【雲中來的音樂 - Computer Audio】家訪 Eton

[asdffx]

asdffx 兄，啊！您每次都問到核心。是的，我的喇叭前後也動了，結果並不是往前靠近 Sweet Spot，而是後退 30mm + 移開中軸線 90mm (各 45mm）+ Toe-out 各5度 （Toe-in 15度 變成 10度)，反而增加了 10K~20KHz 10 dB。我沒第一時間分享，是這三個移動方向都違反了我的喇叭擺位常識，您的建議很好，我會去量 RT60，工作日太忙，我會在週末執行後分享。

那麻烦eton了，慢慢也不急。

[Eton]

這張照片是在這音響室啟用兩年多來喇叭拉的最開、Toe-in 最少的一次，但是 10K ~ 20KHz 卻救回了 10dB，音場中央最聚焦的一次!? 真不知該怎麼解釋。

DSC00174.jpg

asdffx 兄，啊！您每次都問到核心。是的，我的喇叭前後也動了，結果並不是往前靠近 Sweet Spot，而是後退 30mm + 移開中軸線 90mm (各 45mm）+ Toe-out 各5度 （Toe-in 15度 變成 10度)，反而增加了 10K~20KHz 10 dB。我沒第一時間分享，是這三個移動方向都違反了我的喇叭擺位常識，您的建議很好，我會去量 RT60，工作日太忙，我會在週末執行後分享。

[davidchang]

太幸福了 David 兄！ 您覺得 CAS 打敗了 SACD Transport 嗎?

目前尚未購入Transport，朋友最近音響室完工，我先慫恿他買，這樣就有得比較，只是Transport的性價比低了點。

[powerman]

這張照片是在這音響室啟用兩年多來喇叭拉的最開、Toe-in 最少的一次，但是 10K ~ 20KHz 卻救回了 10dB，音場中央最聚焦的一次!? 真不知該怎麼解釋。

小弟的愚見:


不一定要Toe-in, 聲音才聚焦. Toe-in過頭, 聲音不容易寬鬆自然.
沒有絕對的擺位公式, 要考慮Room Mode, ​所以喇叭要嘗試上下左右前後移, 與空間互動. 只要聲音寬鬆自然, 方向就對了.

以及... 要聲音聚焦, 容易得很.


首先,
前級要有雙outputs, 或用Y-cable
不須toe-in, 零度.
只接left or right channel, 錄音變成mono. 理論上, 聲音應該從兩喇叭之間中央一點發聲, 而且這點必須越小越好.
如果這點一大塊不中央, 嘗試上下左右前後移, 把它變小及中央, 還原之後(revert back to stereo out), 音場定位必改善.


但是, 在移動喇叭的過程中, 必須保持聲音的寬鬆自然及有深度.

[Eton]

謝謝 Powerman 兄，調喇叭密技又多一種。
是否 Toe-in 的事，的確也各有擁簇，值得玩味的是，早期的 Avalon 在操作手冊裡皆提倡不要 Toe-in ，自從換了 Isis 後，操作手冊 （page 14) 就改了，衷於原著就恕不翻譯了：
Adjusting the toe-in angle of the speakers is useful in tailoring the sound to best match the characters of your system and listening room.
When the speakers are facing straight forward, they tent to create a large, expansive sound-stage, painted with broad brush strokes.
As they are rotated forward the listening position, the image becomes more compact, with increase focus, creating a greater sense of intimacy.

[asdffx]

謝謝 Powerman 兄，調喇叭密技又多一種。
是否 Toe-in 的事，的確也各有擁簇，值得玩味的是，早期的 Avalon 在操作手冊裡皆提倡不要 Toe-in ，自從換了 Isis 後，操作手冊 （page 14) 就改了，衷於原著就恕不翻譯了：
Adjusting the toe-in angle of the speakers is useful in tailoring the sound to best match the characters of your system and listening room.
When the speakers are facing straight forward, they tent to create a large, expansive sound-stage, painted with broad brush strokes.
As they are rotated forward the listening position, the image becomes more compact, with increase focus, creating a greater sense of intimacy.

呵呵，真巧，我也刚在看isis的说明书呢。网上下的。
会不会是垂直的影响而不是水平的影响呢，isis的高音是从高向下打向人耳朵。会不会之前从上向下的角度有些偏差呢？
见过有些喇叭设计的时候考虑到离轴和正轴的平均值，是会建议不用toe in。因为在家具环境，离轴的影响也不会少于正轴。

powerman兄那个mono的方法，我类似的也试过，因为刚好我在用的喇叭说明有介绍过，不过是用pink noise的。
下面的说明书摘录的：
PINK NOISE LISTENING TEST FOR FOCUS: Play a pink noise source such as a CD with test tones or an FM tuner set between stations (with muting turned off). You will notice one of two sonic effects when listening to pink noise:
1. If you are in the focal point of the array, you will hear a “ball” of noise that appears to float between the speakers. Although noise will appear to emanate from the speakers themselves, the center image will be stronger, with a slightly different tone quality to the phantom center image. The sound from the speakers themselves will seem slightly detached from the pink noise ball, and slightly softer in volume level. The center “ball” of sound will be very strong, and you will be able to detect that the “ball” of sound is round, with depth to the image. This is a ‘locked in’ image with correct focus.
2. If you are not sitting in the focal point, the noise will be diffuse, without a center image. Although you will hear noise coming from the center, it will sound as diffuse as the sound coming from the speakers themselves. Even though the speakers will image music when set up like this, the imaging will be diffuse and non-localized. There won’t be enough ‘body’ to the images, and it will be hard to hear space between the instruments. There is a distinct lack of image focus in this situation.
To find the focal point of the VR system, move forwards and backwards slowly while listening to the pink noise ‘ball’. You can either rock in your listening seat, or move back and forth by several feet if you are way out of the equilateral triangle (using a director’s chair for this works great). If you have a fixed seating position that can not be moved due to furniture considerations, you will have to move the speakers either closer together, or further apart, depending on the strength of the pink noise ‘ball’. If the speakers are closer together than 8 feet, try spreading them further apart.
*Note: if the speakers cannot be separated more than a few feet due to room considerations, as in a video application, it may be necessary to toe the speakers outwards.
If you are placing the speakers further apart than 8 feet, try toeing them in towards you, listening for the ‘ball’ to lock in. It really helps to have two people moving the speakers while you sit in the listening position, giving orders. Toe-in is often critical, with only 1" movements being very apparent. When you have found the correct spacing and listening distance, you will immediately know it, since the image lock-in is very obvious. The pink noise will become a very strong ‘ball’ of sound, just floating in between the speakers. Keep moving the speakers until you hear this ball of sound. As you move the speakers either closer together or further apart, and experiment with toe-in, you will hear changes that enable you to know if you are headed in the right direction. It is not really difficult, it’s fun. Of course, experimenting with several types of placement schemes is valuable, for it will teach you how the speakers interact with your particular room. Only by careful experimentation will you find that ‘magic’ spot for your new VR-4 speaker system.

[Eton]

asdffx 兄所言甚是，Isis 的中、高音單體都是下投式的 （其實這也是ㄧ種垂直向限的 Toe-in），我現在的座位的高低與距離喇叭的長度相對高音的軸線是敏感的，上下移動 Omnimic V2 的麥克風的位置已發現高頻響應的變化，像 Isis 這種單體不在同一平面上的喇叭，對於座位與喇叭的距離，還有座位的高低要有更多調整上的考量。

[Eton]

這次擺位調整的結果，耳聽比以前的擺法是相對滿意的，但是老問題 30Hz 的凹陷沒改變 (不敢苛求)，今天再做一次較完整的OmniMic V2 量測，老實說，這些數據並看不出與以前量測的有明顯的差異，與大家分享，有任何高見，歡迎指點。

2013/03/30 這ㄧ整天的喇叭擺位工作，是聽感輔佐了 OmniMic V2 頻率響應圖而決定下來的：
空間：寬 4050mm，長 9220mm，高 2500mm (平均)
喇叭正面中軸間距：2340mm
喇叭正面中軸與側牆：855mm
喇叭正面中軸與後牆：2663mm
喇叭 Toe-in：10度
聆聽位與喇叭正面： 3600mm (右/左耳距右/左高音單體中心：4120mm)
聆聽位與後牆：2957mm
以上長度量測使用 Bosch DLE40 Professional 雷射測距儀進行



Frequency Response - 紅線是改變後
20130331-speaker-position.jpg

Bass Decay
20130405-1213pm-Bass-Response.jpg

RT60 - 17.54Hz ~ 22.63KHz
20130405-1209pm-RT60.jpg

Energy Time Curve
20130405-1211pm-ETC.jpg

Wavelet Spectrogram
20130405-1226pm-wavelet spectrogram.jpg

這張照片是在這音響室啟用兩年多來喇叭拉的最開、Toe-in 最少的一次，但是 10K ~ 20KHz 卻救回了 10dB，音場中央最聚焦的一次!? 真不知該怎麼解釋。

DSC00174.jpg

[Yeh捨]

我照主人的標示數據以CAD畫平面圖.
3600mm喇叭至人耳的距離.這應該指的是喇叭平面連線的中心到人耳吧 ?
那這時左右喇叭中心到人耳中心實際值應該是3976mm.而非4120mm噎.

差144mm並不大.以音速時間差計算之是0.423msec.(單程沒有往返)
那這一點到後牆的距離是2957mm沒錯的.

這時的 toe in10度角(指的是向中心人耳轉10度吧).
那將會與正對人耳軸線差7度角.兩隻喇叭中軸線交會點.將會在人耳後方3002mm.
屬於人耳後交會.這種方式應該也是多數燒友的擺設法.?

以7度角的偏離.其離軸高音衰退值幾乎可以不計的..
因為一般測喇叭是以30度角為1單位.

[Yeh捨]

不過以我觀察這空間的測量圖有一段時間..
似乎.最高端的測量值應該不至於因為 toe in小改變就會有10dB的變化.?
10khz以上差距10dB是很大的值 !

主人空間的聲學處理絕大部份的面材能作用的頻段應該都在800hz以上.
我指的不是處理低段的圓柱歐(其相對面積是少很多)..
但這也是一個相對算空曠的空間.對於較低端頻率消散是相對偏慢的.
看 bass decay圖就很清楚.
40hz以下(極低)可以先不要管他.90hz處的小峰值超過0.21sec是比較要注意的.
指教囉..