Interesting listening experiences with a few CD players

[Scytales]

There is no One True Definition of "proper", it's good to use our brains to try different methods to counteract biases. Were you aware of this work by Amir, circa 2009? https://www.avsforum.com/threads/establishing-differences-by-the-10-volume-method.1136745/ In it, he deliberately tweaked the volume of one source or other to help him GET RID OF BIASES and detect differences.

I just read the AVS forum thread, and I am afraid that amirm, for whom I have the greatest respect, deserved the criticism that many people addressed to him.

But to my mind, the most important point was not even addressed in the AVS forum thread, whereas amirm has stated it in his very first post (§4) on the other forum:

'I [...] picked material that made it easier to detect differences between DACs. I am not going to disclose what constitutes such content.'

That means that the substance of the listening test he conducted was never fully disclosed, making it impossible to debate it in the first place, however valid his methodology could have been.

As far as your and your son's evaluations are concerned, you have described the sound of one CD player when playing random CDs (not CDs with program material likely to generate a known detectable artifact) as having an 'etched quality, an edge, to the sound' (you) and the other one as having a 'softer' sound (your son).

To me, this kind of qualification may very well be the products of a very small difference in level playback. Once you and your son have been convinced to hear this sound qualities, it is very much possible that you were both biased to perceived them even when the volume of one player was knowingly changed. This is where a listening test after level matching would be useful, in order to confirm or not your assessment of the sound quality of each of the two CD players.

I would add that the description you gave about the sound evaluation of the pro CD player (as having 'less detail retrieval, and its sound had music notes which were kind-of one-note. For instance, the hard drumming in some of the Supertramp tracks sounded like "one note" from Player C and had layering and textural detail') confirms to me that the possibility of very small differences in the output levels between each three players has to be checked, because such subjective assessments of sound quality may typically be the product of subtle discrepancies in playback levels.

 

[tcpip]

To me, this kind of qualification may very well be the products of a very small difference in level playback.

I have explained that I neutralised the possibility of these observations being gain dependent by actually changing the gain of one of the players over a range greater than 2dB. The theory says that if I "like" one device's sound less, and find it rounded, softer or duller, then by increasing the gain, this difference will disappear. I tried this, and by a far wider margin than what the blind-testing experts say. I both raised and lowered the gain of one of the players, and the softer sound remained softer, the edgy remained edgy.

I would add that the description you gave about the sound evaluation of the pro CD player (as having 'less detail retrieval, and its sound had music notes which were kind-of one-note. For instance, the hard drumming in some of the Supertramp tracks sounded like "one note" from Player C and had layering and textural detail') confirms to me that the possibility of very small differences in the output levels...

I checked this too, by increasing the gain when playing back from this player. The characteristics of the sound remained the same. I tried "hearing into the music" with this player by increasing the gain by more than 1dB, and the characteristics remained unchanged.

So I understand that this is not the textbook way to compare players, but I think I took strong measures to neutralise any effect of mismatched gain.

 

[Yuhasz01]

I do not doubt YOU heard a difference(ie,various bias, room properties, , your hearing ). But most people here would like to know if differences go beyond such subjective listening/perception , in realm of objective ,controlled measurements.

 

[Purité Audio]

I have explained that I neutralised the possibility of these observations being gain dependent by actually changing the gain of one of the players over a range greater than 2dB. The theory says that if I "like" one device's sound less, and find it rounded, softer or duller, then by increasing the gain, this difference will disappear. I tried this, and by a far wider margin than what the blind-testing experts say. I both raised and lowered the gain of one of the players, and the softer sound remained softer, the edgy remained edgy.


I checked this too, by increasing the gain when playing back from this player. The characteristics of the sound remained the same. I tried "hearing into the music" with this player by increasing the gain by more than 1dB, and the characteristics remained unchanged.

So I understand that this is not the textbook way to compare players, but I think I took strong measures to neutralise any effect of mismatched gain.

The only really valid comparison is to level match to .1dB and compare unsighted.
Keith

 

[tcpip]

I do not doubt YOU heard a difference(ie,various bias, room properties, , your hearing ). But most people here would like to know if differences go beyond such subjective listening/perception , in realm of objective ,controlled measurements.

In terms of objective, controlled listening tests, yes, not necessarily measurements.

 

[DSJR]

In terms of objective, controlled listening tests, yes, not necessarily measurements.

I've been where you are (believe me, I based most of my career in the industry on it) and today, using really good head/earphones, so much more can be heard as compared to the average pair of passive speakers in a barely or untreated room. Maybe it's my age and current mindset, but I'd rather hear more into the music mix today, rather than into any supposed or presumed sound of the gear upstream...

I think the OP is going to have difficulty with many/most ASR posters, but what would gain real interest, is if a 0.1dB matched and UNSIGHTED A-B comparison was made, the switching by a third party and differences then being consistently heard. Not to spend twenty minutes frantically switching from one to the other, but a bit at a time.

It's an interesting journey for some, but one of the earlier experiences which made me question almost all the programming I had acquired as a dealer, was going to a good few live largely unamplified jazz gigs - not my chosen music genre, but the 'feel' and the 'vibe' as well as judging the raw sound of a grand piano, brass and drums (OUCH if sat too close). This helped no end in 'judging' mere audio gear, which is simply a tool to hear various recordings via... Apologies for being a boring old f**t here.

 

[IPunchCholla]

I have explained that I neutralised the possibility of these observations being gain dependent by actually changing the gain of one of the players over a range greater than 2dB. The theory says that if I "like" one device's sound less, and find it rounded, softer or duller, then by increasing the gain, this difference will disappear. I tried this, and by a far wider margin than what the blind-testing experts say. I both raised and lowered the gain of one of the players, and the softer sound remained softer, the edgy remained edgy.


I checked this too, by increasing the gain when playing back from this player. The characteristics of the sound remained the same. I tried "hearing into the music" with this player by increasing the gain by more than 1dB, and the characteristics remained unchanged.

So I understand that this is not the textbook way to compare players, but I think I took strong measures to neutralise any effect of mismatched gain.

Bias doesn’t work that way. Nor can it be controlled that way.

I have sighted AB tested different PEQs when I was setting up my subwoofer. One was obviously better than the other, It also would change quite a bit adjusting subwoofer gain +-6db to really dial it I.

I spent quite a bit of time playing with the sound before I realized the subwoofer wasn’t even on.

The sound difference was obvious to me. No need to test it. It also was manufactured by my senses, not in the audio signal.

 

[kemmler3D]

Bias doesn’t work that way. Nor can it be controlled that way.

I have sighted AB tested different PEQs when I was setting up my subwoofer. One was obviously better than the other, It also would change quite a bit adjusting subwoofer gain +-6db to really dial it I.

I spent quite a bit of time playing with the sound before I realized the subwoofer wasn’t even on.

The sound difference was obvious to me. No need to test it. It also was manufactured by my senses, not in the audio signal.

This exactly has happened to me a few times also. The only clue that something was wrong was that the difference was smaller than expected...

Things you hear via placebo effect are not only genuinely heard differences, they are totally convincing until you prove to yourself beyond doubt they're not real.

 

[Chrispy]

I too am not inclined to listen on headphones if I get speakers as an option. I use earphones always whenever I want to focus on checking the sound quality of the source media or source component -- it removes the (relatively imprecise) speakers and room from the path.

I've recently bought good headphone amps, and I always had a good pair of reference monitoring earphones (ER4PT), now I have 3 more pairs, so this chain is absolutely superb to let me "look into" the music.

Like I said, even headphone comparisons can be problematic. Still boils down to your preference for the most part than any demonstration of a notable difference.

 

[pinger]

This thread would get a much more favorable response in a magazine like Stereophile

 

[tcpip]

Like I said, even headphone comparisons can be problematic. Still boils down to your preference for the most part than any demonstration of a notable difference.

I don't understand this remark at all about "demonstration of a notable difference". I am not discussing whether I like Player A or Player B, in case I was unable to explain in my original post. I am discussing whether there is an audible difference at all. I believe there is. Preferences only come into the picture after one accepts that there are differences.

 

[pinger]

I don't understand this remark at all about "demonstration of a notable difference". I am not discussing whether I like Player A or Player B, in case I was unable to explain in my original post. I am discussing whether there is an audible difference at all. I believe there is. Preferences only come into the picture after one accepts that there are differences.

Double blind test that proves a detectable sonic difference would be an example of that. You havent shown that.

 

[tcpip]

Double blind test that proves a detectable sonic difference would be an example of that. You havent shown that.

True. I haven't shown that. Have I claimed anywhere that I have?

Look, you and some others seem to be reacting to my posts as if I am appearing for an interrogation. I thought we could all look at this episode as an interesting episode, not put anyone on a dissection table and demand accountability. The episode is under discussion, not I. Are you capable of switching your attitude?

In the meantime, I've prepared a set of FLAC files which I will burn onto a CD-R, play the whole CD back through each CD player, record the output at 24/96, and upload them again to let all of you see if there are audible differences.

Dropbox

www.dropbox.com

 

[pinger]

True. I haven't shown that. Have I claimed anywhere that I have?

Look, you and some others seem to be reacting to my posts as if I am appearing for an interrogation. I thought we could all look at this episode as an interesting episode, not put anyone on a dissection table and demand accountability. The episode is under discussion, not I. Are you capable of switching your attitude?

In the meantime, I've prepared a set of FLAC files which I will burn onto a CD-R, play the whole CD back through each CD player, record the output at 24/96, and upload them again to let all of you see if there are audible differences.

Dropbox

www.dropbox.com

You seem to be taking the questioning personally. Nobody attacked you. We arent discussing you. We are discussing the issue you brought up.

 

[tcpip]

Thank you for the clarification but I know how to recognize when the questioning is personal.

 

[tcpip]

I burned the sampler CD last night. If I get time tonight, I'll play back the CD on both players and record their analog outputs in 24/96 FLAC. Let's see what Deltawave and listening tests throw up, if anything.

Two friends dropped in today in the middle of the day, and I made them listen to the same track on the two players, without telling them anything. They both said that one sounded softer, more "rounded" than the other, and both identified Player B as the one with the softer sound. One of these friends is a designer of awesome Class B amps, and is my only friend who owns an AP analyser -- the AP515, I think. He's struggling with his own observations about audible differences between power amps -- his power amps are regularly getting rave reports from friends who are comparing them with other big brand amps. (I know of comparisons against the Parasound A21+, some other Parasounds, and an Anthem 2-channel amp.) So he too is trying to understand what causes these differences -- they do not seem to be related to distortion and SNR figures.

He listened to my two CD players and immediately asked me whether they were level matched. They weren't, of course, so I invited him to turn up the volume of the softer-sounding Player B to see if the difference went away. That's the theory, right? The sound becomes more dynamic, sharper, clearer, when the SPL goes up. He turned up the headphone amp volume slowly right up to 6dB higher, and brought it down, and gave up. He admitted that the softer sound remained softer even at clearly louder volumes.

 

[majingotan]

I burned the sampler CD last night. If I get time tonight, I'll play back the CD on both players and record their analog outputs in 24/96 FLAC. Let's see what Deltawave and listening tests throw up, if anything.

Two friends dropped in today in the middle of the day, and I made them listen to the same track on the two players, without telling them anything. They both said that one sounded softer, more "rounded" than the other, and both identified Player B as the one with the softer sound. One of these friends is a designer of awesome Class B amps, and is my only friend who owns an AP analyser -- the AP515, I think. He's struggling with his own observations about audible differences between power amps -- his power amps are regularly getting rave reports from friends who are comparing them with other big brand amps. (I know of comparisons against the Parasound A21+, some other Parasounds, and an Anthem 2-channel amp.) So he too is trying to understand what causes these differences -- they do not seem to be related to distortion and SNR figures.

He listened to my two CD players and immediately asked me whether they were level matched. They weren't, of course, so I invited him to turn up the volume of the softer-sounding Player B to see if the difference went away. That's the theory, right? The sound becomes more dynamic, sharper, clearer, when the SPL goes up. He turned up the headphone amp volume slowly right up to 6dB higher, and brought it down, and gave up. He admitted that the softer sound remained softer even at clearly louder volumes.

Until there's a deltawave null test analysis, I would just conclude that it's sighted listening bias (a subconscious/uncontrolled bias association with perceived sound attribute to a gear based on its appearance and knowledge that it's that component that's currently playing) and lack of accurate volume matching based on your anectodal experience

 

[JeremyFife]

I burned the sampler CD last night. If I get time tonight, I'll play back the CD on both players and record their analog outputs in 24/96 FLAC. Let's see what Deltawave and listening tests throw up, if anything.

Two friends dropped in today in the middle of the day, and I made them listen to the same track on the two players, without telling them anything. They both said that one sounded softer, more "rounded" than the other, and both identified Player B as the one with the softer sound. One of these friends is a designer of awesome Class B amps, and is my only friend who owns an AP analyser -- the AP515, I think. He's struggling with his own observations about audible differences between power amps -- his power amps are regularly getting rave reports from friends who are comparing them with other big brand amps. (I know of comparisons against the Parasound A21+, some other Parasounds, and an Anthem 2-channel amp.) So he too is trying to understand what causes these differences -- they do not seem to be related to distortion and SNR figures.

He listened to my two CD players and immediately asked me whether they were level matched. They weren't, of course, so I invited him to turn up the volume of the softer-sounding Player B to see if the difference went away. That's the theory, right? The sound becomes more dynamic, sharper, clearer, when the SPL goes up. He turned up the headphone amp volume slowly right up to 6dB higher, and brought it down, and gave up. He admitted that the softer sound remained softer even at clearly louder volumes.

Hi,
If I may: I'm quite happy to accept that you and your friends hear a difference. I find your post interesting.

However, I have no way of knowing if there is actually a difference in the sounds that you hear. That's the root of most of the comments that you are getting, there's no verifiable evidence.

Looking forward to how this thread develops when you share the recordings.

 

[tcpip]

I will try to share the recordings as soon as I get time to do the recordings.

 

[tcpip]

Recordings done.

Dropbox

www.dropbox.com

These are large FLAC files. Each is a 24/96 recording. I have not trimmed the silences at the start and end of the FLAC files. When trimmed, there will be 64 mins of audio.

I played the same physical CD in all three sessions, without moving or reconfiguring anything in between. I configured Audacity to record samples in 24-bit integer instead of 32-bit floating point.

There are two recordings of Player B, because this player offers two digital filter settings. I can't hear any difference between these filter settings.

I had earlier posted a link here to the original set of FLAC files which I have used. When burning the CS, I saw free space, so I added a couple of longer tracks at the end. Therefore the set of original FLAC files does not have these last two tracks. In any case the comparison can be done now, because the CD played through the different players is consistent.

I have done zero post-processing. I have not changed the gain, and therefore the tracks are not level matched. The two recordings from Player B are expected to be at identical levels, but the Player A track is expected to be (slightly) different.

I strongly suggest that you don't come to a conclusion about subjective differences until you check them out with IEMs or good clean monitoring headphones.