[stang]

obviously not. I know it is not a lot better. I was clearly stating judging by the graph, if we were using graphs ONLY and not our ears, it should be a lot better.

 

[xnor]

Mids look recessed, bass looks bloated. What do you see?

 

[jynweythek]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif I'm not buying because of a certain graph, but graphs can help you make decisions.

I bought the V6 based on most reports saying that it was very flat, and the HeadRoom graph backing that up, which combined seemed like pretty convincing evidence to me. However, it is most definitely not flat (as I explained earlier). The HeadRoom graph is simply wrong for this headphone. But I guess in this case both the graph and (most) people's reports were wrong.

 

[Shike]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif I understand that it's not totally useless (also didn't say that), but who cares about mechanical damping in headphones?

When you're talking a dynamic transducer you better care about motor structure. Think of it this way . . . remember foam surrounds on speakers and how it's changed to rubber? It allows for greater xmax and faster recovery of the diaphragm. With loose mechanical damping nothing really controls the driver to keep it from over excursion and other nasties. This is why sometimes the bass will climb extremely fast in the 50hz, or worse, lose linearity.

Quote:

Please compare the HD800, PX100 and K601 in terms of mechanical damping with both 50 and 500 Hz square waves. What is the best and why? For me they seem to be on par, but maybe I'm just blind or don't know what to look for...

In terms of constant pressure the order of performance is K601, HD800, and PX100, and linearity is K601, PX100, and HD800. On the 500Hz you're right, the performance is all relatively close.

Quote:

I've seen measurements of HPs at 80 dB SPL with ~3% THD. Let's take the "new" HD238 for example, people claim that they can hear bass distortion.. I'd love to see that on paper.

I'd like to see said measurements, considering off headroom charts the greatest harmonic is -70dB which is less than .5% distortion.

Quote:

Have you read that faq entry? Nothing new besides this:

The headwize one? Yes.

Quote:

"The motional impedance of headphone transducers varies very little (or should vary very little - someone can always do it wrong!) with frequency, so the source impedance can be high with no ill effect." rotflmao. Is this a joke?

Hardly. Tubes have a high output impedance and can perform fine. Some headphones are closer to a resistive load so to speak. You'll also note that the damping factor doesn't need to be extremely high for headphones due to the actual amount of mass. If you're talking a sub with a 50lbs motor structure then damping factor is going to matter a heck of a lot more.

Quote:

I know. That's actually a big problem. Ordered a CAL! a while ago (for modding) and the right side produced over 3 dB more bass (~60 Hz). phhhh..

Yep, that's probably the best thing about moving up a product line is hopefully the tolerances tighten.

Quote:

Tight bass and a good looking 50 Hz square wave are mutually exclusive!

Not . . . really. The only way they would be is if something has an outrageous decay (most headphones don't).

Quote:

Originally Posted by jynweythek I bought the V6 based on most reports saying that it was very flat, and the HeadRoom graph backing that up, which combined seemed like pretty convincing evidence to me. However, it is most definitely not flat (as I explained earlier). The HeadRoom graph is simply wrong for this headphone. But I guess in this case both the graph and (most) people's reports were wrong.

I think you may not be used to hearing a flat phone actually. Judging from ryumatsuba's SR-60 graph I'm guessing he doesn't use a measurement method that takes HRTF into account.

 

[nhancakes]

Quote:

Originally Posted by stang /img/forum/go_quote.gif obviously not. I know it is not a lot better. I was clearly stating judging by the graph, if we were using graphs ONLY and not our ears, it should be a lot better.

yeah that's assuming people think that those bumps on the graph are good things and that people want those. what of the many prefer a flat response curve and many people that pay top dollar for it?

 

[Redo]

It's not just headphone FR graphs, speaker FR graphs start to look identical across the board when you get into higher end gear. They can give you a small idea of what to expect, especially on the bottom end, but they must be heard.

 

[iDriveAGTI]

FR graphs, if you have a good idea about them you can tell with fair accuracy which will be "bright". How? Notice where the bumps and dips are, the human ear DOES NOT have a flat frequency response! You hear certain frequencies way more than others. Particularly frequencies in the range of the human voice and just a bit higher.

Subtle differences in graphs at specific frequencies can produce very different impressions over the long term... Especially if the differences are at frequencies that people are by our nature very sensitive too.

The other thing graphs can tell you is how well matched the drivers on the headphones are. If a graph is included for L & R, closely matched drivers will let you know that the sound stage will be more focused and stable.

 

[jynweythek]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif I think you may not be used to hearing a flat phone actually. Judging from ryumatsuba's SR-60 graph I'm guessing he doesn't use a measurement method that takes HRTF into account.

A different person in the same thread concurred with me that that was exactly how the V6 sounded to them as well. Furthermore, I tried using SineGen and pretty much confirmed the accuracy of the Japanese site's graph. It is most definitely not a flat headphone. It is actually quite colored, but in a way that sounds flat at least to some people (I think the inaccuracies kind of fall into people's blind spots, so to speak). The upper midrange and treble is very exaggerated (making it very bright) and the upper bass/lower midrange has a major dip (causing a lack of warmth). It sounds really quite unnatural. It makes it pretty bad for listening to an awful lot of music, if you want to actually enjoy the music. This sound signature may work fine for monitoring, and this is what the headphone is designed for. For musical enjoyment, most of my cheapo $5-10 headphones are better overall (though there is certainly some music where the V6 sound signature works great).

I also don't buy the idea that a (perceptually) flat headphone wouldn't sound good. It definitely ought to sound good, because if it is perceptually flat that means it is reproducing the music as it is meant to sound (if it is a live recording, it should sound like the live performance). If you mean that it would be flat to a microphone but not to a human ear then I don't see what the point of such a headphone would be.

 

[Shike]

Quote:

Originally Posted by jynweythek /img/forum/go_quote.gif A different person in the same thread concurred with me that that was exactly how the V6 sounded to them as well. Furthermore, I tried using SineGen and pretty much confirmed the accuracy of the Japanese site's graph. It is most definitely not a flat headphone.

If your results shadowed his then I can guarantee it didn't take HRTF into account. Without that the measurements are inherently off . . . it's not like measuring normal speakers.

Quote:

If you mean that it would be flat to a microphone but not to a human ear then I don't see what the point of such a headphone would be.

The headphone, when taking head related transfer function into account, is flat too a good extent. Your perception though may be different . . . but empirically the headphone is fine in the FR area. Thus, when I say you may not be used to a flat phone, I mean when measured properly a headphone that a calibrated mic reads as flat.

 

[jynweythek]

I don't understand the rationale behind HeadRoom's approach. Why would you use an expensive dummy head with simulated ears and all, then correct for the way it hears sound to get the raw frequency output that a simple microphone would measure? If you want to know that, why not just use a simple microphone setup? (I hope I am understanding you correctly).

If the Japanese site's graphs are closer to the human perception of the frequencies put out by a headphone, then it seems far more useful to me, in general. I realize people's ears vary somewhat, but showing the way it sounds to the average ear seems pretty useful.

If that is what is meant by "flat" (though it is often used to mean the same thing as neutral) then I ask again, what is the point of a flat headphone? And why would anyone consider this a good attribute?

 

[b0dhi]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif In terms of constant pressure the order of performance is K601, HD800, and PX100, and linearity is K601, PX100, and HD800. On the 500Hz you're right, the performance is all relatively close.

The "constant pressure" is just a measure of low freq response. In fact the whole square wave response graph is another way of plotting frequency response. Constant pressure just signifies DC response.

Quote:

Yep, that's probably the best thing about moving up a product line is hopefully the tolerances tighten.

This is mostly not the case. Only a few headphone models specify a meaningful higher tolerance. Most tolerance specs are irrelevant since they're only done at a single frequency (usually 1khz). You can find headphones at all price ranges on HeadRoom and compare the L/R matching and see that there isn't really much improvement with price except it special cases.

Quote:

I think you may not be used to hearing a flat phone actually. Judging from ryumatsuba's SR-60 graph I'm guessing he doesn't use a measurement method that takes HRTF into account.

Unless that HRTF is that poster's own HRTF, the graph will be largely meaningless for predicting if it will sound "flat" to him or not.

 

[Shike]

Quote:

Originally Posted by jynweythek /img/forum/go_quote.gif I don't understand the rationale behind HeadRoom's approach. Why would you use an expensive dummy head with simulated ears and all, then correct for the way it hears sound to get the raw frequency output that a simple microphone would measure? If you want to know that, why not just use a simple microphone setup? (I hope I am understanding you correctly).

They use a dummy head to simulate cartilage and how the frequency generally reacts to a canal. The calibration file is used to correct for shortcomings of the mic to make it (the mic) flat.

Quote:

If the Japanese site's graphs are closer to the human perception of the frequencies put out by a headphone, then it seems far more useful to me, in general. I realize people's ears vary somewhat, but showing the way it sounds to the average ear seems pretty useful.

It's backwards. The Japanese site is more likely to just strap on a microphone it seems.

Quote:

Originally Posted by b0dhi The "constant pressure" is just a measure of low freq response. In fact the whole square wave response graph is another way of plotting frequency response. Constant pressure just signifies DC response.

Hardly. Square wave is made to measure overshoot, control over time, and to some extent transient attack.

Quote:

This is mostly not the case. Only a few headphone models specify a meaningful higher tolerance. Most tolerance specs are irrelevant since they're only done at a single frequency (usually 1khz). You can find headphones at all price ranges on HeadRoom and compare the L/R matching and see that there isn't really much improvement with price except it special cases.

Really? That's disappointing . . . I know Grado tries though.

Quote:

Unless that HRTF is that poster's own HRTF, the graph will be largely meaningless for predicting if it will sound "flat" to him or not.

Still, it's more measuring the impact of flesh, cartilage, etc. While there will be some deviations, this helps cover a lot.

 

[Bradan]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif Please compare the HD800, PX100 and K601 in terms of mechanical damping with both 50 and 500 Hz square waves. What is the best and why? For me they seem to be on par, but maybe I'm just blind or don't know what to look for...

Since there is no air cushion to act as a mechanical dampener, I'm sure the point of comparison between these open-backed cans would be the surround of the diaphram, the voice coil, and magnet assembly

Quote:

Originally Posted by xnor /img/forum/go_quote.gif @Bradan: Take the same headphones (as you posted) and compare them using the normal FR. Now rank them again based on how flat their response at the lower end is.

Yeah, I know, I'm in agreeance with you, this part is obvious. That's part of the reason I picked up my D2000

Quote:

Originally Posted by xnor /img/forum/go_quote.gif another example: compare DT700 and HD800 ... so I think that you're looking at the wrong thing. That's why I said that those square waves are not very useful - for me at least.

Yeah, FR graph shows linearity and signature of headphone, but you need a little more info IMO.

Headroom used to have the FR graph for the Apple iBud and, it was pretty damn good and very linear, but only the squarewave will show the quality of the systems used for mechanical dampening.

I think they took the graph off for obvious reasons.

Quote:

Originally Posted by xnor /img/forum/go_quote.gif I didn't. .. and these phase-errors can be found in square wave graphs...?

Here are quotes from headroom's learning guide that can best explain this.

Quote:

Square waves are signals that have many frequency components; the rising and falling edges are fast, and the flat top and bottom are slow. This is a bit of an oversimplification (in fact, square waves can be shown to be made of an infinite series of sine waves at the fundamental square wave frequency and all the odd harmonics in a very specific amplitude and time (phase) relationship. When phase is smeared, the square wave starts to look quite ragged as all it's components become misaligned. For the upper mid and treble frequencies the 500 Hz square wave is very sensitive to phase errors

Quote:

The principles are very similar to the high frequency test, but the lower 50Hz square wave test tells you more about bass and low-mids performance. The headphone's ability to maintain a constant pressure for the length of the flat top and bottom is a measure of how well it can reproduce low frequency notes. This is very difficult as the driver is small with a limited excursion and the earcup is fairly leaky and lets pressure out easily.

 

[jynweythek]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif It's backwards. The Japanese site is more likely to just strap on a microphone it seems.

But if the V6 graph on the Japanese site is matching what I hear, doesn't that mean it's more accurately representing how the headphone sounds?

Quote:

To perform this test we drive the headphones with a series of 200 tones at the same voltage and of ever increasing frequency. We then measure the output at each frequency through the ears of the highly-specialized (and pricey!) Head Acoustics microphone. After that we apply an audio correction curve that removes the head-related transfer function and accurately produces the data for display.

This is what HeadRoom says they are doing. My interpretation is that the dummy head has its own simulated ears that reflect the acoustics of the human ear. What the mic reports should pretty much match what a human would hear (ignoring equal loudness curves, I suppose). Then they remove the HRTF (the acoustic effects of the simulated ears), giving the raw output of the headphone. So the head adds its own HRTF (physically) and then they remove it. Let me know if I am confused here...

The Japanese site explains their measurements here (near the middle of the page). Unfortunately I can't make much sense out of the Google-translated Japanese. Linked here in case anyone wishes to try...

 

[xnor]

Quote:

Originally Posted by Bradan /img/forum/go_quote.gif Headroom used to have the FR graph for the Apple iBud and, it was pretty damn good and very linear, but only the squarewave will show the quality of the systems used for mechanical dampening. I think they took the graph off for obvious reasons.

Marketing reasons?

I'd love to see those measurements.

About the phase-errors:
Headroom says to look at the raggedness of the wave. Do they mean that I should look for how "square" the wave is, or do they mean to look at the oscillation (ringing)?

Anyway, ryumatsuba started to publish graphs that show phase shift errors a while ago. Take a look at this: DT770 and HD650 and look at ~50 Hz. HD650 clearly wins.
Now let's try to do the same with the =853&graphID[]=713]headroom graph.