Long time ago I asked about how hitting nebula, because the reason(and some developers like UAD or the old analog way along the history) tells you it must be -18 dBFS RMS, but somewhere I found this was related to PEAK, and it was so strange to me, because -18 dBFS peak implies an ultra low RMS level. I asked on a post and the ultimate answer giancarlo told me was that nebula need (for sweet spot I mean, and except for libs from cdsoundmaster or others who advice it on a same way) to be feed with a -18 dBFS being this a PEAK level... It was strange to me but if giancarlo tell this is the way I believe it must be.
I still find it strange but maybe nebula needs it. Please let me know if Anyone of you knows really if this is the right rule for nebula or if something was changed or misunderstood.
Hi, the way I do it: I use Klanghelm VU meter, and treat it like you would on an analog console, adjust the gain either in the plugin or preferably using clip gain to get the needle towards 0dbVu. With drums, I check that I don't get the yellow or red light...
Well, I can see my last posts have not much sense. I posted them without having thought about it, just because the topic was alive again one year after.
So now I re-read the whole thread and some things begin to put in place.
First at all I think there are two differents matters or questions regarding to gain stage, and they use to be mixed on our conversations so it contributes to increase the mess. Maybe we could try a different approachment to the matter so we can clarify it. The two things I am telling about are:
1 - How to do gain stage in order to record on digital in a proper way and using general plugins(not relating analog gear behaviour), where we obviously must consider two aspects: the limit of 0 DBFS and the needing of leaving enough headroom
2 - How to hit Nebula when using a library of an analog gear and (this is important) our goal is to get the closet behaviour to that analog gear
Usually on the forums(not just ours) people tend to mix one or other scenario so the advices use to be "not trespassing 0 DBFS", "use your ears", "leave headroom" and so on, things which give benefits but they are good practices for a general context of working on digital workflow.
A different thing is when you want to emulate a piece of analog gear using nebula. For example, let us consider a vintage console emulation. This console has its own behaviour and its response depends on the input level. The sampled console is not exactly the real console, I mean you can send any input level to real console and it always will give you a "consoled" response, but you can not send any input level to the sampled one and expecting it gives you exactly what the real console does. There is a range of input levels, inside it you will feed the sampled console and you will get the console response(a close response), but outside that range you will get a response "coloured" by the library but different from what the real console does(it maybe can be a good sound but it is not what console does).
So when trying to get the best from the nebula presets in order to recreate(in the closet way) the sampled real gear we must know which levels excite the presets. If we send a -35dBFS peak signal to the console preset we will get a "not very related to original console" response, and if we send a very ver very hot signal we maybe get the same, a not very close to original response. Of course maybe there are presets for ultra hot or ultra low signal levels(like Michael's ones for some libs from him, it depends on the developer way of doing it), but even in this scenarios we should send the right level to the right preset: we should send a -5 level to a preset called "-5". If we send a -18 signal to a "-5" labelled preset we are not feeding the tool with the right food, the response may be a good one and we could even benefit from it on a creative way, but if our original purpose is "getting the response of that analog gear" we are not on the way. It is like playing with the Drive knob of presets, it may gives you not bad or even good responses but it gets away from the sampled gear response or THD.
"Use your ears" or "leave headroom" advices are OK but they are pointed towards getting a general good sound. When our goal is translate hi end gear to our digital tracks we must go beyond, of course we have to leave headroom and use our ears, but getting a close(for example)"N**e" console response for our tracks implies we have an extra task which is taking care of the level we send to the sampled tool.
So I can see the importance of knowing what level needs a preset in order to feed it with it and so we can get the closet response. That is why I (and I think there are several ones amog us) am often asking about this matter.
I think a solution must be getting a list of every library asking its developer the input level needed for every one of them(I know this may be a herculean task). SOme devs tells about it on theirs manuals, some not. And other fantastic idea I just read about is having an indicator of kernels being used , which would let us know if our input level is exciting the "original" behaviour of the preset.
Please let me know your thoughts about this. Maybe this approachment helps to separate things and we can address the advice on one or other way.
Thanks in advance for reading this messy post and excuse my surely messy english
Can soneone clarify what the correct levels are to work with the CDSoundmaster libraries & stand alone plugs ? There seems to be the suggestion above they are different with the gain staging than regular Nebula/ Aqua, which is news to me (?)..