 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
I love the Exact Instructions track. Really tense.
That is one of my favorite tracks as well. Already was on the old LP.
|
|
|
|
|
|
|
|
|
|
I have a few high-res files myself, though I never actually tried to blind test them against CD-quality audio. There is no question that CD-quality audio has always been good (better in fact than anything on vinyl). So I am not one who claims that I could easily or at all hear differences between high-res and CD quality audio. (I do make the claim and have verified though that I can hear the difference between MP3 and CD quality audio, and if an average Joe like me can do it, I suppose just about everyone could.)
The linked Waldrep article also conflates MQA and other high-res formates, which are obviously not the same. (Yeah, I'm sure he knows that too, it just was not the point he is making). I have not heard any MQA encoded CDs yet myself. The Waldrep test is also not a "study" but more of a survey, whereas the QMU study is actually a compilation of various studies, so regardless of whether some of the results can be questioned, its results are still more significant.
The thing is: if even just a small percentage of the population could even hear a difference between CD quality audio and high-res files, it would be reason enough to validate the format. There is not question that high-res files provide a higher audio resolution when compared to Redbook standard CDs, so even if many cannot hear the difference, the fact that there IS indeed a difference is sufficient enough reason for some to prefer that format. (Just like people buy 8K Ultra HD TVs, even if you may not see the difference when you sit on your couch.) Maybe my dogs could hear it. They sure hear sounds from my stereo system that I cannot hear. :-)
After all, these days, high-res files often cost not much more than CD quality audio, in some cases -- like Benjamin Wallfisch's soundtrack for "IT Chapter Two" even less. (The high-res download costs actually less on Qobuz than the physical CD release, with a subscription even just about half that.) So even if there is no audible difference, the mere fact that better files are available are sufficient reason for many to use that format. The advantages are at least theoretically better sound (regardless of whether it is audible) without any downside. (There used to be the reason of space, but in the age when even portable hard disc space is measured in Terrabytes, that is hardly an issue anymore.)
Agree with most of what you say. I've enjoyed the back and forth. So many people get all kinds of bent out of shape around this topic 
Agree that the Waldrep test is more of a survey, but if you look at his more recent comments, he is trying to structure a more scientifically valid methodology with his latest listening trials, and has proposed a test protocol to the AES.
Good point about 8K TVs. Talk about a marketing gimmick. The visible improvements have all come with the implementation of wider color gamuts and high dynamic range, NOT resolution.
|
|
|
|
|
|
|
Most of the Hi-Res files actually don't benefit from the qualities allowed by the format.
The problem first lies in the mastering.
An example with Bill Conti's Rocky:
http://dr.loudness-war.info/album/view/108534
http://dr.loudness-war.info/album/view/96457
The first CD release is much better than the 24-bit digital file.
About The Swarm, a 13-bit CD may be enough considering that La-La Land usually doesn't retain the full dynamic range of the recordings.
Good point about the actual dynamic range vs. the theoretical. Most real world listening environments have such a high noise floor that if you had a true 16 bits of dynamic range, you would have to constantly boost the volume control to hear the quieter passages, then have to turn it down drastically whenever the louder sections came blasting in. That's what led to the loudness wars to begin with - overcoming noisy listening environments in the car, on the street, etc. Without compression, all that low level detail got lost in the background noise of the listening space.
Ideally, we'd have a playback system that retained dynamic range while giving the user control over dynamics in order to adapt to the environment. For critical listening at home, full dynamic range. For listening on the go, the ability to add quality compression to taste.
|
|
|
|
|
|
|
And expensive systems doesn't mean these systems are good.
One of my friend is an electronic engineer who makes amplifiers (he made mine) and actually set up a small Hi-Fi brand which also makes speakers. From my experience, many expensive systems are a joke, and they actually don't worth their price.
Hi-Res can brings subtleties that are lost by most systems, including the most expensive, whoever made them.
Stradivarius and Guarneri were not "Dr." and certainly better knew how to make sound a bit of wood. The same in Hi-Fi, the best speakers I have ever heard were made by a cabinetmaker.
"And expensive systems doesn't mean these systems are good."
Indeed! Are you familiar with Harman's MLL? It's a double blind listening chamber that features a pneumatic speaker shuffler, allowing listeners to precisely match speaker location and volume. They have found over almost 40 years of evaluating speakers in a double blind fashion that some of the worst sounding speakers were also some of the most expensive. Most of this research correlating listener preference with speaker measurements was funded and accomplished years ago by the Canadian government; it was refined more recently at Harman's labs in Northridge CA.
What makes a good sounding speaker is now well understood. With the new speaker measurement standard developed by Drs. Toole and Olive (CEA standard #2034, to be precise), they are now able to predict with 86% accuracy which speaker will win the double blind listening tests just by looking at the measurements. This standard has been peer reviewed in scientific journals and is now officially endorsed by the Consumer Technology Association. As mentioned, this research started with the Canadian National Research Council and continued with an even bigger budget at Harman. With bookshelf speakers, the correlation between measured performance and listener preference is actually 99%.
What they discovered is that the speakers with the most neutral (i.e., flat) frequency response both on and off axis invariably won the listening tests. This was true regardless of listener age, nationality, sex, or ethnicity. It should probably not be a surprise that the speakers that least color the sound are perceived as sounding the best by listeners of all stripes. And it has nothing to do with what the cabinet is made out of, as long as it is free of resonances. If we listen to a vocalist or instrumentalist in an acoustic space, our ear / brain combination has an expectation that a recording of it should sound the same when played back through speakers. In other words, the speaker should not alter the timbre of the original recording.
www.soundstage.com has a compilation of speaker measurements taken in the Canadian NRC anechoic chamber. It's pretty easy to figure out which speakers will do well in the listening tests, just by looking at the frequency response graphs. Check out this typical B&W response, with the huge hole in the upper midrange caused by the mismatch of the too large midrange driver not blending well with the tweeter, and the lack of a waveguide contributing to the same problem (for reference, the B&W 704S2):
Compare this to a Revel speaker measurement. Revel speakers are designed based on the data coming out of the above mentioned research (for reference, the Revel F206):
As you might guess, the B&W sounds recessed and distant as compared to the Revel, especially in the critical vocal range.
Other speaker companies availing themselves of this research include PSB and KEF (though KEF has some issues with intermodulation distortion due to its UniQ design - the problem is that the woofer becomes the waveguide for the tweeter, and it's almost always in motion). Paradigm used to be engineered around the NRC research, but that changed when they became part of the Martin Logan conglomerate. With the adoption of this new CEA2034 by the industry, one can easily narrow down the list of speakers to audition by eliminating speakers that do nothing but distort the sound.
|
|
|
|
|
|
|
Wow, I wrote a novel! Sorry to take this so far off topic (though to be fair, MQA became a major topic of conversation).
|
|
|
|
|
|
|
 |
Posted: |
Mar 10, 2020 - 11:50 AM
|
|
|
|
By: |
Nono
(Member)
|
Good point about the actual dynamic range vs. the theoretical. Most real world listening environments have such a high noise floor that if you had a true 16 bits of dynamic range, you would have to constantly boost the volume control to hear the quieter passages, then have to turn it down drastically whenever the louder sections came blasting in. That's what led to the loudness wars to begin with - overcoming noisy listening environments in the car, on the street, etc. Without compression, all that low level detail got lost in the background noise of the listening space.
Ideally, we'd have a playback system that retained dynamic range while giving the user control over dynamics in order to adapt to the environment. For critical listening at home, full dynamic range. For listening on the go, the ability to add quality compression to taste.
On a good system, you don't need to adjust the volume for the quieter moments. I could have had such problems in the past, but not anymore. And a good system doesn't mean an expensive one, but one which sounds "credible" whatever the informations and dynamic range it can deliver.
Today, I almost always go back to the older editions because the mastering levels are louder and louder, even in film music. The La-La Land edition of Krull is louder than the Super Tracks which is louder than the SCSE which is louder than the Southern Cross first CD release.
When will it stop?
Since labels have to sell the same music again and again, I unfortunately know that the only way to give the listener the feeling that the new remastered edition is better than the previous one is to make it sound louder, more EQs, more polish, more beautiful, more spectacular than ever!
Which is nonsense.
|
|
|
|
|
|
|
|
|
|
|
How loud you need to adjust your system has very little to with quality, it has everything to do with the noise floor of the listening environment. The dynamic limit of CD is about 105 db. That's means 105 db of difference between the very softest sounds (say, a pianissimo woodwind solo) and the loudest sounds (the sound of the entire orchestra playing triple forte). Let's say that range is real world around 85 db in the concert hall.
Now, let's say you have a recording that captures all 85 db of that dynamic range, and you play that on your home system. Immediately you have a problem. A typical quiet listening room has about a 40 db noise floor (traffic noise, HVAC sounds, etc). To overcome that noise floor, you have to make sure that that oboe solo is already at 41 db just to be audible above the room noise. Then when the whole orchestra comes in at 84 db louder - now at a sound pressure level of 125 db - you might not only blow out your speakers, you might blow out your ears as well. Even theatrical reference level on a mix stage is 105 db MAXIMUM for sound level peaks. Much louder and you will sustain permanent hearing damage.
Now let's take that CD with a dynamic range of 85 db (well short of its theoretical maximum) and put it in the car, where the ambient noise is approximately 70 db. Now you can even more dramatically see the problem, and why you need to raise the volume level even more to hear the oboe solo over the din. And again, when the full orchestra comes in you'll get blasted out of your seat.
This is why a moderate level of compression under almost all listening circumstances can be desirable. I totally agree that the use of compression has gotten out of control and needs to reigned in, but purists wanting no compression at all are ignoring a large part of the picture. This is why I advocate user adjustable compression based upon listening environment. Reduce compression in quiet spaces, increase it in noisy spaces.
Smart mastering engineers understand both sides of the argument and apply compression judiciously, keeping all the above in mind.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Some of my friends are in the business, making good speakers is not an exact science. If it was the case, anybody could make great speakers.
That's the thing, in fact, there are some loudspeakers that are perfect for one type of listener (and/or room) but will sound terrible for another listener (and/or in another room).
Though of course, a flat frequency response and a "natural" sounding speaker is what just about everyone desires. I updated my system about a year ago and selecting loudspeakers was without a doubt the most difficult part. The other equipment is more or less a question of specs. And I am not an audiophile either, I listen under regular living room conditions, not in a laboratory. I'm just someone who likes to enjoy music. But the Harman White Paper is certainly very interesting, but indeed: I have heard some speakers with extremely flat frequency responses (when measured) that sounded terrible to my ears.
|
|
|
|
|
|
|
|
|
|
Some of my friends are in the business, making good speakers is not an exact science. If it was the case, anybody could make great speakers.
Of course, making the cabinet free of resonance is the basic of any good speakers. But music is also more than just frequencies, and what we can hear is also shaped by what we don't hear (and can't see on a graph).
Then how do you explain almost 40 years of published and peer reviewed research that shows almost exactly the opposite, that "good sound" from speaker can be quantified and measured, so accurately that they can predict with 86 - 99% confidence which speaker will win double blind listening tests just by looking at detailed measurements? (Note - the reason why the figure varies between 86 and 99% is due to bass response - a good measuring speaker with more bass extension will outperform an equally good speaker with less bass extension).
To be fair, and perhaps to your point, it's not a simple frequency response graph we are talking about. The standard I am discussing - CEA2034 (https://speakerdata2034.blogspot.com/2019/02/spinorama-cea-2034-2015-ansi-data-format.html) - relies on a specific set of measurements called a "Spinorama." It shows not only frequency response measurements on axis, but off axis as well (the research shows that flat frequency response off axis is almost as important as flat frequency response on axis). To get the Spinorama measurements, each loudspeaker is measured at 70 different angles around its horizontal and vertical orbits in order to fully characterize the quality of its on and off-axis sounds, and allow removal of acoustical interference effects from resonances, which can cause harmful colorations to the reproduced sound. In a CEA2034 / Spinorma graph, the frequency curves represent, from top to bottom, the quality of the direct sound, the average listening window, the first reflections, the sound power, and the directivity indices for the first reflections and the sound power. Case in point - here is the Spinorama measurements for the Revel F228Be, which just won Stereophile's Speaker of the Year award (not that such awards are always an indicator of quality):
The article by Kal Rubinson of Stereophile about his experiences at Harman evaluating the Spinorama / CEA2034 measurement system, and how well it works in terms of predicting listener preference (btw, the mystery speaker in the shootout was the Paradigm Persona 3F):
https://www.stereophile.com/content/blind-listening-harman-international
Anyone can make good speakers, if they have the skills, knowledge and resources at their disposal. A major problem is that many boutique speaker manufacturers do not have the resources to build anechoic chambers or invest in a full suite of Klippel speaker measurement software, which is what is used to generate the Spinorama graphs.
Documentation, references, and subreferences on the Canadian NRC and Harman research, which resulted in the CEA2034 industry accepted standard:
https://seanolive.blogspot.com/2008/12/part-3-relationship-between-loudspeaker.html
Hope some find this information useful. I know many of the engineers and scientists involved, and have done lectures on the subject (like this one, lol). 
|
|
|
|
|
|
|
How loud you need to adjust your system has very little to with quality...
When the problem is heard in a quiet room, it comes from the quality of the system and not from the surrounding noise.
Some systems don't have any dynamic gradation between the lower and the louder moments, which is disturbing. If the sound is full with a natural dynamic gradation from the softest to the loudest moments, a wide dynamic range is not really a problem.
And the fuller and more audible the softest moments are, the less you have to raise the volume, the more you can benefit from the dynamic range, which is also related to the quality of the system.
It's what I meant.
Got it, and agree
|
|
|
|
|
|
Some of my friends are in the business, making good speakers is not an exact science. If it was the case, anybody could make great speakers.
That's the thing, in fact, there are some loudspeakers that are perfect for one type of listener (and/or room) but will sound terrible for another listener (and/or in another room).
Though of course, a flat frequency response and a "natural" sounding speaker is what just about everyone desires. I updated my system about a year ago and selecting loudspeakers was without a doubt the most difficult part. The other equipment is more or less a question of specs. And I am not an audiophile either, I listen under regular living room conditions, not in a laboratory. I'm just someone who likes to enjoy music. But the Harman White Paper is certainly very interesting, but indeed: I have heard some speakers with extremely flat frequency responses (when measured) that sounded terrible to my ears.
Actually, it's more than a white paper - it's decades of research that has been validated externally over and over through standard scientific peer review procedures. Floyd and Sean have tested literally thousands of speakers during their years in Canada and in CA, and all the tests have supported these conclusions.
One of the things they discovered is that a bad speaker will sound bad in any room, and a good speaker will sound better in any room. The ranking scores do change, but always relative to one another. In other words:
Bad speaker in a good room - rating 4 out of 10
Bad speaker in a bad room - rating 2 out of 10
Good speaker in a good room - rating 9 out of 10
Good speaker in a bad room - rating 6 out of 10
RE: flat measurements. Who was taking the measurements and what was the resolution? Were they taking into account off axis response as well as overall sound power in the room? Most speaker measurements are worthless, since they are usually low resolution and are rendered with a great deal of smoothing. A speaker can have great, flat on axis response but very poor off axis response, and the end result is a poor sound speaker (this is why it used to be so important to absorb side wall reflections).
This is why the Spinorama measurement is so important - it quantifies all of these variables.
One of the things I offer here is for anyone to bring in any speaker regardless of price range, and I will shoot it out blind with a speaker known to do well in Spinorama graphs. In almost 5 years, each time the predictive power of the measurements was borne out. There's a huge thread about the shootout we did between the Revel Salon2 and the JBL M2 Master Reference Monitor on the AVS Forum, two speakers that do extremely well in Spinorama tests. It was almost a dead heat! You can read it here:
https://www.avsforum.com/forum/89-speakers/2907816-speaker-shootout-two-most-accurate-well-reviewed-speakers-ever-made.html
One of the most viewed threads in the history of the AVS Forum, which is in itself one of the most read forums on the internet.
BTW, here is the Spinorama of the JBL M2 monitor, along with a picture of Bruce Botnick, who mixes on the JBL M2:
https://jblpro.com/zh/news/legendary-engineer-and-producer-bruce-botnick-outfits-his-studio-with-three-of-harman-s-jbl-professional-m2-master-reference-monitors
See! I did bring it back around to THE SWARM
|
|
|
|
|
|
|
|
|
|
Posted: |
Mar 10, 2020 - 2:15 PM
|
|
|
|
By: |
Nono
(Member)
|
Then how do you explain almost 40 years of published and peer reviewed research that shows almost exactly the opposite, that "good sound" from speaker can be quantified and measured, so accurately that they can predict with 86 - 99% confidence which speaker will win double blind listening tests just by looking at detailed measurements? (Note - the reason why the figure varies between 86 and 99% is due to bass response - a good measuring speaker with more bass extension will outperform an equally good speaker with less bass extension).
I said music is not "just" frequencies. There's something more that can't be quantified.
Anyone can make good speakers, if they have the skills, knowledge and resources at their disposal. A major problem is that many boutique speaker manufacturers do not have the resources to build anechoic chambers or invest in a full suite of Klippel speaker measurement software, which is what is used to generate the Spinorama graphs.
Tools are nice, and are needed at some point, but at the end it's only the ear of the designer/maker that will make the difference, with adjustments which won't be on paper, and which could be different from what the science has to say.
Hope some find this information useful. I know many of the engineers and scientists involved, and have done lectures on the subject (like this one, lol).
It's very interesting, thank you!
|
|
|
|
|
|
|
|
|
|
Posted: |
Mar 10, 2020 - 2:24 PM
|
|
|
|
By: |
CCW1970
(Member)
|
And expensive systems doesn't mean these systems are good.
One of my friend is an electronic engineer who makes amplifiers (he made mine) and actually set up a small Hi-Fi brand which also makes speakers. From my experience, many expensive systems are a joke, and they actually don't worth their price.
Hi-Res can brings subtleties that are lost by most systems, including the most expensive, whoever made them.
Stradivarius and Guarneri were not "Dr." and certainly better knew how to make sound a bit of wood. The same in Hi-Fi, the best speakers I have ever heard were made by a cabinetmaker.
"And expensive systems doesn't mean these systems are good."
Indeed! Are you familiar with Harman's MLL? It's a double blind listening chamber that features a pneumatic speaker shuffler, allowing listeners to precisely match speaker location and volume. They have found over almost 40 years of evaluating speakers in a double blind fashion that some of the worst sounding speakers were also some of the most expensive. Most of this research correlating listener preference with speaker measurements was funded and accomplished years ago by the Canadian government; it was refined more recently at Harman's labs in Northridge CA.
My friend works for Harman, and I've been in that room. It's really, really cool, and very illuminating. It's a great way to get rid of one's preconceived notions about speakers.
|
|
|
|
|
|
|
|
|
|
|
|
| |
|
|
|
|
|
|
|
|
|
