How do you SUBJECTIVELY quantify improvements in sound

[gks333]

2nd post. I was curious to see how you all quantify increased improvements in sound based on percentage.

Example 1: About 2 years ago I upgraded my front end from a Marantz SR 7005 receiver to separates; Anthem AVM60 and MCA 525. I told myself “if I get a 5-7% increase in sound I would be happy”. Well, that’s exactly what I got, about a 7% increase. The speakers at the time stayed the same, Focal chorus 826v/800CC. The dialogue was a bit cleaner, more detailed. Sound throughout was a hair less strained/more effortless but all in all, a minor but noticeable improvement.

Example 2: This year I replaced my Focal chorus 826V/800CC/700SR with Revel Performa3 F208/C208/S206. I told myself the same thing as before, 5-7% improvement hoping much of the audible improvement would lie with the C208 since my system is 97.31% movies but knowing that there is so much music within movies. Well, I got about 5-8%, initially to be fair, before I started using REW to integrate subs and mains. Yes, dialogue was a bit clearer. Left and rights, very minimal but less treble across all (to be expected since focal is on the more lively side). Biggest improvement was from the S206 surrounds which really opened up the rear sound stage. Probably due to their bipole configuration vs focals single/direct.

Trying to be as unbiased and unemotional about what I am hearing as possible, my thoughts are generally that a $700 5 speaker plus 1 sub system plays close to the hole frequency spectrum say 45-20khz. It will not play as loudly, cleanly, accurately, have as good of sound stage etc but it is playing, for the most part the majority of the source. I watch youtubers like Zero fidelity, Z reviews, Andrew robinson, steve guttenburg, Youthman, Home Theater Guru’s, Jay iyagi, Joe n Tell and Thomas and stereo. When it comes to home theater (HT), it seems like the hometheater youtubers and I are not as picky or particular to hear “lips smack” or the “saliva in a singer’s mouth” during a song. The “sizzle” or “open airiness” of whatever instrument. I understand and believe myself that when listening to certain music you can really hone in on strengths and weaknesses of your system. Certain sounds on recording are more easily revealed and are more apparent in certain musical recordings. Also, are more intimate and reveal more of a speakers/systems ability and/or personality while a movie is more like “talk talk…sing music…boom bang” etc. When there is a great score, say Lord of The Rings, I seem to not be as critical at all of the sound vs a stereo recording and just go a long for the ride as the director intended. Yet that score within a movie can bring so much emotion out of me that music rarely does. Apples and oranges no doubt since movies are not just audio but visual and storytelling as well.

So when I bring up percentage of improvement, my current system is 100%. Then upgrade is integrated, reviewed/critical listening and then conclude 5-7% increase. Now that system becomes current system at 100%.

The other thing I do is compare my system to the current reference system I have heard. That system (yes I know, memory is fleeting and with time distorted) is 100% and my system, lets say, is 88% of that system based on accuracy, dynamics, sound stage etc.

It reminds me of going to a beerfest and having very good beers and people would just say “damn that’s good”. Verse going to wine tasting and all the words they use to describe the wine while swishing, sniffing and gargling the wine while I just drink the damn thing. Seems that stereophiles are more sophisticated vs hometheaterphiles just want to have fun and not dive as deep into the nuances of the performance.

This may come off as “stereo vs HT” but that is not my intent. Just observation from over the years.

Whats your thoughts? Again, I am 97.31% movies with my system.

 

[DVDdoug]

I don't think in those terms... I've been "into audio" for a long time so I tend to think in decibels. Decibels can be converted to percentage but I wouldn't convert dB to percent "quality".

Generally with modern electronics noise, distortion, and frequency response are all better than human hearing and at that point there can be no improvement in actual "sound quality". Sometimes you might hear noise but that's a very identifiable defect. You can measure and quantify the noise but I wouldn't say 10% less noise is 10% better...

Of course, amplifier power is also a consideration if you can't get enough loudness without (audible) distortion.

Speakers (and headphones) are a different story. There are always frequency response variations and every speaker sounds "different". The room also makes a difference... In a room and you get more frequency response variations as well as reverb and other timing-related variations.

 

[bravomail]

Whats your thoughts? Again, I am 97.31% movies with my system.

I'm in the same boat- movies r majority of my use case.
But I'm willing to experiment.
So after listening to AIYIMA A07 - my Yamaha AVR was retired.
I suggest u do the same - experiment. Esp since u r "financially talented"

 

[amper42]

Chinese rover found house on back side of the moon. They are exploring the area for months. Once inside they found Topping Pre90, D90SE DAC and Topping PA5 amp powering a pair of BMR Monitors. The hardware was bolted to the ground as gravity on the Moon's surface is only about 1/6th as powerful as the earth. No sub used for fear of lift off with each 30Hz tone. The lack of gravity amazingly improves the quality of hifi sound for those few seconds when you have your space suit helmet off.

Chinese rover spots strange ‘mystery house’ on the dark side of the Moon

The ‘house’ is likely to be a large boulder, but astronomers will not know the truth until the month-long trip along the satellite

www.independent.co.uk

 

[gks333]

I don't think in those terms... I've been "into audio" for a long time so I tend to think in decibels. Decibels can be converted to percentage but I wouldn't convert dB to percent "quality".

Generally with modern electronics noise, distortion, and frequency response are all better than human hearing and at that point there can be no improvement in actual "sound quality". Sometimes you might hear noise but that's a very identifiable defect. You can measure and quantify the noise but I wouldn't say 10% less noise is 10% better...

Of course, amplifier power is also a consideration if you can't get enough loudness without (audible) distortion.

Speakers (and headphones) are a different story. There are always frequency response variations and every speaker sounds "different". The room also makes a difference... In a room and you get more frequency response variations as well as reverb and other timing-related variations.

So, like I said, I'm curious, how do you come to a conclusion that, example, this system that you just heard compared to yours is X amount better? Or, the new front end I got made my system X amount better or worse? How do you convey to someone else the quantitative difference so they understand how much better (or worse) it is/was? I grade most everyday things on a 10 scale like food, drink. 10 being the best burger I have ever had. A whopper would be a 5 or 6. Tastes like a burger. Kinda looks like a burger. Feels like a burger but does not make me say "DAMNNN....THATS A BURGER!!!!

 

[gks333]

Chinese rover found house on back side of the moon. They are exploring the area for months. Once inside they found Topping Pre90, D90SE DAC and Topping PA5 amp powering a pair of BMR Monitors. The hardware was bolted to the ground as gravity on the Moon's surface is only about 1/6th as powerful as the earth. No sub used for fear of lift off with each 30Hz tone. The lack of gravity amazingly improves the quality of hifi sound for those few seconds when you have your space suit helmet off.

Uh....what???

 

[Willem]

I was curious to see how you all quantify increased improvements in sound based on percentage.

What is the unit on the measurement scale? Anyway, with electronics much of it is better than human hearing, so no further improvement is audible.

 

[hex168]

My understanding is it can't be done. Percentage of what, exactly? For example, Sean Olive's predictive preference score looks at variances from flat frequency response and derives a single number score. That score correlates well with listener preference. However, all that means is that if speaker A has a higher score than speaker B, people are likely to prefer Speaker A.

A 10% higher score compared to a 20% higher score means that the chances of a listener preferring speaker A are greater for a 20% higher score than for a 10% higher score. But there is no meaning to 10% or 20% "better." It just gives better odds of getting the speakers in the right preference order.

There's a term for an ordered list with no scale, but I can't think of it. But that is all a preference score is good for.

Olive Score:

https://patentimages.storage.googleapis.com/95/03/0a/a4a8dbd7d8042c/US20050195982A1.pdf

 

[Willem]

It is an ordinal scale.

 

[hex168]

It is an ordinal scale.

You know that tip-of-your-tongue feeling when you can't quite dredge up a word or phrase? It's happening to me more often now. You just fixed at least this instance of it. Thanks!

 

[BDWoody]

Or, the new front end I got made my system X amount better or worse?

I think I'm starting from a different place...

I don't expect ANYTHING to make a meaningful audible difference unless it has moving parts...like a speaker, TT cartridge, or headphones.

The rest, competently chosen, should be indistinguishable. My Krell and Bryston amps sound no better than my Denon receiver if I don't push the receiver outside of its happy operating envelope. Same with DAC's, cables, etc.

It may be easier for me, because I've built my systems with an eye on competence for the job at hand, accuracy and the highest fidelity from source to transducer, rather than using gear to color the signal, as many seem to enjoy. I can always add distortion, but I can't take it away.

I wouldn't have any idea how to quantify comparisons between different systems at different times...

 

[gks333]

I think I am being misunderstood or I did not convey my point properly. I am meaning in a subjective way to someone else or yourself. Do you just say after an upgrade to your own system that, "its better" or do you try and put a number on it NOT being mathematically or scientifically proven correct but to perceptualize the improvement. To quantify in terms that make sense to you or who your are trying to convey your observations/experience to in more detail than "sounds better".

 

[weesch]

hi
with the resolution of the treble and high medium....
because it's the more difficult to reproduce...
best regards
weesch

 

[Headchef]

I do often wonder how much it would be to have a system based on the top performing kit inspected here.

I’d assume that it would be built around those rather tasty Genelec active speakers and then the Topping D90SE?

 

[Peterinvan]

Why bother trying to describe sound quality improvements by a percentage? We have all kinds of subjective aspects where we just trust our ears, such as Andrew Park’s standards:

• Detail Retrieval
• Speed & Dynamics
• Soundstage & Imaging
• Timbre
• Frequency Response & Tonality
• Response to EQ

 

[Blumlein 88]

https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1284-2-201901-I!!PDF-E.pdf

 

[Jimbob54]

I think I am being misunderstood or I did not convey my point properly. I am meaning in a subjective way to someone else or yourself. Do you just say after an upgrade to your own system that, "its better" or do you try and put a number on it NOT being mathematically or scientifically proven correct but to perceptualize the improvement. To quantify in terms that make sense to you or who your are trying to convey your observations/experience to in more detail than "sounds better".

I think you like numbers more than I do. I just think in terms of better or worse and then if trying to explain to someone else why, will resort to the perceived qualities I think one brings over the other. Then the question becomes "is the improvement worth the price?".

If I started trying to say that a change had led to a 7% improvement in the bass without having measured the before and after response, people would rightly ask me how I had arrived at that number. Likewise , if I said burger A was 7% better than burger B.

 

[escksu]

Ask yourself, how do you quantify any subjective experience. Like a car, you felt car A is a better and more comfortable ride over B. Even food, taste better. How much better?? In what way its better?

 

[phoenixdogfan]

Generally I use percentages. I go by how much it improves in relation to the Platonic ideal. So if a speaker is 50 percent of the Platonic ideal, and, say a new pair of speaker cables improves such that it moves it 20 percent closer to that ideal, so the combined speaker+cables is now 70 percent of the Platonic ideal, I just report a 20 percent improvement.

 

[JRS]

It reminds me of going to a beerfest and having very good beers and people would just say “damn that’s good”. Verse going to wine tasting and all the words they use to describe the wine while swishing, sniffing and gargling the wine while I just drink the damn thing. Seems that stereophiles are more sophisticated vs hometheaterphiles just want to have fun and not dive as deep into the nuances of the performance.

If you ever sit in on a homebrew judging (many of the clubs will often do a tasting after the event where various bottles--usually not the really good stuff which seems to suffer from evaporation--are opened and ruthlessly, though politely, dissected and you follow along learning the vocabulary of mostly defects, but learning a helluva lot about beer enjoyment along the way), you'll be surprised by the numbers of poets who can wax eloquently for some time. During a typical competition, a few hundred beers will be sampled (beers have a 50 point scale) and scored. Maybe 5% will score over 40, and none over 44. The majority slump in the middle in the 27 to 35 range, with some real nasty shit scoring in the teens. Carbonated septic tank might get you some points for color and head retention.

I bring all of this up because it strikes me that we have this metric adopted from Harman Labs which focuses on two metrics, on axis FR, directivity discontinuities and distortion. These numbers carry a great deal of weight with members, almost reaching holy grail status, but they don't include the authority of a loud speaker, imaging, compression and a sense of ambience. So it's clearly a huge step forward in my mind, but a fair amount of wiggle room for exceptional loudspeakers that may miss a metric. One of the great features is having a numerical scale that obviously won't correlate linearly with quality, but at least a bench mark.

It seems there should be a way to go beyond the more or less objective criteria of the Harman curve and include some of these less tangible measures. Indeed an entire system should be amenable to the same kid of objective analysis. Of course subjects are involved in making these judgements but as Toole and Olive discovered, results are more consistent than what you might believe.

Getting back to the beer which is very "subjective," the interrater reliability in scoring is striking. Rarely will two judges differ by more than 2 points and only once have I seen a disparity of 5. When this does occur, the judges go thru an item by item analysis and generally reach a compromise within 2 minutes. We should be able to do this with audio.