- 第 4 页 -
原帖 由 haku_wang 于 12-5-13 03:02 发表
另外，觉得VI的弦乐音色即使在MIR中也还是有点无法达到实录的那种厚重感。我曾经在VSL论坛上问过Dietz有关这件事情。因为即使是MIR也一样是通过卷积来进行混响处理的。而对于弦乐组这样的乐器来说，只有两个声道的录音以及两个声道交叉的卷积还是不足以还原现实中一组弦乐在那么多不同位置所造成的混响效果
MIR在VE PRO5里是可以设置成环绕输出的,如果设置了环绕输出,它的话筒也显示成多只话筒,这应该不是两个声道交叉的卷积吧?
而且在对音乐厅采样时,它也是每个采样点采8次,就是水平方向每60度采一个,向上/向下各采一个, 按维也纳的说明,如果拿钢琴音色来做比喻,其它采样混响效果器就是采一个C3键,然后分布满整个88键做成的音色, 而MIR是每个键都采了(音乐厅中的多个采样点), 而且每个键还分了力度层(同一点的多方向采样), 就是它说的多重采样.
也许我没理解您的意思,如果您是说一组弦乐人很多是一片,不可能在用一个点上的采样来还原,当然这的确是MIR没做到的
原帖 由 xixi 于 12-5-14 14:56 发表
MIR在VE PRO5里是可以设置成环绕输出的,如果设置了环绕输出,它的话筒也显示成多只话筒,这应该不是两个声道交叉的卷积吧?
而且在对音乐厅采样时,它也是每个采样点采8次,就是水平方向每60度采一个,向上/向下各采 ...
MIR是每个采样点都进行了采样，但是这个采样仍然是基于一个有限空间内的点（比如每1平方米一个采样点，采集各个角度和方向的IR）。具体进行卷积运算的时候，也仍然是根据输入的音频的通道位置来进行的。输入双声道的话，如果输出双声道则是4次卷积。如果输出多声道则会更多（实际上，可以认为是，双声道的输入与多声道或双声道的卷积进行运算）。
MIR并不会将所有位置都与输入位置的声音进行卷积。如果只在MIR中使用一个双声道输入，其实它也就与Altiverb的插件没有太大的区别（顺便说，MIR的核心卷积处理算法和技术正是Audioease也就是Altiverb的开发商提供的）。只是如果有多个输入，可以根据不同的位置，用不同的位置相对mic的IR进行卷积而已。我不知道MIR有没有考虑现实中mic的漏音交叉问题。
我说的意思的确是指在弦乐组中，每个独立的弦乐的位置所造成的不同的混响的叠加，造就了弦乐组的厚重感。这种每个琴的位置不同所造成的混响区别和叠加效果MIR确实没有做到。这也是为什么VI的弦乐很难做的很丰满的主要原因。毕竟对于弦乐组来说，这是一个类似积分的过程。无法用两个声道或者几个声道简单的解决。唯一最好的方法就是真的模拟出14个人的独奏小提琴，放在MIR中的不同位置（如果MIR的采样点分布足够密集的话）。我曾用chamber strings和orchestral strings做过实验，确实有提高，但是还是很难做到实录的感觉。
MIR的多方向采样的目的是模拟不同的方向的区别。这点上是一般卷积混响没有做到的。虽然本质来说这只是占用了更多数据的缘故。MIR的真实感其实还有一个原因是来自于其针对不同乐器的Profile。主要包含了不同乐器类型在现实中对空间的声音的频率响应的区别，以及一些其他的参数。
我是听了VSL上Jay Bacal的Williams示范得出的结论，还跟CD作了对比，听起来Insp:IR更理想。那也许是Jay的问题，但我不敢买也不能用，因为跟其他工具不同，VE PRO 和MIR都是在DAW之外运行的，要单独设置联接甚么的。作为盲人，我只能在Sonar内部操作，而且VSL所有的软件读屏都不能支持，我只能靠别人帮助搞一些Track Templates，然后通过Track Inspector调整Vienna Suite插键的参数。而MIR还要摆位甚么的，还有很多参数，如果不调，甚至还不及一斑的卷积。也许你们中的人可以比Jay作得更好，那就是我因为无法实践而得出的错误结论。
冈收到考稃漫关于此疑问的回答：
How should music sound? First of all it is a matter of taste, secondly it is a matter of how people are used to listen to music and a third point is also, that it isn't a good idea to simulate a real concert atmosphere for speakers. So a good mix is contains a bit of all. It should be optimized for the use with speakers and headphones and should also simulate a bit the situation of a real concert.
Unfortunately a lot of people are only used to listen to music with headphones. Further a lot of people do not know any more how it sounds in a real concert and so on.
There were endless discussions in a infinite number of forums. Further, people who don't use samples but midi sounds will have another opinion anyway.
The problem of MIR is (in my opinion), that all the sound comes out from one "acustic pot" - if you use it without adjust several things.
That's the situation of a concert visitor but not what we know from classical CDs and Mixes at home. So in a way MIR sounds natural but also not.
A good mix should be transparent so that we can make out most of the instruments.
It should not sound muddy and nasty, no we should be able to listen to it for a long time with pleasure.
This means it should be balanced in the frequency and in its dynamic as well.
I personally can adjust all these things in a better way with single effects than with MIR.
Nevertheless, MIR could be a fast alternative for first results for those people who don't have the time and the knowledge to use all the effects.
TIP: If you adjust the wet/dry ratio within the output section of MIR a bit more to dry you often get a better result with less "pot-sound".
You are at the point now to find your own sound,...
A very good way could be to search for mixes which you like and with a sound you think: That's the one I would like to have as well.
Then it's up to you to analyse this mix and finally to copy it so to say with your effects and possibilities.
[ 本帖最后由 osaka2011 于 12-5-15 12:32 编辑 ]
冈收到考稃漫关于此疑问的回答：
How should music sound? First of all it is a matter of taste, secondly it is a matter of how people are used to listen to music and a third point is also, that it isn't a good idea to simulate a real concert atmosphere for speakers. So a good mix is contains a bit of all. It should be optimized for the use with speakers and headphones and should also simulate a bit the situation of a real concert.
Unfortunately a lot of people are only used to listen to music with headphones. Further a lot of people do not know any more how it sounds in a real concert and so on.
There were endless discussions in a infinite number of forums. Further, people who don't use samples but midi sounds will have another opinion anyway.
The problem of MIR is (in my opinion), that all the sound comes out from one "acustic pot" - if you use it without adjust several things.
That's the situation of a concert visitor but not what we know from classical CDs and Mixes at home. So in a way MIR sounds natural but also not.
A good mix should be transparent so that we can make out most of the instruments.
It should not sound muddy and nasty, no we should be able to listen to it for a long time with pleasure.
This means it should be balanced in the frequency and in its dynamic as well.
I personally can adjust all these things in a better way with single effects than with MIR.
Nevertheless, MIR could be a fast alternative for first results for those people who don't have the time and the knowledge to use all the effects.
TIP: If you adjust the wet/dry ratio within the output section of MIR a bit more to dry you often get a better result with less "pot-sound".
You are at the point now to find your own sound,...
A very good way could be to search for mixes which you like and with a sound you think: That's the one I would like to have as well.
Then it's up to you to analyse this mix and finally to copy it so to say with your effects and possibilities.
[ 本帖最后由 osaka2011 于 12-5-15 12:32 编辑 ]