这几天在群里看到讨论summing的势头又强了一些,当然我并不是说准备影响商家和制造商的利益,我就是单纯的谈谈summing mixer的一些经常被我们关注到的方面。
大家基本上想到模拟硬件,一开始会想到谐波失真、频响失真,SNR。
我在waves群里谈到crosstalk,也跟其他老师私下里交流过,其实这个概念有时候还是蛮有趣的,具体我就不再说了,大家可以看看SOS的编辑评论。
SOS Technical Editor Hugh Robjohns replies: Crosstalk is something most audio designers have striven to minimise and avoid for over a century! It is the unintentional and (usually) unwanted ‘bleeding’ of sound from one channel into another, and it is an inevitable and inherent issue in analogue systems.
In analogue systems, crosstalk almost always gets worse as the frequency rises, because it generally manifests as a capacitive coupling — so crosstalk normally sounds like a thin and tinny low-level bleed of whatever is on the other channel(s). Clearly, if you’re multi-tracking something you don’t want crosstalk from one channel to interfere with the other. Crosstalk can be minimised (and often reduced well below audible levels) by careful PCB layouts, grounding and good wiring arrangements, with balanced wiring being much less prone than unbalanced, for example.
However, there are some beneficial aspects to crosstalk in certain circumstances. Most notably, when working with a stereo mix, crosstalk can affect the stereo image and width. If the crosstalk is in phase it will tend to reduce the image width, but if it’s out of phase it will tend to widen it. Sometimes that can be helpful, especially as it tends to be a frequency-conscious effect, generally getting stronger at higher frequencies.
There’s no accidental crosstalk within a digital system, and it has to be deliberately coded in if required.
SOS Reviews Editor Matt Houghton adds: In addition to what Hugh says, if you wish to satisfy your own curiosity about the sonic effects of crosstalk, you can simulate the process quite easily in your DAW — and can even tweak it for creative effect. For example, start with a stereo source (a mixed track, say, or a drum loop). Set up a send from your stereo source to a new stereo bus on which you’ve reversed the stereo panning, so that the source’s left channel is sent to the right channel of this bus, and the right to the left. Keep the bus’s fader at unity gain, and slowly increase the send amount from the source, so that its left and right channels start contributing to the bus’s right and left channels, respectively: you’re introducing crosstalk. (When comparing the untreated source with the ‘crosstalk-enhanced’ version, the latter will be slightly louder, which can skew your perception of which sounds ‘best’).
For a more detailed imitation of the capacitive-coupling crosstalk effect that Hugh describes, keep the send level quite low, and apply some EQ to simulate the way this emphasises higher frequencies — a combination of high-shelving and low-shelving boosts (ie. a ‘tilt’ EQ) should give you the general idea. Choose a minimum-phase rather than linear-phase EQ, as this will also introduce some phase rotation that the capacitive coupling typically introduces. For authenticity, you’re looking for 6dB per octave shelves, but for creative effect, you can do as you please.
This will get you into the right ball park... though it’s worth pointing out that in the real world of analogue electronics the crosstalk might enter the signal path at more than one point, and could affect unrelated sources on adjacent channels. A decent console will have been designed to avoid crosstalk in the internal signal path as far as possible, making the insert points and cables to patchbays more likely places for significant capacitive coupling to occur than the mixer’s own channel strips.
Hugh Robjohns阐释了crosstalk是怎么一回事,它有点像“程度较低的流血效应”,也就是我们常常说的漏音。
同时给出了它的影响,例如,对立体声声像的影响。
Matt Houghton主要是讲了一下如何模拟出crosstalk。
高端调音台往往会强调对crosstalk的影响,所以他们会在PCB设计上精心调整,以保持crosstalk在较低水平。
但是值得一提的是在数字系统,例如DAW中,不存在串扰,除非你故意为之。例如Studio One的内部summing系统:Mix Fx可建模调音台的crosstalk,并经由passthrough特性传递至所有的子Mix Fx,就从所谓的软summing来说,studio one远远强于所谓的Pro Tools HDX系统,因为HDX系统并不准备模拟crosstalk.
总得来说,crosstalk在较大的情况下,能对立体声声像产生较为明显的影响,当然这也取决于频率。
因此,为什么“summing系统听起来更好、更宽、更顺滑、更刺激、更有力、更加具有深度”,其实是可以考究的。
当然,根据本人多年经验,summming的作用时常被夸大了,往往它们并
不会产生过于深刻的影响——取决于你选择的summing设备、summing设备的水准,以及混音的频谱分布。
summing box相较studio mixer来说,它起到的作用很简单,你在ITB中混音,在OTB中求和,但是它们又回到了ITB。
大多数情况下你直接通过母带路由系统监听母线,或者只监听summing mixer,但是毫无疑问,通过这两种路径都会影响你的混音判断。
如果你准备购买一台,我的观点是你可以购买,在某些时候它能让作品更好。
你可以不购买,因为它并不是任何情况下都适用的,特别是你的混音并不是很好....
混音过程可以让你的作品本身具有很深的深度,summing系统不是必需品。
如果一定要给出一个答案,那么summing box的好处和坏处是看脸的,你可以混完,过一次然后对比。
切记,当你觉得好的时候,一定要给出:哪里好了?是否真的这里好了?
我们有义务保证素材的质感,而不是疯狂的加入各种调料:这样炒皮鞋也很香啊!
以及:
64bit可提供三百多db的动态范围。即使32bit float也能无视动态范围的损失,所以别tm扯淡了,至于hd accl能比native音质好纯属脑袋掉粪坑里头糊涂了,hd accl只有32bit的处理精度,完全是个弟弟,你说hdx也就算了,最起码它求和精度是64bit float。
同时数字系统它们的dr就是dr,而不是snr,SNR引入了所有的噪声源。
宿主求和虽然在数字系统中产生digital artifacts,但是他们的失真远远小于硬件,所以实际上所谓summing可以增加动态范围实际上是一句悖论——可能意义不同,看广告词时候忽略这句话就完事了。
如果还搞不清楚SNR与DR的关系可以参考https://en.wikipedia.org/wiki/Dynamic_range,特别要说的就是厂商比较会忽悠,因为音频、音乐、电子三者对于动态范围的定义是不同的。
以上属于本人口嗨欢迎纠错。
本帖最后由 qianyao 于 19-12-16 22:41 编辑