How do you make sure you are hitting Nebula at -18dbfs?

Tips & tricks, working results, technical support
User avatar
zabukowski
Expert
Expert
Posts: 1645
Joined: Sun Mar 28, 2010 8:44 pm
Location: Ljubljana, Slovenia
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by zabukowski » Sat Jan 10, 2015 3:51 pm

I haven't read whole thread ... in case you didn't mention this yet...

NebulaMan offers two types of input normalization - before processing with Nebula.

1. peak RMS
2. TRUE peak

Normalizing to -18 dbFS peak/max RMS is a very close approximation to 0 dbVU (since VU meter is a different thing than RMS meter). Of course it depends on source material - you can check this in NebulaMan 2 loading your VST VU meter plugin directly after normalization stage.

Peak RMS is not always suitable - for material with hi-peak and low RMS power (drums for example), TRUE peak normalization should be used - value -6db is a good starting point. Again it depends on dynamic range of the material, but in most cases it works nice.

Cheers,
Zabukowski

User avatar
Sheikyearbouti
Member
Member
Posts: 318
Joined: Wed Feb 16, 2011 6:49 pm
Location: London, UK
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by Sheikyearbouti » Sat Jan 10, 2015 4:14 pm

--
Nikolay Georgiev
http://georgievsound.com/

User avatar
ngarjuna
Expert
Expert
Posts: 782
Joined: Tue Mar 30, 2010 5:04 pm
Location: Miami

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by ngarjuna » Sat Jan 10, 2015 4:37 pm

Perhaps I'm missing something vital but I don't entirely understand the very need for batch processing in this manner. If you're a professional engineer then you should already understand what's been pointed out (and was effectively just reiterated by Zab): that a straightforward system of rules are wholly insufficient for gain staging audio in the first place. My clients want, need, expect and receive personal attention to every detail of their production; that's why they pay me. And working in television I can pretty well guarantee my deadlines are as tight as anyone in the business; so yeah, I get wanting to save time but not at the expense of actually doing ones job. And if you're not a professional how many tracks are dealing with that you're so inundated that you can't even address, stage and process everything a track at a time? How much time are you even saving if you have to listen, measure and pre-decide which set of rules to use before you batch process anyway?

I can't help but feel the attempt to automate audio production has done little to enhance the field. I also can't imagine how this goes along with the ethos of Nebula (helping users to return to the days when sound processing mattered and sound quality trumped convenience).

thomasd
User Level VII
User Level VII
Posts: 73
Joined: Fri Nov 19, 2010 6:49 am

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by thomasd » Sat Jan 10, 2015 4:37 pm

Google Chrome can do that for you.
Macbook Pro 2013, Apogee Symphony IO Thunderbolt,
Logic Pro 10.2 , Protools 12, Cubase 8.5,Nebula Server, all Aqua's.
Acustica Audio dealer for the Netherlands
http://www.dippelstudio.nl, http://www.dippel.nl

User avatar
david1103
Expert
Expert
Posts: 517
Joined: Wed Mar 31, 2010 12:26 am

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by david1103 » Sat Jan 10, 2015 4:41 pm

Sheikyearbouti wrote:
Translation? ;)
https://translate.google.co.uk/translat ... edit-text=

along with the great new phrase "You're getting the cock a mess" :lol:

User avatar
zabukowski
Expert
Expert
Posts: 1645
Joined: Sun Mar 28, 2010 8:44 pm
Location: Ljubljana, Slovenia
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by zabukowski » Sat Jan 10, 2015 4:43 pm

ngarjuna wrote:Perhaps I'm missing something vital but I don't entirely understand the very need for batch processing in this manner. If you're a professional engineer then you should already understand what's been pointed out (and was effectively just reiterated by Zab): that a straightforward system of rules are wholly insufficient for gain staging audio in the first place. My clients want, need, expect and receive personal attention to every detail of their production; that's why they pay me. And working in television I can pretty well guarantee my deadlines are as tight as anyone in the business; so yeah, I get wanting to save time but not at the expense of actually doing ones job. And if you're not a professional how many tracks are dealing with that you're so inundated that you can't even address, stage and process everything a track at a time? How much time are you even saving if you have to listen, measure and pre-decide which set of rules to use before you batch process anyway?

I can't help but feel the attempt to automate audio production has done little to enhance the field. I also can't imagine how this goes along with the ethos of Nebula (helping users to return to the days when sound processing mattered and sound quality trumped convenience).
For example, I often use auto normalizing when receiving all kind of material for mixing which is usually far from optimal.

On the other hand, If everything is perfect from the start and you have control over it, there is no need for additional normalizing etc. So i am trying to use the right medicine for the specific problems :)

User avatar
ngarjuna
Expert
Expert
Posts: 782
Joined: Tue Mar 30, 2010 5:04 pm
Location: Miami

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by ngarjuna » Sat Jan 10, 2015 4:51 pm

zabukowski wrote: ...
For example, I often use auto normalizing when receiving all kind of material for mixing which is usually far from optimal.

On the other hand, If everything is perfect from the start and you have control over it, there is no need for additional normalizing etc. So i am trying to use the right medicine for the specific problems :)
You're right, Zab, I should have made this distinction:

I think it's a poor practice to let batch processes make decisions for you. Once you've made a decision it's clear that there are cool tools to facilitate such choices. So my criticism is not that such tools exist but rather how we try to use them to replace decision making. If you have to batch process everything 20dB to make it usable that's a lot different than using normalization to decide how hard to drive any particular Nebula instance. To me, anyway.

User avatar
david1103
Expert
Expert
Posts: 517
Joined: Wed Mar 31, 2010 12:26 am

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by david1103 » Sat Jan 10, 2015 4:53 pm

ngarjuna wrote:Perhaps I'm missing something vital but I don't entirely understand the very need for batch processing in this manner. If you're a professional engineer then you should already understand what's been pointed out (and was effectively just reiterated by Zab): that a straightforward system of rules are wholly insufficient for gain staging audio in the first place. My clients want, need, expect and receive personal attention to every detail of their production; that's why they pay me. And working in television I can pretty well guarantee my deadlines are as tight as anyone in the business; so yeah, I get wanting to save time but not at the expense of actually doing ones job. And if you're not a professional how many tracks are dealing with that you're so inundated that you can't even address, stage and process everything a track at a time? How much time are you even saving if you have to listen, measure and pre-decide which set of rules to use before you batch process anyway?

I can't help but feel the attempt to automate audio production has done little to enhance the field. I also can't imagine how this goes along with the ethos of Nebula (helping users to return to the days when sound processing mattered and sound quality trumped convenience).
I think in the future everything will be automated and it will be done to an amazing high standard. Right now we are just starting on this. There will be very few jobs done by humans in the future. Either we will all be sitting in the sun drinking beer or the nut jobs will have blow us all up.

Seriously though. I understand what you are saying, but it really depends on exactly what you are doing. Here is an example that has happened to me.

Mastering a compilation album from 15 different bands all totally different volume levels. I want to listen right away to them all to hear what is going on, and also to start playing with Nebula EQ.

1. Load all the tracks into Reaper
2. Normalize to -18dbfs at a press of a button with the SWS extension
3. Stick Nebula on the master bus and start to listen and experiment

I just saved a hell of a lot of work that had no real artistic or technical merit. I will of course adjust everything by hand eventually. Without bringing all the tracks to about the right level Nebula would be distorting all over the place and I would have no headroom to make boosts.

User avatar
zabukowski
Expert
Expert
Posts: 1645
Joined: Sun Mar 28, 2010 8:44 pm
Location: Ljubljana, Slovenia
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by zabukowski » Sat Jan 10, 2015 4:55 pm

ngarjuna wrote:
zabukowski wrote: ...
For example, I often use auto normalizing when receiving all kind of material for mixing which is usually far from optimal.

On the other hand, If everything is perfect from the start and you have control over it, there is no need for additional normalizing etc. So i am trying to use the right medicine for the specific problems :)
You're right, Zab, I should have made this distinction:

I think it's a poor practice to let batch processes make decisions for you. Once you've made a decision it's clear that there are cool tools to facilitate such choices. So my criticism is not that such tools exist but rather how we try to use them to replace decision making. If you have to batch process everything 20dB to make it usable that's a lot different than using normalization to decide how hard to drive any particular Nebula instance. To me, anyway.
-
Very well said !!!

Cheers,
Franci

RJHollins
Expert
Expert
Posts: 4050
Joined: Sun Mar 28, 2010 5:53 pm

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by RJHollins » Sat Jan 10, 2015 6:08 pm

additionally ...

Gain staging is a dynamic procedure.

It starts at the source level recorded. As the signal is processed [eq, fx, etc], levels within the chain need to be monitored/ re-balanced.

The 'standard' is called 'Unity Gain'. What goes IN = what goes out. Once that is established/understood, creative gain structures are then predictable. 'Unity Gain' is still the guide as not every process handles extremes.
i7-5820k, MSI X99A Plus, 16 GIG Ram, Noctua NH-D14, Win-7 Pro [64-bit], Reaper-64

NVC [Nebula Virtual Controllers]

User avatar
giancarlo
Founder
Founder
Posts: 11215
Joined: Mon Sep 21, 2009 10:40 pm
Location: Italy

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by giancarlo » Sun Jan 11, 2015 11:31 am

budda963 wrote:
ngarjuna wrote:
giancarlo wrote:-18dBFS is a peak value,not RMS
Pay attention if you want to address to nebula usage: most of compressors and preamps are based on peak (L infinite norm) and not RMS (L2 norm)

Gdrive itself is based on peak.
You need to normalize your audio to -18. Rms will work in most of cases, but there is at least a 6dB difference in most of cases
Why would you use peak values to gain stage? And why normalize when it's as simple as reading a meter and adjusting the input on that basis? The meter is a guide to get you into the neighborhood, the adjustment (of how hot you're going into a Nebula instance) should be based on how it sounds. Do you want to drive the input a little? Do you want to go in cool and clean? There's no magic normalization number that's going to make that adjustment for you.

Gain staging is an important subject if you want to use analog equipment same as Nebula. Take the time to learn and understand it and you'll never have another question about how to hit Nebula.
Giancarlo: I don't understand how the -18dbfs could be a peak figure when the equipment of those days didn't measure peaks. Many signals normalized to -18dbfs peak wouldn't even move the meter on a VU.

ngarjuna: I've been gain staging manually on a track by track basis but I was interested in seeing if I could speed up the workflow a little bit.

My general idea has been to get the VU meter hitting 0 or maybe +1 during the loudest parts of my signal. But that is definitely not -18rms or -18dbfs. If I'm not mistaken that figure is referred to as "-18db rms at max". It's almost like taking a peak reading from the VU meter. Of course the peaks are louder than -18dbfs but the meat of the signal is at -18. (Do i sound like a lunatic? lol. Maybe I have this whole idea wrong)

simple answer: -18dBFS is a "peak" value by definition.
Peak is also easier than RMS to the end user (you normalize your audio and everything is ok).

Peak is also the way the engine works in most of cases

We are speaking about a level you need in order to get more headroom, it is a convention.

User avatar
flipnautold
Expert
Expert
Posts: 716
Joined: Sat May 18, 2013 3:36 pm
Location: Germany, Düsseldorf
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by flipnautold » Sun Jan 11, 2015 12:44 pm

I always record as loud as possible: So i push my Hardware preamp very loud until -3db but without peaking and then put perhaps -3db headroom on top with guitars for example.

Using nebula forces me to use groups and then to lower the output volume of the recorded material to get no peaks... i always look and hear if something gets peaking in nebula on the goup.

But when i use my hardware compressor, a small tla 50, i also have to lower the volume from daw to around -12db... because otherwise the tla-50 got too much volume. Recording again into daw i record again back loud as possible.
Studio One on Intel i7 4790k.
Aquas i like are: Lime, Gold, Pink2412, Amethyst2, Honey

User avatar
Support
Expert
Expert
Posts: 3685
Joined: Mon Feb 24, 2014 11:33 am
Location: Lodi, Lombardia, Italy.
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by Support » Sun Jan 11, 2015 1:26 pm

flipnaut wrote:I always record as loud as possible: So i push my Hardware preamp very loud until -3db but without peaking and then put perhaps -3db headroom on top with guitars for example.
Depends technology and topology if each preamp, try to do it with DACS preamps, you almost will not get any nonlineal distortion. For an example Nikolay test N**e Console at +26 dBu no distortion but non lineal distortion and at +26,2 dBu distortion appears, I'm no talking about coloration due non lineal distortion, complete analog distortion at +26,2 dBu.
If you record with no regulation, we can call it full scale regulation, you get good signal to noise ratio, you use all 24 bits of your system and you also get non lineal distortion from your preamps but you must to re level your files to adapt it to a norm.
In CD production delivering PPM at -0.20 dBFS were a no regulation productions. To understand this +24 dBu are around 15.5 volts so you must to pad this signal to connect to any other device.
In studio wrong gain staging means distortion but in a touring live sound system means destroy a 1.000.000 euros system (example 18 speakers per channel flown system plus subs).
If you use a correct mics (sensibility is important) and good preamps you can record at -24 LU (20 bits are used only) and you don't need any re-level your audio for send it to broadcast station.
Best regards,
Acustica Audio help-desk service and support.
https://www.acustica-audio.com/
Acustica Audio is a trademark of Acusticaudio SRL.

To contact Acustica Audio support go to the ‘Support’ in your Acustica Audio account.
We do not provide support via social networks, public forums, Acustica Audio forums, or email accounts.

User avatar
babiuk
Vip Member
Vip Member
Posts: 384
Joined: Tue Oct 19, 2010 11:16 am

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by babiuk » Sun Jan 11, 2015 4:17 pm

I am even more confused...

I think an audio with -18 rms and -10 peaks will get more benefits through a neb preset than an audio with -10 rms and -18 peak levels

So you support guys are saying we don't have to worry about rms level coming into nebula just be careful about peak levels?
Mac Pro 5,1 / 6-core Intel Xeon / 16 GB Ram / El Capitan / Reaper / H******* Mixbuss

User avatar
Sheikyearbouti
Member
Member
Posts: 318
Joined: Wed Feb 16, 2011 6:49 pm
Location: London, UK
Contact:

Re: How do you make sure you are hitting Nebula at -18dbfs?

Post by Sheikyearbouti » Sun Jan 11, 2015 4:59 pm

flipnaut wrote:I always record as loud as possible: So i push my Hardware preamp very loud until -3db but without peaking and then put perhaps -3db headroom on top with guitars for example.
But why would you ALWAYS do this?

Here is how I see it (and I record either on large format analogue consoles - S*L, N**e - or use portable preamps/converters. Depending on which case it will change my strategy).

You start with the source and try to establish a good level out there - guitar amp, keyboard output, how loud the drummer plays, etc. (in case there are a few people in the liveroom, you also aim for good balance between them in the liveroom)

Than you add enough gain on the mic in to reach a good line level (and ideally you won't need to touch the line in gain if it is a line signal). Good level to me is a balance between what the VU meter is showing (around 0VU) and the peak one does (do not clip, try to peak at least 6dB below 0dBFS).

If the preamp/console is of the kind where you could get a sonic benefit from pushing it harder (and this largely depends on the type of music/production, the topology of the hardware, the instrument, etc.), than trim your level to tape/DAW after you've pushed the input hard (so trim via channel fader, group trim, etc.).

Further, if you boost with EQ, bring down the level a bit (I've got no rule if I would do this before or after the EQ - it will sound differently, as after will possibly let you generate more harmonic distortion within the EQ section to the trim section).

If you compress - use the gain make-up so that if you bypass the comp you do not hear any audible change in the loudness of the source.

Than you enter around 0VU or louder in your DAW. This for most of our studios is -18dBFS (peak), but some are calibrated at -14dBFS. I would also add that if the song does not comprise of many instruments, I would go hotter to DAW. For example on a classical guitar recording done with 2 mics, which will see no processing, I ask the musician for the loudest section of the song and try to get to about -6dBFS in my DAW. Also, if I want to re-enter back into the console at hotter levels for the mixing stage, I would record them like this into the DAW in the recording stage.

The point of going into your DAW at levels with enough headroom is that (and there are probably more):

1) you will not have to trim your master bus (fader)

2) you will not have troubles with clipping internally plug-ins that do not allow levels above 0dBFS - ha!

3) if you decide to go back to the analogue domain you will not have to trim (for example your TLA)

To me the point of gain staging is this - start where you want to be, finish there. So you address the source and try to maintain this level for the entire production process.

And BTW, in the digital domain at 24bits we've got about 144dB to operate [debatable and depends on some factors], in the analogue it is a lot less - say maybe about 120dB - than than you've got noise. So why would you want to have very hot digital levels?
--
Nikolay Georgiev
http://georgievsound.com/

Post Reply