When is hi-rez overkill?

[Almen]

Yeah, that would be cool to test!

I guess you would have to be careful with the dithering when converting 24/16/24, so we are not listening to some sloppy algorithms instead.

I am planning to do some testing similar to the report from JAES, that is, inserting an A/D/A 44.1 kHz/16 bits loop into an arbitrary analog signal. I am waiting for some A/B/X equipment, though, so this might not be until next year.

 

[ssully]

24/96 vs 16/44 has been done at least once or twice on hydrogenaudio.org. Also you can download what claims to be true 24/96 here and do the test yourself

http://www.soundkeeperrecordings.com/format.htm

HA thread (there are others too):

http://www.hydrogenaudio.org/forums/index.php?showtopic=74123

Note the detailed post from 'Werner' about the hidden assumptions and differences involved in the sort of test you propose.

 

[daved64]

It's a good idea to keep in mind that the audiophile-nonsense world has a vested interest in expensive, limited-edition things like 50-gram vinyl discs, because that's what they're selling -- specialty.

Wow, that's one skinny LP!

 

[Key]

24/96 vs 16/44 has been done at least once or twice on hydrogenaudio.org. Also you can download what claims to be true 24/96 here and do the test yourself

http://www.soundkeeperrecordings.com/format.htm

HA thread (there are others too):

http://www.hydrogenaudio.org/forums/index.php?showtopic=74123

Note the detailed post from 'Werner' about the hidden assumptions and differences involved in the sort of test you propose.

Everything Werner has said has already been taken into consideration by me. I just did not want to bore you guys - leave the technical junk up to me and the conversion. I can't tell if Werner is also overlooking that settings other than linear phase are used because they have been subjectively judged to be the most audibly transparent even though measurably they introduce a phase distortion.

Barry on the other hand seems to think that the 44.1 16-bit should be kept at that resolution when comparing to 96kHz 24-bit. There are a lot of problems with this and it would make the test require much tighter controls imho. You would have to measure the output of the DAC in both settings and they would have to match in terms of FR, THD, IMD, and Phase within a certain margin of error. Also switching between the two formats introduces intermediate switching distortions which make fast ABing impossible.

The idea of comparing untainted 96kHz 24 bit vs 96 24->44.1 16->96kHz 24-bit does throw an arguably unneeded step into the chain but it is the real world compromise that has the least tradeoffs and is easily reproducable since you only need a DAC which is decent in 96kHz and not one which is decent in both 44.1 and 96kHz - much much harder to find.

 

[ssully]

Everything Werner has said has already been taken into consideration by me. I just did not want to bore you guys - leave the technical junk up to me and the conversion. I can't tell if Werner is also overlooking that settings other than linear phase are used because they have been subjectively judged to be the most audibly transparent even though measurably they introduce a phase distortion.

Barry on the other hand seems to think that the 44.1 16-bit should be kept at that resolution when comparing to 96kHz 24-bit. There are a lot of problems with this and it would make the test require much tighter controls imho. You would have to measure the output of the DAC in both settings and they would have to match in terms of FR, THD, IMD, and Phase within a certain margin of error. Also switching between the two formats introduces intermediate switching distortions which make fast ABing impossible.

The idea of comparing untainted 96kHz 24 bit vs 96 24->44.1 16->96kHz 24-bit does throw an arguably unneeded step into the chain but it is the real world compromise that has the least tradeoffs and is easily reproducable since you only need a DAC which is decent in 96kHz and not one which is decent in both 44.1 and 96kHz - much much harder to find.

Key,

I quite agree that it's the best compromise, but was just noting the possible factors of an ABX should it turn out to be 'positive' -- as a rare one or two has on HA. Then drilling down into the technical details of e.g. filters and dither becomes necessary.

Negative ABX result of course also has its own set of possible explanations depending on how rigorous the test was, e.g. inadequate training, no positive control, inadequate equipment or source material.

But what interests me is when listeners claim to hear big differences sighted between a certain A and B, and that difference fails to be supported, with the same listener, the same A and B, using the same equipment, under blind conditions. That, I think , at least indicates that the listener was probably hearing his own imagination and expectation bias. Which is very, very common.
(It does not rule out that someone , somewhere, might detect a real difference between A and B)

 

[Key]

Yes this is what I think as well. It's not so much that they don't hear it though in these cases of differences heard sighted and them going away with ABX - they actually do in some situations. But I would personally attribute this to something psychosomatic and related to the overlapping of the senses and it's utility in proprioception. That is a whole other hypothesis though ha.

Here is something I have personally found with ABX comparisons of small measurable differences. I can actually ABX very small hard to hear differences if I go through the test very fast and trust my senses when I hear a difference that I am looking for - I am usually looking for this difference based on the way they measure. In some of these situations when I go through very fast I am able to take the ABX down to a very low percentage of guessing. The thing is as time goes on I honestly can not perceive the differences with certain things like measurable phase distortions. I think there is something cognitive taking place where my mind is actually working to make A match B or B match A. Don't hold me to that though I still need to test more and make sure I wasn't just getting lucking when I get the ABXing right. I make a rule not to guess though so at the time I hit the button I do believe I am hearing a difference. As soon as I can't I stop hitting the button because I don't know which is which.

 

[daved64]

But what interests me is when listeners claim to hear big differences sighted between a certain A and B, and that difference fails to be supported, with the same listener, the same A and B, using the same equipment, under blind conditions. That, I think , at least indicates that the listener was probably hearing his own imagination and expectation bias. Which is very, very common.

I'm nowhere near as smart or technical as you guys, but would you not agree that the whole audiophilia disease is based on imagination and expectation. We WANT to believe that the bitchin' cables we bought for $40 a foot are making a difference, even though lamp cord would probably do the trick. We WANT to believe that the SACD we're playing sounds 1000 times better than the well mastered CD, because dammit, we're told it does!

 

[ssully]

Yes this is what I think as well. It's not so much that they don't hear it though in these cases of differences heard sighted and them going away with ABX - they actually do in some situations. But I would personally attribute this to something psychosomatic and related to the overlapping of the senses and it's utility in proprioception. That is a whole other hypothesis though ha.

Here is something I have personally found with ABX comparisons of small measurable differences. I can actually ABX very small hard to hear differences if I go through the test very fast and trust my senses when I hear a difference that I am looking for - I am usually looking for this difference based on the way they measure. In some of these situations when I go through very fast I am able to take the ABX down to a very low percentage of guessing. The thing is as time goes on I honestly can not perceive the differences with certain things like measurable phase distortions. I think there is something cognitive taking place where my mind is actually working to make A match B or B match A. Don't hold me to that though I still need to test more and make sure I wasn't just getting lucking when I get the ABXing right. I make a rule not to guess though so at the time I hit the button I do believe I am hearing a difference. As soon as I can't I stop hitting the button because I don't know which is which.

ABX is very sensitive...level differences down to the physiological limit of hearing (molecules banging randomly against eardrums) can be detected with it. However, there are plenty of measurable differences that cannot rationally be expected to be audible. Phase is not necessarily easy to hear, so I would not be disappointed if it was not foudn in your ABX; but also don't be surprised if you get fatigued doing serious blind comparisons for more than 30 min or so. If you do, just stop and resume at another time, but be sure to compile all the results, not just cherry-picked.

http://www.hydrogenaudio.org/forums/index.php?showtopic=16295

 

[simonwagstaff]

I'm nowhere near as smart or technical as you guys, but would you not agree that the whole audiophilia disease is based on imagination and expectation. We WANT to believe that the bitchin' cables we bought for $40 a foot are making a difference, even though lamp cord would probably do the trick. We WANT to believe that the SACD we're playing sounds 1000 times better than the well mastered CD, because dammit, we're told it does!

I promised myself I would stay away but I guess it really doesn't matter. do you think that an entire industry is based upon delusional customers? Perhaps. There are certainly A LOT of pretty silly things out there. Have you ever listened to a system that had these silly bitchin' cables? Maybe. This is by no means an endorsement. Just recently I have been playing with cables and thought I would check out some of these silly speaker cable. Audioquest CV-4 speaker cables and Colorado interconnects to run between my processor and tube amp. Why these things are so silly that they have battery packs on them to set up a bias field around the wires to improve shielding and who knows what else. These were products of the year named by Absolute Sound for whatever that is worth. I wasn't about to spend $1200 bucks on this set though that is what their list price is. I was able to get some on E-bay for what I considered a reasonable price, a fraction of that, a small fraction.

Without going into detail, this is one of the best audio purchases I have made in a long time.

I really didn't want to hear a difference, for what I spent it wouldn't have mattered if I heard an improvement, but I did.

There is actually a lot of science and engineering that goes into some of this stuff, delusional or not.

I still think there is some magic in audio and my goosebumpometer tells me so.

 

[Key]

I'm nowhere near as smart or technical as you guys, but would you not agree that the whole audiophilia disease is based on imagination and expectation. We WANT to believe that the bitchin' cables we bought for $40 a foot are making a difference, even though lamp cord would probably do the trick. We WANT to believe that the SACD we're playing sounds 1000 times better than the well mastered CD, because dammit, we're told it does!

Well it's hard to say in all cases really. Many things can change the sound including cables which are basically designed to do such things by manipulating capacitance, inductance, or resistance. But does that mean you are approaching the most accurate reproduction because you heard a difference and it subjectively sounded better? Not necessarily. Honestly it seems as if a lot of audiophiles are trying to use a cable as a missing component of the passive crossover based on the false assumption that this is somehow more pure than using active electronics with tone controls.

To me audiophilia is made up of a combination of similar false assumptions and expectational biases. As well as in a lot of instances rational people reacting to superfluous information - things we can measure but not necessarily hear.

ABX is very sensitive...level differences down to the physiological limit of hearing (molecules banging randomly against eardrums) can be detected with it. However, there are plenty of measurable differences that cannot rationally be expected to be audible. Phase is not necessarily easy to hear, so I would not be disappointed if it was not foudn in your ABX; but also don't be surprised if you get fatigued doing serious blind comparisons for more than 30 min or so. If you do, just stop and resume at another time, but be sure to compile all the results, not just cherry-picked.

http://www.hydrogenaudio.org/forums/index.php?showtopic=16295

I was using that particular example because I didn't exactly know what to do with the results. The results looked like I could successfully tell the difference in phase but it was only because I stopped hitting the button as soon as I stopped hearing the difference. So in reality I both heard a difference and didn't. Mind you this was very fast like within the first 2 minutes of testing phase linear vs phase distorted on my 2 ways and I did cherry pick the song in that I thought it would be an example that would make it easier to hear the difference.

I am planning on going back to the test soon. But honestly the difference that I could detect seemed insignificant in the scheme of things. It wasn't a huge improvement.

 

[ssully]

I promised myself I would stay away but I guess it really doesn't matter. do you think that an entire industry is based upon delusional customers? Perhaps. There are certainly A LOT of pretty silly things out there. Have you ever listened to a system that had these silly bitchin' cables? Maybe. This is by no means an endorsement.

Wow. Of all the things to support your argument: audio cables?? Really?? Not exactly a technology in its infancy.

Just recently I have been playing with cables and thought I would check out some of these silly speaker cable. Audioquest CV-4 speaker cables and Colorado interconnects to run between my processor and tube amp. Why these things are so silly that they have battery packs on them to set up a bias field around the wires to improve shielding and who knows what else. These were products of the year named by Absolute Sound for whatever that is worth.

That would be: nothing.

I wasn't about to spend $1200 bucks on this set though that is what their list price is. I was able to get some on E-bay for what I considered a reasonable price, a fraction of that, a small fraction.

Without going into detail, this is one of the best audio purchases I have made in a long time.

I really didn't want to hear a difference, for what I spent it wouldn't have mattered if I heard an improvement, but I did.

There is actually a lot of science and engineering that goes into some of this stuff, delusional or not.

'Sciency' web pages and brochures != science. Pseudoscience, perhaps.

I still think there is some magic in audio and my goosebumpometer tells me so.

Interesting. Look what happens when I see 'battery packs' attached to audio cables:

 

[ssully]

Mind you this was very fast like within the first 2 minutes of testing phase linear vs phase distorted on my 2 ways and I did cherry pick the song in that I thought it would be an example that would make it easier to hear the difference.

Nothing wrong with picking a song you think will expose the difference best. That's a good idea. I was referring to cherry picking ABX results...as described in that HA post.

 

[simonwagstaff]

Whateva!!!!

 

[timbre4]

I'm sure glad this thread is keeping the other threads safe!

 

[Almen]

ssully and key: Thanks, great info.

timbre4: I am sure glad this thread exists at all, so I can learn more about ABX testing. You too keep coming back.

 

[daved64]

I promised myself I would stay away but I guess it really doesn't matter. do you think that an entire industry is based upon delusional customers? Perhaps. There are certainly A LOT of pretty silly things out there. Have you ever listened to a system that had these silly bitchin' cables? .

Well, the biggest industry in the universe is based on an imaginary guy in the clouds that hears and sees all and is keeping track of your every move, ready to cast you into either eternal bliss or eternal damnation. Audiophelia is a close second.

If YOU think you hear a difference then that should be good enough for you, but I fell victim to many of these "tweaks", only to sit there and say "I don't hear a fucking thing different". I am blessed with very good ears (my wife begs to differ). Spikes on speakers, no difference (other than changing overall driver height which may cause a "difference", but not necessarily an improvement). Home made CAT-5 cables. No difference. Even tried the ridiculous "ball bearing" tweak that was supposed to bring me to orgasm. Only thing it did was make my speakers vulnerable to tipping over.

What DID help was bass traps, and cutting down on reflections. These are tangible and proven tweaks where one can cite many examples of hard science to back them up.

Oh well, at least this all gives life back to this thread.

 

[Almen]

Well, the biggest industry in the universe is based on an imaginary guy in the clouds that hears and sees all and is keeping track of your every move, ready to cast you into either eternal bliss or eternal damnation. Audiophelia is a close second.

Hehee, sorry, had to chuckle a little, there.

What DID help was bass traps, and cutting down on reflections. These are tangible and proven tweaks where one can cite many examples of hard science to back them up.

Hear, hear! I too, have this experience - moving speakers, adding carpets and curtains, and smoothing the base will really change the sound, whereas esoteric tweaks most of the time will not.

 

[Key]

Even tried the ridiculous "ball bearing" tweak that was supposed to bring me to orgasm. Only thing it did was make my speakers vulnerable to tipping over.

Haha so he got you with the hip-joints eh? You are more open minded than me I will give you that. I suspect I wouldn't hear a difference on my setup as well. I think I already get the same effect from these cheap samson stands to a lesser extent. When I push on the speaker it moves around back and forth like it is spring loaded. All you need are metal stands with a little bit of elasticity to the poles.

Ah but the tiny resonant frequencies and not totally decoupled etc.. To me none of that really matters because sound travels through the air and hits solids - speakers - directly thought the air. You are not going to stop sound from hitting the speakers completely because sound travels through all matter - air, solid, plasma, liquid.

What DID help was bass traps, and cutting down on reflections. These are tangible and proven tweaks where one can cite many examples of hard science to back them up.

This is where I start to part ways with most all of the audio community. I think that beyond removing early reflections, getting rid of resonant objects, and fine tuning your speaker placement - getting them at least 2 or 3 feet away from the wall - that room treatment is another vain attempt to control things which are so complex we can't accurately predict. I think room response is mostly misunderstood in the audio community and there are a few people who actually get what is going on in relation to a sound environment with countless reflections of reflections and how your brain processes that information and discards the sounds in the room which aren't useful.

I will not deny that room treatment can change the sound of a room. But I have serious doubts that it is a prerequisite for an accurate illusion of a sound stage or that by somehow absorbing the late reflections in a room through the use of human technology it will get me any closer to accuracy.

Siegfried Linkwitz is one of the most outspoken advocates for the type of systems I think will work independent of room treatment. But to tell the truth I am not even as perfectionist as him with some of these things. But here is a great lecture about what he has found in terms of speakers that will work independent of the room's character.
http://www.linkwitzlab.com/publications.htm#23
http://www.zshare.net/audio/6941674293626fd8/
http://www.linkwitzlab.com/AES-London'07/AESUK_lecture_0711.pdf

the lecture is a little long but I think he makes a lot of sense

Another person who has done research along these lines is Earl Geddes
http://www.gedlee.com/Loudspeakers.htm

To put it simply I think that if you use speakers which are either Controlled Directivity, Constant Directivity, Omni, Dipole, or Cardiod that you should not need room treatment in most situations.

 

[G E]

.... What DID help was bass traps, and cutting down on reflections....

Agreed. Adding floor to ceiling bass traps in 4 corners and addressing reflection points with absorption panels TRANSFORMED my listening experience. The only things measured were the materials I cut to build the devices, but the Grinmeter (similar to Wagstaff's Goosebumpometer) has been pegged ever since.

 

[G E]

...
What DID help was bass traps, and cutting down on reflections.

...


Other major improvements ocurred when I upgraded my power amp, changed tubes in my preamp and upgraded source components (now a Pioneer Elite 79 avi). My first hi-res digital player was a stock Pioneer 563a and it was thoroughly underwhelming no matter if redbook, sacd or dvd-a. But it did open the world of multi-channel playback of music. After stiffening the power supply AND swapping out the OPamps it made delightful music. I bought a second 563a and modded it but used different OPamps. More delightful music but with a sonic signature completely different than the first: Incredible PRT and slightly forward presentation versus a wonderfully full, almost "smokey" warmth that has a special something for female vocalists and acoustic bass but not the last word in resolution. And with these two players I could finally appreciate the benefit of hi-res digital compared to red book.