The problem with subjective impressions

[Kimo]

Subjective impressions may be problematic in that they generally will reflect the listener's subjective impression of what a good system should sound like,  which may not align with your own. 

 

For example, I have a strong dislike for almost all metal dome tweeters to the point that I find most systems difficult to listen to when I first sit down with them.  They typically sound harsh and flat to me.  I am aware that the Raal ribbons I prefer probably don't measure as well as some dome tweeters, but they also don't annoy me.  For those who don't share my affliction, the metal dome may be the superior choice, and my subjective view isn't worth squat to them.

 

If you do happen to find someone who shares your impressions in general, I believe that their subjective impressions can be quite valuable.  

[Kimo]

1 minute ago, Summit said:

 

Yes it’s true that subjective impressions generally will reflect the listener's subjective impression and preference, it is why they are subjective. I am not talking about preference per se, but that most reviews and impressions are written the same and with the same words. I mean you can almost take any review and how the gear they liked the most is described and just change the name of the gear for another gear and voila. The problem as I see it is this type of review will not reflect the SQ of the gear in the grand scheme of things.

 

Even if a good USB cable for example is important for obtaining really good SQ I would not say that difference from one good USB cable to better USB cable to be as big as the difference between a good amp and a better amp. I think that it would be great if we would get some sort of classifying of how big the difference is in relative terms (1-10), so that our subjective listening impressions would be more objective. Not objective as in measurement or DBT, but more like how big was the overall SQ effect by changing gear A to gear B in their reference audio system. Most impressions focus too much on the reviewed gear IMO and in the effort to describe the gear in depth we may not know how big the upgrade/change really is as part of a system and together with the other equipment in the audio chain.  

 

I don't think that we are going to be able to set an objective scale for subjective impressions for much of anything in life.  How much hotter is vintage Jennifer Garner than vintage Pam Anderson?  We are still stuck with subjectively even when talking about different kind of 10s.   

 

[Kimo]

4 minutes ago, Archimago said:

 

Curious Kimo, have you ever tried blind-testing a dome tweeter speaker with similar type of speaker (be it 2-way monitor or maybe floor stander) using ribbons at the same volume level to confirm that impression?

 

 

Nope.  I got tired of trying to "fix" the treble on my former Vandersteen, Dynaudio, Triangle, etc. housemates and banished them.  I am okay with most Tannoys though, for whatever reason.  I used single drivers for many years, my dislike of most tweeters being so strong.

 

I now use objective response.  My narrowing eyes.  The sneer of disapproval of my wife provides my confirmation bias.  

[Kimo]

4 hours ago, STC said:

But we have audiophiles still believe a 60s amplifiers or speakers supposedly to be the holy grail of the ultimate SQ. 

 

We also have respected and successful tube amp manufacturers that would say the same thing, though not about Scott and Fisher.

 

I also know of speaker manufacturers that will tell you the greatest sounding driver ever built was the RCA.

[Kimo]

1 minute ago, Summit said:

 

No the scale is subjective, but reflects the importance of the audio gear in the audio chain and not only as one individual equipment.  

 

I would agree with you on the importance of amps and speakers, but others would not.  I know people that will tell you that a good preamp is the foundation of great sound in any system, and others that would say no preamp is the best preamp.

[Kimo]

2 minutes ago, Summit said:

 

To be objective is to reduce subjective factors to a minimum, but we can never eliminate them, and most people will tell it like they hear it. Preamp or not. 

 

You would think so, but that is not what I have experienced.  One time about dozen audiophiles all older than myself were mulling around in a shop and the owner, who was messing around with some test equipment asked us to raise our hands as he played test tones rising in response with each subsequent play.  As he ran through ascending tones, one fellow who admitted his hearing was damaged dropped out rather quickly.  I was next at 16k.  

 

The rest of the golden ears kept popping up their hands as he advanced well in the mid 20K region.  After the test I told him that I was surprised that my hearing dropped off so quickly compared with the group.  He told me that the last tone he actually played was the 16k tone, and that he was just saying he played the higher tones for a laugh.

 

I guess that story doesn't do much for the subjective cause.

[Kimo]

4 minutes ago, fas42 said:

 

Here's a good example of allowing the "shape of the camera body" to loom like a large monster, getting in the way of being able to take "good shots" ... a tweeter is merely part of the system delivering the recording; if you can "hear the tweeter" then you are listening to the rig, and not to the musical event - the equipment used is dominating your perspective, such that the capture of the musical event is well down on the totem pole ...

 

Nope.  If I disconnect the offending tweeter, the offense disappears.  If I insert a single driver, the offense remains at bay.  Some parts of the system are more critical than others.

 

First rule, do no harm, neither subjectively nor objectively.  

[Kimo]

28 minutes ago, plissken said:

 

Measurements would suffice. I blind tested reconstruction filters of a DAC over headphones and could 100% pick out when it was A or B. The cool thing is seeing what it looked like measurement wise.

 

 

 

I have tried switching through the filters on my DAC and found that I can't reliably pick them out on my monitor speakers.  

 

I was able to pick out a misbehaving ESS 9038 implementation, however.  

[Kimo]

20 hours ago, Ralf11 said:

misbehaving ESS 9038 implementation as in "hump"?

No hump, just high order distortion.  Fixed by the company, once they received updated information from ESS. 

 

Apparently, more care needs to be taken when using the 9038, as opposed to other ESS chips.  

[Kimo]

Speaking of problems with subjective comparisons, or maybe even on objective ones for that matter.  Has anyone ever had the chance to compare multiple samples of the same unit?  I wonder what type of sound variation you might find, though I would hope it would be pretty subtle if everything was working properly.

[Kimo]

7 minutes ago, fas42 said:

 

Well done two channel does everything that's necessary to create a powerful, intense, immersive listening experience. Provided the system is capable of distortion free SPLs from the two speakers that match live music, then adding extra channels adds nothing to what one experiences. Of course, there are very few rigs that can run cleanly to the higher volumes without introducing excess, disturbing anomalies, so extra channels are an easy workaround, "sharing the load" - each channel has to work less hard, so distortions are significantly lower ... job done!  😉

 

The downside is, that the spatial information now presented to the user is much more complex. so much of the ambience encoded in the recording may no longer make sense - swings and roundabouts, 🙂.

 

If these are your ultimate objectives, you will be slightly limited in your quest.  Wide baffle pro style speakers with active crossovers and amplification.  Best bets would be ATC, Grimm, PMC, Westlake, etc.  

 

You will also need to listen in the near field in general, and have plenty of room treatment and maybe some DSP.  You ain't getting there with tubes, horns, and wide dispersion speakers.  Nothing wrong with the active route, but I would say most of those pro audio focused manufacturers aren't aiming for a subjective approach.

[Kimo]

22 minutes ago, fas42 said:

 

Those methods are shortcuts using currently available  equipment - my goal is to hear the recording, and only the recording. at any volume level I choose - the gear is purely a means to that end. It turns out that nearly all hifi systems are too severely handicapped when set up without thought; so lots of DIY and/or tweaking is needed to get them to the right status.

 

If the accuracy of the sound coming from the drivers is good enough, then the ear/brain does all the sorting out to balance the sound in the room - it always "sounds right", whether listening from within a few inches from the speakers, or from another room, or heard through the windows from outside.

 

They are not shortcuts.  They are state of the art equipment designed for maximum accuracy to be used by mastering professionals in rooms designed for minimum interaction.  If you really want to hear the differences in how recordings sound, this is the way to do it with speakers, and yes, you need volume to get all the way there.  I can't imagine too many engineers master at a low volume, even when using a horn based model like an M2.

 

All your set up magic and tweaking isn't going to overcome a high noise floor, or a a phasey crossover, though it might improve on what is already there.   The mastering studio is pinnacle of set up, no?

[Kimo]

They are still not shortcuts in that you won't get there without them, if you are looking to hear all that is on the recording, and even if they don't all sound alike.  You need to hear what a pro mastering solution sounds like to hear what one designer's interpretation of neutral playback of your recording really sounds like,or maybe at least a Benchmark system on a pair of really good monitors.  Hey, it may only come down to studio effects and reverb, and even stuff the engineer couldn't hear when he finished the recording.  But if your goal is to really hear what is on the recording, this is the route you pretty much need to take.

 

I can give you very specific examples of what is lost with 16 bits with Doors and CCR recordings, per the guy who mastered them.  If your system (not likely), room (a little more likely), or ears (more likely) aren't up to it, you won't notice the difference and it won't matter how loud you play the system  It needs to be loud and revealing.  As I understand it, most of the engineers basically damage the hearing over time, due to this practice.

 

That being said, maybe you can get the best for you, without going to extremes, if you are older and suffering from the inevitable decline in hearing that goes with advancing years, a system with an SNR of 120 may mean little to you.  I would like to think that it is not inevitable, but I understand the exceptions will be few.

 

Of course, none of this means that you will have the most subjectively pleasurable system going this route.

[Kimo]

11 hours ago, fas42 said:

 

What "specific examples" are lost, with Doors and CCR recordings?

 

 

S/N is actually highly irrelevant  - a rig may blast you with theoretically low noise sound, but if it's contaminated with disturbing distortion anomalies, then your hearing will give it a big thumbs down.

 

The best I've heard from a living room type system used Bryston monoblocks, and close to the best Dynaudio speakers - this had clarity, and all the nuances that matter - I've heard plenty of lower grade pro monitors, and they're basically pure shite; so many things wrong with the SQ they're laughable. The Bryston combo showed how it should be done - could do standing a couple of feet away from a drum kit being given a full workout, with ease - you know, your head is almost exploding from the intensity of the sound type of thing, 😊.

 

I don't do this loud, but the quality is at that level.

1. CCR: The drums at the beginning of Fortunate Son on the AP SACD.  On the the high resolution area, the reverb and decay is obvious.  On the CD layer it is truncated, changing the presentation significantly.  Hoffman has offered this as an example of how 16 bit alters the sound of the original recording, due to the loss of low level information.  It is a striking example.

 

2. The Doors: The beginning of Moonlight Drive is marked by some rolling light piano in the right channel.  On the DCC CD, the 16 bit standard, the decay is once again altered, as compared with all other high resolution versions, both download and disc.  This results in a thicker, less delicate presentation, which is not so true to the recording.  Another example of how the sound is changed by the loss of low level information.

 

3. You are correct about the pro systems offering up a bit too much high order distortion for many people's liking.  I am guessing that the emphasis on low overall distortion exaggerates this problem, but playing the AB and D amps closer to their limits helps with results. 

 

4. Lower grade pro stuff really has gotten pretty low grade.  This stuff I was referencing is actually rather pricey.  The other problem is that you have to buy amps, crossovers, and speakers all at the same time with the active stuff, which makes it feel even pricier.  

[Kimo]

3 hours ago, fas42 said:

 

Just had a listen to a few versions of this on YouTube - some have lots of reverb on the drums, others don't. So, there are obviously different masterings of this track out there ... since the quality of data compression of YouTube clips is not going to do the SQ any favours, as heard over tiny laptop speakers, 😉 I can't see how 16 bit vs.high res comes into the equation ,,, 🙂.

 

https://forums.stevehoffman.tv/threads/steve-i-conducted-your-creedence-ccr-fortunate-son-cd-vs-sacd-test.170124/

 

[Kimo]

3 hours ago, John Dyson said:

I called on some friends elsewhere in the audio world, and found approx 3 (not sure -- haven't compared them all) versions of Moonlight drive) -- I was curious, never listened ot the Doors...  Which one are you writing about?  Everything available on original digital distribution  apppears to be DolbyA encoded -- I might be able to decode the section that you are talking about...  I don't have any sources for analog rips- so have no references.

 

John

 

 

There are two versions of Strange Days on HDtracks.  The original high resolution master, and a deluxe mono/stereo version featuring a new high resolution master.  As well, there is a DSD transfer available from Acoustic Sounds in download and SACD.  The CD I used was the DCC Gold Disc, the 16 bit digital standard according to most.

[Kimo]

29 minutes ago, fas42 said:

 

Which deals quite a bit with the quality, and setup, of the ADC used ... a post,

 

 

This, to me, is another "everything matters!" - extrapolating from a situation where a whole variety of factors may have impacted, to then say this "proves" that something is true is going waaaay out on a limb...

 

It is not going way out on the limb.  Quite literally everyone who bothers with this stuff knows that more bits is better for AD mastering, or even necessary for proper AD mastering, and that a final downsampled copy can often come quite close to a penultimate 24 bit copy.  That latter has nothing to do with what is lost during transfer from the analog source.  

 

Are you suggesting that the SACD process adds ambiance?

 

 

[Kimo]

48 minutes ago, fas42 said:

 

That's nifty! I have the CD, a German 1985 release - the YouTube versions were too downgrade - I tried playing from the CD-ROM drive, still not good enough, ripped to the hard drive - niice!!

 

Which piano, again? There are 2 acoustics, with separate pianos, on the left and the right ... nice cymbal work in the middle of the track, good vocals - I wouldn't call it a "thick" presentation ....

 

The opening piano riffs in the right channel, you know da dum, dum, dum...

 

Do you have a high rez sourece to compare with?  I wonder what the original vinyl sounds like.  I have never seen a clean copy of Strange Days on original vinyl.

[Kimo]

1 minute ago, Rexp said:

Some well described subjective difference there that current measuring devices wont pick up. OT but I wouldn't necessarily blame 16 bit. 

 

Well, it was the guy who mastered it who called out the loss at 16 bits. 

 

The alternative is that someone added reverb to the SACD, which would make Mr. Hoffman, or some audio sneak, a bit of a mischief maker at best.

[Kimo]

2 minutes ago, fas42 said:

 

Getting confusing ... it's a call and response pattern - the left piano calls, "da dum", the right one responds "dum, dum" - subjectively it's two tracks, in the mix ...

Sorry for confusion, but you got it, the response.  

 

I have to believe this is pretty easy to hear on most systems.  I could pop in the tube amp tomorrow for comparison.

[Kimo]

Just now, Rexp said:

Are there any versions on Tidal you like? 

Sorry, I don't do Tidal.  

 

I actually like the DCC CD, but the decay is lost a bit.  I believe he mastered this with a tube based system, so maybe it was lost in a 12ax7 somewhere?

 

I guess I could track down a non DCC CD to see if it is closer to the DCC or the high resolution.

[Kimo]

7 hours ago, fas42 said:

What's with people swapping left and right channels, on uploaded YouTube?? 😨 ... Found a few dodgy versions, but this one, gets pretty close to what I'm hearing,

 

 

Is this good enough to say that the right piano isn't up to it? Or is the decay OK?

 

Actually, the DCC might even be worse than this one, but I am comparing this on my computer with my system, which isn't that easy.  That being said, either high res version on HD tracks sounds different.  

 

The overall effect is a hardening of the notes being struck.  No warble.  

[Kimo]

49 minutes ago, John Dyson said:

Here is the 'best match' of my 3 copies for the You Tube example.  It was a bit of a hasty decode, and only had the single cut from the album, so could onlly make the compenstaing EQ based on the single cut off the album.

From the apparently early CD copy that I have, in feral DolbyA format, I had to do the following EQ before decoding:

 

bass -3 2250 0.50q bass +3 250 0.50q treble +3 2250 0.50q

bass -6 2750 0.50q bass +6 750 0.50q treble +6 2750 0.50q

treble -9 6k 0.50q treble 0 12k 0.50q

(The fact that the gain at 12k was '0' or '-3' is a good sign that the EQ is correct.)

The calibration level was -12.45dB (a totally expected value),

decoded in MS mode.   LR mode resulted in things shifting around.

 

Sorry that this is a limited length snippet.  The fact that it is the 'A' version means that it is the second full attempt after finding the correct MF eq (MF eq is the 2250 and 2750 EQ for the needed dip in the midrange.)

 

12 - The Doors - Moonlight Drive-A-snippet.mp3 2.1 MB · 0 downloads

This pretty much sounds completely different from anything else with the drum rolls being noticeably snappier.  

[Kimo]

3 minutes ago, Ralf11 said:

Repeatedly insisting that power or interconnect cables cannot improve sound is needed to counter such insanity.

 

 

 

Would the effect or no effect of power cords have anything to do with unregulated power supplies in tube amps?  Seems like you could measure this one pretty easy.  

 

 

[Kimo]

2 minutes ago, esldude said:

You do know the overwhelming majority of solid state amps have unregulated supplies don't you?

But still much less voltage than a tube amp supply, so less demanding, no?