What is more important to you: frequency response or imaging/soundstage?

[KeithPhantom]

For me, I prefer headphones/speakers that have the correct frequency response even if I have to forfeit some of the imaging/spatial qualities of sound. What do you prefer?

 

[Daverz]

Of course I don't really think that imaging/soundstage is more important than frequency response. But I find the idea of not caring about imaging to be equally ridiculous.

 

[KeithPhantom]

But I find the idea of not caring about imaging to be equally ridiculous.

It isn't not caring, but since there is a trade-off between those two, I rather getting better performance in the frequency response department than getting the best imaging/soundstage performance.

 

[Daverz]

It isn't not caring, but since there is a trade-off between those two, I rather getting better performance in the frequency response department than getting the best imaging/soundstage performance.

Yeah, if we have to prioritize them, although as we've seen from many of Amir's reviews, if the speaker has good directivity and you can EQ it, frequency response doesn't have to be perfect. Though I'd prefer to avoid the requirement for EQ.

I could not be satisfied by speakers that imaged poorly, so they would be a waste of money, however good they were otherwise. I think good imaging is some combination of good frequency response, good directivity, and careful driver matching, so creating this dichotomy is wrongheaded.

 

[levimax]

since there is a trade-off between those two, I rather getting better performance in the frequency response department than getting the best imaging/soundstage performance.

Trade off? Not sure what you mean... I dont think more accurate FR negatively affects imaging.

 

[pozz]

The second depends on the first. I think you tonal balance, not frequency response.

 

[KeithPhantom]

Can you explain to me more about this relationship? It would be helpful to understand more about this topic scientifically.

 

[bigjacko]

I think image and sound stage is way more important than frequency response because we have eq, unless the driver is just plain too bad that eq cannot bring the frequency response back.

Can you explain to me more about this relationship? It would be helpful to understand more about this topic scientifically.

At different angle, the sound will be transformed into different frequency response due to our ear and body, that is how we know the direction of sound. At further distance, the sound will become lower in SPL, that is how we know the distance of sound. So if we can manipulate the frequency response, we can manipulate image and sound stage. SPL change will be fixed once the music is mastered, but people still experience difference in sound stage across different headphone or speaker, so there are still more things that "makes us feel" there is more sound stage, but in reality it is not in the recording and it is not sound stage but just the feeling of it.

 

[MDAguy]

tonal balance is the combination of both ... and you need both.

 

[MakeMineVinyl]

I honestly can't see how that would have to be a choice. The experience is lessened if either is deficient unless we're talking mono.

 

[witchdoctor]

If you are starting a speaker company and doing a poll for your marketing if you can't do both reasonably well the competition will be too much, better not start yet.

 

[Inner Space]

They're not mutually exclusive. What matters for imaging is accurate pair matching and absence of cabinet-generated sound. A bad FR will give a weird image, with things in the wrong places, too far away or right in your face, but it's still an image, as sharp as can be.

But without the either/or - in terms of pure preference, if for some reason I couldn't have both - I would value imaging over flatness.

 

[Chromatischism]

In addition to what Inner Space pointed out with pair matching, I find imaging to be a product of off-axis response that closely matches the on-axis, dispersion width consistency throughout the frequency range, along with placement in your room so that reflections are equal.

You can make a bad-sounding speaker from a tonal perspective still image well, so I wouldn't say this is a choice. I would say that placement in your room often does cause a tradeoff between soundstage, imaging, and avoiding SBIR.

 

[pozz]

Can you explain to me more about this relationship? It would be helpful to understand more about this topic scientifically.

When you say frequency response, you're talking about an engineering feature, a characteristic of the output. Imaging is a perceptual experience based on timing and frequency-dependent amplitude of an acoustic event. So you're mixing domains, somewhat.

Take speakers. Panel types have characteristic imaging effects because of their radiation pattern, especially if you take the time to align the back wave and get it to bounce to the listening position. The combo of direct sound from the front wave and indirect but strong early reflections from the back wave leads causes a stronger sense of depth. Omni speakers work similarly to affect imaging by generating strong early reflections, but with greater diffuseness given that side radiation isn't cancelled like it is in panels. The differences due to "radiation pattern" are really another way of describing complex frequency response, which is separate from "tonal balance", i.e., your final impression of how good the bass/midrange/treble delivery is, when all is said and done. (The Harman research into this question concluded that a flat listening window and smoothly declining sound power lead to the best tonal balance.)

Same goes for headphones. Only that frequency response as measured by anthropomorphic coupling systems (like the GRAS rig, which simulates just the ear) is a combined, complex result because the acoustic space is so small. I don't know of reseach that has tried to disambiguate the various reflections from the direct sound inside the cup. So some of what's going on is not really understood.

Something to understand about headphones is that they isolate your face, head and shoulders from sound. IEMs further isolate concha and pinna. This means that you're removing some of the complex frequency response effects caused by very small refllections and absorption of those parts of your body. Some of the cues you are used to expecting in every day listening are simply not there. This leads to "in the head" localization. By using a more humanlike setup, like the B&K HATS, and taking measurements (without headphones) of sound sources coming from in front of the dummy, or from the side, or at various distances and heights, you can then compare what final frequency response differences there are between headphone/IEM results and those spatial sources, and try to get the former to match the latter. This is what Sennheiser did with their diffuse sound setup and what led to the design of the HD8xx series of headphones. They attempted to push localization outwards, to externalize it, and succeeded to some extent. However, a lot of people disliked the final tonal balance of the initial HD800, despite the spatial effects.

So you could rephrase and ask if tonal balance is more preferred than imaging. Harman would say the preference research shows direct radiators (designed to their specs) and tonal balance are more important at the end of the day than whatever spatial effects are caused by other designs. Some people obviously opt for spatial effects and different types of radiation anyway. Unfortunately there is no way to arbitrarily control imaging but keep tonality the same, so there are always trade offs.

For headphones, I like the sound of the Sennheiser HD820 despite the weird tonality. It leaves me wanting more, so I will switch to my AKG K371 Harman-designed headphones sometimes, but I dearly miss the spatial effects. For speakers, I'm in the tonality/Harman camp (I can get into my reasons why if you like, which are do with how I understand "fidelity").

 

[Snarfie]

Imaging/soundstage but i got their since i used Mathaudio Room EQ an speakers that are build specific around a phase coherent an a time-alignment design. However Mathaudio had the most influence about the final result regarding imaging/soundstaging because i guess it created a level playingfield for all freqencies which eliminated most of my horrible roommodes.

 

[Kal Rubinson]

I honestly can't see how that would have to be a choice. The experience is lessened if either is deficient unless we're talking mono.

.........or multichannel.

 

[tvrgeek]

Both provide information to your brain to fool you into thinking what you hear is real. There is no "right" answer. Every condition depends on your preference and the degree of each.

Your brain make assumptions on equalization when you walk onto a room. As long as the FR is not too far off, it is happy. It also makes a decision based on what you see and ambience from sound and makes decisions on soundstage. It is a mater of degree on both. If you have "reasonable" eq, and do a decent job on near space reflections and refraction, you get a happy overall image. You can make easily audible changes to eq or soundstage, but which is right you may not know.