem to tell the compete story of the resulting subjective human experience. I believe that because my subjective listening experience leads me to suspect that something psychoacoustic in nature is being missed in the translation of ear-brain response in to parametric design objectives.

For example, the CD standard was designed to a set of outstanding specifications, just as AVRs are. Most objectivists declared that if your ears didn't like the sound of CD, then your ears were necessarily in error. Later, it was realized that other technical parameters, which evidently, were not much considered when CD's spcs. were orignally set down, were audible factors in the subjective experience. Parameters such as clock jitter, just to name one, but there are others which are maybe still less well recognized.

No, I'm not picking on high-feedback, nor any other particular technique. I'm also not suggesting that I have some universally applicable secret formula to superior sound. I've been suggestng that there's more going on with the human subjective reaction to audio system playback than is strictly dictated by traditional audio design specifications. The reaction by many objectivists to that suggestion is much the same as it was by CD objectivists to subjective observation reports that something was lacking with CD. As I've siad earlier, I believe that if the sound is different then the signal must also be different. I simply suspect that there are still such difference left to find.

While the story of CD digital audio provides a relatively recent example, here's another example, an old one, Which is the discovery that another distortion parameter was important to the ear that once wasn't recognized as such. We learned some long time ago that a singular THD figure isn't the only distortion parameter important to the ear, but that it's spectral distribution can also affect the perception of that distortion.




I'm not questioning whether 0.1% THD is a sufficiently low metric. I'm questioning whether a distortion test traditonally utilizing static test signals is sufficent to determining the subjective expeience of sound based on a dynamic signal, especially a spectrally complex and dynamic signal such as music. I'm not declaring that staic test signals are necessarily inadequate, I don't know that answer, but it seems a fair question. However, what I have observed is that there doesn't seem to be an accurate and 100% reliable correlation between traditional specifications and the subjective experience.

Lastly, I don't see why my offering personal observations provokes such an emotional response from you, including personal attacks on my motives, when I'd made none towards you. Enough said from me. You, no doubt, will feel compelled to respond. However, I expect to be finished with this as it's become a non-productive discussion with you at this point.

I may not agree with everything you write, but I appreciate your contributions to the discussion!

 

[Blumlein 88]

Again, with the negative assertions of my intent. You have no way of knowing my intent. You would have been better off leaving it alone, as you're becoming increasingly reactionary and emotional. I hope to make this my final response to you on this subject.



I've already attempted to explain multiple times that my concern is only that traditional specifications don't seem to tell the compete story of the resulting subjective human experience. I believe that because my subjective listening experience leads me to suspect that something psychoacoustic in nature is being missed in the translation of ear-brain response in to parametric design objectives.

For example, the CD standard was designed to a set of outstanding specifications, just as AVRs are. Most objectivists declared that if your ears didn't like the sound of CD, then your ears were necessarily in error. Later, it was realized that other technical parameters, which evidently, were not much considered when CD's spcs. were orignally set down, were audible factors in the subjective experience. Parameters such as clock jitter, just to name one, but there are others which are maybe still less well recognized.

No, I'm not picking on high-feedback, nor any other particular technique. I'm also not suggesting that I have some universally applicable secret formula to superior sound. I've been suggestng that there's more going on with the human subjective reaction to audio system playback than is strictly dictated by traditional audio design specifications. The reaction by many objectivists to that suggestion is much the same as it was by CD objectivists to subjective observation reports that something was lacking with CD. As I've siad earlier, I believe that if the sound is different then the signal must also be different. I simply suspect that there are still such difference left to find.

While the story of CD digital audio provides a relatively recent example, here's another example, an old one, Which is the discovery that another distortion parameter was important to the ear that once wasn't recognized as such. We learned some long time ago that a singular THD figure isn't the only distortion parameter important to the ear, but that it's spectral distribution can also affect the perception of that distortion.




I'm not questioning whether 0.1% THD is a sufficiently low metric. I'm questioning whether a distortion test traditonally utilizing static test signals is sufficent to determining the subjective expeience of sound based on a dynamic signal, especially a spectrally complex and dynamic signal such as music. I'm not declaring that staic test signals are necessarily inadequate, I don't know that answer, but it seems a fair question. However, what I have observed is that there doesn't seem to be an accurate and 100% reliable correlation between traditional specifications and the subjective experience.

Lastly, I don't see why my offering personal observations provokes such an emotional response from you, including personal attacks on my motives, when I'd made none towards you. Enough said from me. You, no doubt, will feel compelled to respond. However, I expect to be finished with this as it's become a non-productive discussion with you at this point.

I'll make a short comment. Your CD example is a good one though not the way you think it is. JITTER was never a real problem with CD. Nor is there anything wrong with the specifications to this day. The format is fine. The use of it took some people time to understand vs practice developed for an analog tape world.

 

[Blumlein 88]

http://archimago.blogspot.com/2016/04/retro-measure-1994-sony-mdp-750.html

Here is some measurements done on a 1994 Laserdisc CD player from Sony. You can see the jitter levels on it are not bad at all. Wish I had measures of the first generation or two to put up. I did some on a 3rd gen Magnavox player at one point. Not too different than the Sony here on the Jtest. A bit higher noise floor than modern players. We are talking like a 1985 or so player being 3rd generation.

 

[watchnerd]

http://archimago.blogspot.com/2016/04/retro-measure-1994-sony-mdp-750.html

Here is some measurements done on a 1994 Laserdisc CD player from Sony. You can see the jitter levels on it are not bad at all. Wish I had measures of the first generation or two to put up. I did some on a 3rd gen Magnavox player at one point. Not too different than the Sony here on the Jtest. A bit higher noise floor than modern players. We are talking like a 1985 or so player being 3rd generation.

I'd say better than "not bad", it's actually comparable to the modern day budget DACs.

 

[mvil]

Today i tested first the TIDAL trial of 60 days (https://tidal.com/try-now?campaignId=5c90db3427dd79001024654e)

I immediately felt a very big difference between Spotify premium at 320 and tidal hifi, being tidal the best, thne i recorded both streams to audacity, and used delta wave to compare them.

At 10 seg specifically

https://tidal.com/browse/track/116887361





How can this HUGE diference be explained?

 

[renevoorburg]

How can this HUGE diference be explained?

Perhaps because the timing of the two files gradually gets more and more out of sync?

 

[mvil]

Perhaps because the timing of the two files gradually gets more and more out of sync?

DeltaWave v1.0.45, 2019-10-07T10:40:36.2007882+02:00
Reference:  spotify2.wav[?] 824216 samples 48000Hz 16bits, stereo, MD5=93f6b0dc77da5eb43344a991923874ce
Comparison: tidal2.wav[?] 824216 samples 48000Hz 16bits, stereo, MD5=e537679487fb18996d3dcfe75ed13405
Settings:
    Gain:True, Remove DC:True
    Non-linear Gain EQ:False    Non-linear Phase EQ: False
    EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -160dB
    Correct Drift:True, Precision:30
    Non-Linear drift Correction:False
    Upsample:False, Window:Hann
    Spectrum Window:Hann, Spectrum Size:32768
    Spectrogram Window:Hann, Spectrogram Size:4096, Spectrogram Steps:2048
    Dither:False
    Trim Silence:True
    Enable Simple Waveform Measurement: False

Discarding Reference:  Start=0s, End=0s
Discarding Comparison: Start=0s, End=0s

Initial peak values Reference: -0.125dB   Comparison: -0.124dB
Initial RMS values Reference: -10.133dB   Comparison: -10.142dB

Null Depth=12.49dB
X-Correlation offset: -960 samples
Trimming 0 samples at start and 1 samples at the end that are below -90.31dB level

Drift computation quality, #1: Excellent (1.01μs)


Trimmed 7862 samples ( 163.791667ms) front, 133399 samples ( 2779.145833ms end)


Final peak values Reference: -0.125dB   Comparison: -0.039dB
Final RMS values Reference: -10.112dB   Comparison: -10.084dB

Gain= 0.0055dB (1.0006x) DC=-0.00004 Phase offset=-20.000879ms (-960.042 samples)
Difference (rms) = -29.22dB [-31.42dBA]
Correlated Null Depth=55.99dB [49.56dBA]
Clock drift: 0.28 ppm


Files are NOT a bit-perfect match (match=0.19%) at 16 bits
Files match @ 49.9803% when reduced to 7.4 bits


---- Phase difference (full bandwidth): 39.9483512777137°
    0-10kHz: 24.88°
    0-20kHz: 34.28°
    0-24kHz: 39.95°
Timing error (rms jitter): 18.1μs

RMS of the difference of spectra: -84.2748743276308dB
gn=0.999372072620961, dc=-3.9267776708044E-05, dr=2.76621E-07, of=-960.0422090838

DONE!

Signature: 54631493ab1944fb7f0ee059dc6abc40

 

[Soniclife]

Today i tested first the TIDAL trial of 60 days (https://tidal.com/try-now?campaignId=5c90db3427dd79001024654e)

I immediately felt a very big difference between Spotify premium at 320 and tidal hifi, being tidal the best, thne i recorded both streams to audacity, and used delta wave to compare them.

At 10 seg specifically

https://tidal.com/browse/track/116887361





How can this HUGE diference be explained?

Can you describe the method used to capture the streams. Lossy will often use a low pass to remove high frequency data, but that looks way bigger then I would expect. Did you capture the whole track or a snippet?

 

[mvil]

Can you describe the method used to capture the streams. Lossy will often use a low pass to remove high frequency data, but that looks way bigger then I would expect. Did you capture the whole track or a snippet?

Loopback on Windows (no effects). Both volumes on the apps maxed out.

Record to start on audacity with -50dB, after I hit play on Tidal or Spotify it starts recording, for about 15 seconds of the music I posted.

Then trim them for the same amount in the end.

Spotify defined for Very High.

Tidal defined for HIFI.

Export individual wav file on audacity and compare and match them on delta wave.

 

[Soniclife]

Loopback on Windows (no effects). Both volumes on the apps maxed out.

Record to start on audacity with -50dB, after I hit play on Tidal or Spotify it starts recording, for about 15 seconds of the music I posted.

Then trim them for the same amount in the end.

Spotify defined for Very High.

Tidal defined for HIFI.

Export individual wav file on audacity and compare and match them on delta wave.

What do you mean by -50db?
Did you set the sound card to 16/44?
Did you set audacity to 16/44?
Try with a longer recording. I record the whole track then set the window in deltawave.

 

[mvil]

What do you mean by -50db?
Did you set the sound card to 16/44?
Did you set audacity to 16/44?
Try with a longer recording. I record the whole track then set the window in deltawave.

Recording is trigered when the level reaches -50.

Both are set into 16/48.

I can try with a longer one, but if I can listen to it, it is that different on all the track.