知道HEADROOM 的请进.(看headroom 提供的耳机频响图要注意的问题。)

Yari

交易区新帖推荐

headroom 估计地球人都知道.看它的图要注意
对了,你说了.就是这样的图.你就是3楼楼长.


原文:Frequency Response

Frequency response of headphones, as we measure it, is the apparent loudness of sound produced by a set of headphones for all frequencies at a fixed drive voltage. Headphones that produce a lot of energy in the lower end of the frequency scale have more bass and are said to be 搘armer?or (if the excessive energy is concentrated in the very low end) 揵oomy.?Headphones with excessive energy in the high frequencies have more treble, and are said to be 揵right.? A frequency response curve (also called a sweep tone plot) is made by putting a pair of headphones on a high quality dummy-head microphone, playing back an audible sweep of all frequency tones from 20Hz to 20kHz (driven at a fixed voltage), and recording the response with an Audio Precision System Two Cascade ($20,000 of the coolest piece of audio test gear you抳e ever seen). The measurements gathered are then compared against an ideal headphone measurement made from the averaged frequency response of ten very good headphones. The resulting difference between the measured response curve and the 搃deal?is then plotted. This basically means that when you look at a headphone graph of a particular pair of headphones, you抮e seeing how they vary from the ideal.

How we measure headphones咱家(headromm)评耳机方法)


     Headphone frequency response is the most important measurement in determining how “good” a headphone sounds, but because frequency response curves are so complicated, they are also very hard to interpret. Even to a trained eye, these things sometimes look like just a bunch of noise. Instead of simply presenting the standard frequency response curve, we try to make it easier for everybody by displaying headphone frequency response as deviations from the frequency response of an “ideal” headphone. We create this ideal curve by averaging the frequency response curves of the ten best sounding (in our opinion) headphones. Thus, when we display the frequency response of a particular pair of headphones, what you are actually seeing is the raw headphone data for that pair of headphones minus the ideal headphone frequency response data. It’s how the headphones you’re looking at differ from a very good sounding pair of headphones.

We know this method is based in part on our perception of what “good” is, but the list of headphones included in the composite measurement include some undeniably great headphones. We feel that the subjective guidance the method gives when interpreting the data is very helpful in selecting headphones that are a good value for the money.

In order to understand how we measure headphones, you have to understand why we measure headphones. Our main purpose in measuring headphones and headphone amps is to make it easier for people to buy them. While this doesn’t mean quite as much when we measure amps (since they are relatively simple to measure and fairly well understood), it means a lot when we measure headphones.

Headphones are surprisingly complex things to measure meaningfully. Significant and relevant interactions between the characteristics of the headphone driver; the acoustic enclosure between the driver and the ear; the acoustic impedance of the enclosure behind the driver; acoustic input impedance and reflections within the ear; and the perceived psycoacoustic importance of each of these things make correlating headphone measurements to experienced performance a daunting task.

Since we don’t manufacture headphones, we aren’t measuring headphones in order to design changes in performance. As mentioned above, we’re measuring headphones to help you buy the right headphones. Thus the most important thing for us is to try to make the measurement more meaningful in terms of your personal satisfaction. We can always just tell you how we feel personally, so our measurements should be fairly objective to give you an “unbiased” way to evaluate headphones.

Unfortunately, as anyone in audio can tell you, it’s almost impossible to design a completely objective way of measuring subjective performance. But rather than do nothing at all, we said, “What the heck, let’s just stick our necks out, take the best intuitive leap we can, and see what happens.”

So we bought Fritz. Fritz is a binaural recording microphone based on a Neumann head. He is basically $6000 worth of the best fake human head microphone you can buy. Then we got our Audio Precision System Two Cascade (the coolest piece of audio testing equipment on the planet) to listen to the headphones through Fritz’s ears. (You have to remember that we do this kind of thing for fun.) You know what we got? A bunch of noisy graphs, that’s what!

We really had a hard time knowing what was what. There were clearly some trends, and we could tell generally whether a particular pair of headphones were bright or dark. But the curves had a lot of high frequency artifacts that just looked like noise.

So we went about looking for way to dredge some meaning from all this noise. We decided to create an “ideal” headphone measurement for comparison. To do this, we picked the ten best sounding headphones we could find (this is the point at which our objective measurements gets a subjective injection), averaged them together, and smoothed the result a bit. We then used this ideal frequency response as a background subtraction against which we could present the measured performance of various individual headphone models. What this means is that when you’re looking at the frequency response curves on one of the headphone product pages, what you are actually seeing is the difference between that headphone and an ideal headphone. The advantage of this system is that it is repeatable, and that we can only influence it subjectively through the “ten best headphones” we pick.

We want to say right up front that this method represents only our first efforts at making an objectively sound and subjectively meaningful headphone measurement. We already know that there are places where this methodology can use improvement, but we also know that it will take at least six months to come up with the next revision of our measurement methods. Since we think there is significant meaning in the headphone data resulting from the current methodology, it’s worth publishing the data.


也就是说该网站提供的图加工过，已经比较“圆滑”了。最重要的这些图表示的频响是同该网站所采用的理想标准（由其所认为的10副最好音质的耳机的数据平均后制成）对比后的产物。所谓明暗,肥瘦都是相对于"标准"的.我理解就是越平越好。

另一点注意的是，这些结果也只能是尽量客观（10副参考耳机本来就是主观选的）。所以结果只能是有胜于无的，主要是方便同志们看明白，先心里有个谱。

还有该网站对耳机评价方式不仅是音质而且包括别的因素，价格（性价比），舒适度等。

我看得不很仔细了，也没按句翻（我也翻不准）。各位还是自己看吧。主要是看它的图时要知道是经过与“标准”比较并且加工后的，就行了。

咱家,对是她家弄这点图不易呀.花了不少票子.那个假脑袋6000美刀,那个分析系统2万美刀,就整出这点图来,每张图都是血汗.啥?你想改进,我还不瞎说,咱不提钱,你重搞一下少说也要半年.呵呵

最后注意，我看得不仔细，不会写错了吧。哈哈