Relationship between "brightness" and "resolution"

[TheFrator]

What is the relationship between these two terms? There seems to be a common sentiment that the more treble a transducer has, the more "resolving" it sounds. But this isn't always the case with my subjective, sighted listening impressions of various headphones and speakers. I've heard various dark transducers (confirmed by headphone measurements online and UMIK-1 with REW for speakers) that sound more "resolving," especially with acoustic music.

My definition for "resolving" I am using is from from this site (Headphonesty) and states: "The equipment’s ability to reproduce and separate the sounds between individual instruments. Detail is the interpretation of the reproduced resolution." Also related to this definition is "detailed." Same source defines it as: "A strong midrange and treble, as well as fast transient response, that conveys the most subtle elements in the music."

Forgive me if this belongs in the Psychoacoustics section of the forum. Thank you for reading and commenting. I'm simply trying to learn and make sense of my auditory experiences.

 

[fpitas]

As I got my speakers to be flatter and flatter in response, the resolution increased. Of course, that takes treble, as well as midrange and bass. Bass can indeed cover up details, but now you're down to asking if you want accurate reproduction, or a special effect that attenuates bass and allows hearing details better. I can't answer that, except for myself.

 

[TheFrator]

As I got my speakers to be flatter and flatter in response, the resolution increased. Of course, that takes treble, as well as midrange and bass. Bass can indeed cover up details, but now you're down to asking if you want accurate reproduction, or a special effect that attenuates bass and allows hearing details better. I can't answer that, except for myself.

I currently DSP my speakers (Philaharmonic BMR Towers with Benchmark preamp and amp) to a flat response using a software called Audiolense. This software generates a correction convolution filter that I then load into my playback software.

Sorry for the colors being not the best but I can't figure out how to adjust them. Shown are how my left and right speakers are altered when using convolution and the target curve (in the deepest shade of red/purple)

And then their simulated response with the target. (Upon writing this out I realized I need to figure out a way to measure it for real. Simulations are cool but need to be verified.)

I provide this context because my headphones (Audeze LCD-5) are more "resolving" than my speakers and any other headphone I've heard (even beyerdyanmic treble monsters) especially with acoustic music (jazz and Eva Cassidy's Live at Blue's Alley album). I guess a problem is that the headphone's measurement can't precisely be transformed into how it would sound as a speaker or vice/versa.

Edit : clarified graphs

 

[NTK]

How did you measure? Are these in-room responses or quasi-anechoic measurements?

 

[MRC01]

Brightness and resolution are often confused with each other, but they are 2 different things. A common "trick" is to voice a speaker or headphone (or a recording) brighter so it sounds more detailed or higher resolution at least on first impression. Yet the opposite can happen, meaning excessive brightness - or other anomalies in frequency response - can mask lower level sounds.

There are examples that demonstrate they really are different. Headphones or speakers having a darker signature that are high resolution (the Audeze LCD series, among others), and cases where I listen to an overly bright recording that sounds "detailed" in the first (superficial) few moments, but then I use DSP to tame the brightness and it reveals more of what's happening in the music - it reveals subtle cues in the cello or bass line that was masked by the excessive brightness.

 

[fpitas]

I currently DSP my speakers (Philaharmonic BMR Towers with Benchmark preamp and amp) to a flat response using a software called Audiolense. This software generates a correction convolution filter that I then load into my playback software.

Sorry for the colors being not the best but I can't figure out how to adjust them. Shown are how my left and right speakers are altered when using convolution and the target curve (in the deepest shade of red/purple)

View attachment 244696

And then their simulated response with the target. (Upon writing this out I realized I need to figure out a way to measure it for real. Simulations are cool but need to be verified.)
View attachment 244697

I provide this context because my headphones (Audeze LCD-5) are more "resolving" than my speakers and any other headphone I've heard (even beyerdyanmic treble monsters) especially with acoustic music (jazz and Eva Cassidy's Live at Blue's Alley album). I guess a problem is that the headphone's measurement can't precisely be transformed into how it would sound as a speaker or vice/versa.

Edit : clarified graphs

I'll point out, your speakers are in a room that has reflections. The brain is good at sorting that out, but not perfect. Headphones are basically all direct sound.

 

[TheFrator]

How did you measure? Are these in-room responses or quasi-anechoic measurements?

In room measurements.

Lets ignore my impressions with speakers since the room and reflections matter a whole lot.

I've heard bright headphones (beyerdynamic DT990, DT1990, Meze 109 Pro, and a few Hifiman) but they're not nearly as "resolving" as the dark LCD-5. I own both the DT1990 and LCD-5 at the moment so they're easy for me to switch back and forth with.

LCD-5 measurement

Beyerdynamic DT1990 measurement

Sorry but I'm unable to overlay the two onto a single graph. The main point of contention I'm focusing on is the treble (so above ~4 kHz). The 1990 is bright, while the LCD-5 is dark. However; the LCD-5 sounds more resolving, especially on acoustic music. This goes against the common sentiment (or at least the sentiment I've read ad nauseam) that "resolution" = "treble elevation". So I'm trying to reconcile my subjective point and get some ideas on where exactly resolution comes from? Is it from the headphone drivers ability to reproduce transients more effectively?

 

[TheFrator]

Brightness and resolution are often confused with each other, but they are 2 different things. A common "trick" is to voice a speaker or headphone (or a recording) brighter so it sounds more detailed or higher resolution at least on first impression. Yet the opposite can happen, meaning excessive brightness - or other anomalies in frequency response - can mask lower level sounds.

There are examples that demonstrate they really are different. Headphones or speakers having a darker signature that are high resolution (the Audeze LCD series, among others), and cases where I listen to an overly bright recording that sounds "detailed" in the first (superficial) few moments, but then I use DSP to tame the brightness and it reveals more of what's happening in the music - it reveals subtle cues in the cello or bass line that was masked by the excessive brightness.

OK so you share a similar subjective impression as me. I guess the crux of what I'm asking is, what metrics can we use to differentiate a more resolving transducer from a less resolving one? You'd think it be in a frequency response graph like I posted in my comment above this. But as we've experienced with the LCD series, it's not the treble (or at least not solely the treble).

 

[fpitas]

In room measurements.

Lets ignore my impressions with speakers since the room and reflections matter a whole lot.

I've heard bright headphones (beyerdynamic DT990, DT1990, Meze 109 Pro, and a few Hifiman) but they're not nearly as "resolving" as the dark LCD-5. I own both the DT1990 and LCD-5 at the moment so they're easy for me to switch back and forth with.

LCD-5 measurement
View attachment 244865
Beyerdynamic DT1990 measurement
View attachment 244866

Sorry but I'm unable to overlay the two onto a single graph. The main point of contention I'm focusing on is the treble (so above ~4 kHz). The 1990 is bright, while the LCD-5 is dark. However; the LCD-5 sounds more resolving, especially on acoustic music. This goes against the common sentiment (or at least the sentiment I've read ad nauseam) that "resolution" = "treble elevation". So I'm trying to reconcile my subjective point and get some ideas on where exactly resolution comes from? Is it from the headphone drivers ability to reproduce transients more effectively?

Something like that, maybe. One of the claims TAD makes about my horn drivers is the ability to better respond to tiny signals. I doubt that's just hyperbole, they do reveal a lot of detail if you listen closely. Perhaps the better headphone drivers share that ability.

 

[NTK]

In room measurements.

Lets ignore my impressions with speakers since the room and reflections matter a whole lot.

I've heard bright headphones (beyerdynamic DT990, DT1990, Meze 109 Pro, and a few Hifiman) but they're not nearly as "resolving" as the dark LCD-5. I own both the DT1990 and LCD-5 at the moment so they're easy for me to switch back and forth with.

LCD-5 measurement
View attachment 244865
Beyerdynamic DT1990 measurement
View attachment 244866

Sorry but I'm unable to overlay the two onto a single graph. The main point of contention I'm focusing on is the treble (so above ~4 kHz). The 1990 is bright, while the LCD-5 is dark. However; the LCD-5 sounds more resolving, especially on acoustic music. This goes against the common sentiment (or at least the sentiment I've read ad nauseam) that "resolution" = "treble elevation". So I'm trying to reconcile my subjective point and get some ideas on where exactly resolution comes from? Is it from the headphone drivers ability to reproduce transients more effectively?

In the ~500 Hz to 1.8 kHz range, the LCD-5 is above the "target" curve, whereas the 1990 isn't. See if you apply some EQ in the region and notice any differences. Sometimes a bump in the bass region can give an impression of a depressed treble.

 

[dorakeg]

What is the relationship between these two terms? There seems to be a common sentiment that the more treble a transducer has, the more "resolving" it sounds. But this isn't always the case with my subjective, sighted listening impressions of various headphones and speakers. I've heard various dark transducers (confirmed by headphone measurements online and UMIK-1 with REW for speakers) that sound more "resolving," especially with acoustic music.

My definition for "resolving" I am using is from from this site (Headphonesty) and states: "The equipment’s ability to reproduce and separate the sounds between individual instruments. Detail is the interpretation of the reproduced resolution." Also related to this definition is "detailed." Same source defines it as: "A strong midrange and treble, as well as fast transient response, that conveys the most subtle elements in the music."

Forgive me if this belongs in the Psychoacoustics section of the forum. Thank you for reading and commenting. I'm simply trying to learn and make sense of my auditory experiences.

While I do agree that more treble could give a perception that its more "resolving", its not always the case. I have never listened using headphones so I am not sure about it. However, for speakers, there are those which are excessively "bright" don't not much details heard or resolving. There are also those which aren't bright and yet resolving.

 

[Chrispy]

Got me, most subjective terms like that seem to mean many different things to many different people with different gear (even electronics!)

 

[fpitas]

Got me, most subjective terms like that seem to mean many different things to many different people with different gear (even electronics!)

I go with my definition. As a famous egg said, "When I use a word, it means just what I choose it to mean. Neither more or less."

 

[animalelder]

Sorry but I'm unable to overlay the two onto a single graph. The main point of contention I'm focusing on is the treble (so above ~4 kHz). The 1990 is bright, while the LCD-5 is dark.

The graphing tools are fun. Here are the two, plus an image where LCD-5 is zeroed as well as the Beyer as the 0.

Apologies if I misunderstood your post.

 

[TheFrator]

Got me, most subjective terms like that seem to mean many different things to many different people with different gear (even electronics!)

Maybe it's a bit futile as of now to determine a metric from what makes a transducer "resolving." I don't want to mention the word "speed" when it comes to a driver because I don't think that's it. But I hypothesize it has something to do with the transients and how accurately they're portrayed. This statement assumes the music recording a high quality one.

Apologies if I misunderstood your post

You didn't misunderstand at all! Thank you very much for taking the time to graph these for me. A 10dB difference at 8 kHz and a 15 dB at 10 kHz ! Of course these measurement rigs have some resonant frequencies with the upper treble so the measurement is not as trustworthy as say the mid range measurement. But hey it's a comparison.

I'd just really like a cut and dry answer but there doesn't seem to be one when it comes to measuring a subjective trait like resolution. In addition to this, we also have our own unique HRTF at play.

 

[Da cynics]

In the ~500 Hz to 1.8 kHz range, the LCD-5 is above the "target" curve, whereas the 1990 isn't. See if you apply some EQ in the region and notice any differences. Sometimes a bump in the bass region can give an impression of a depressed treble.

Complete answer