Please explain dBu, dB & how much headroom is required for

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
RELATED
PRODUCTS

Post

Hi guys,

I read this statement by Bob Katz in his Mastering book: -
"...Note that a 16-bit recording fits entirely in the bottom 91 dB of the 24-bit. You would have to lower the peak level of a 24-bit recording by 48db to yield an effective 16-bit recording..."

My own research tells me that the dynamic range for 16-bit systems is 96 dB and for 24 bit digital audio it is 144 dB. So, Bob is telling the truth :roll:

1. I'm rather a bit confuse now. When I look at my Cubase channel, I see measurements in dB, correct? So what is dBu?

2. I've been recording & mixing in 24 bit. However, I only make sure there's no clipping. I did not have any conscious to make them to be around -48dB as indicated by Bob. What do you guys think? (Hmmm, is this the root of the reason why my mix is always not as loud as others?)
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

This has nothing to do with how loud your mixes are. Bit depth only measures dynamic range. What it's getting at is that the highest possible dynamic range (quietest to loudest) for 16 bit is -96db to 0dbs, whereas a 24 bit mix would be -144db to 0db. Think of on orchestral piece, where the orchestra goes from extremely quiet to extremely loud. A 16 bit recording would not accurately capture the entire range of the orchestra if the orchestra's range was to exceed -98dbs whereas a 24 bit recording would.

The reason your mixes aren't as loud as commercial mixes is probably because you haven't sucked enough of the dynamic range out of your audio. To do this is easy (extreme limiting+make up gain), but to do this without artifacts is an artform and the reason why mastering engineers are paid so well.

Post

dB's are always relative. But to what?

In the real world (acoustics) 0dB is "absolute silence" (that is, for the human ear)

In electronics 0dBv indicates an RMS level of 0.775 volts with a 600 ohm load. Since the actual current (volts) fluctuate depending on the load, it's preferred to measure unloaded. In that case it's called dBu instead of dBv to indicate it's not connected. When you read specs saying a device operates at -10dBu or +4dBu, then it's talking about nominal electricity levels it's designed for.

In digital audio however the "zero" reference point is the DA/AD clipping point which happens at full scale. Here it's called dBfs (relative to the full scale)

That should have answered question one.

2: You're doing OK. Once you start recording at a -48dB level things will become very faint and noise levels increase drastically. Something was put way out of context here.

Your mixes may be relatively quiet if you don't use any compression or limiter. These devices / plugins squash the dynamics headroom. Extreme peaks are tamed, so the average level can be boosted without clipping.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

Thanks gentlemen,

It's a comfort to know that recording in 24bit like what I've been doing is not a shocker. And, yes, I know about limiting/dynamics. Just for a moment I was rather confuse with Bob's book.

'You're doing OK. Once you start recording at a -48dB level things will become very faint and noise levels increase drastically. '

Actually, I'm not even sure how to record at -48db even if I want to. Most of the recordings are done via VSTi and sometimes via a real guitar via a Line6 POD. For guitar for example, I was told to record as loud as possible without any clipping (as indicated from my input channel). And then pull it down during mixing(if required). So far normal?

My 24-bit mix in general hops around near clipping but no red light. Please confirm this is normal? There is much talk about headroom. But, how much headroom is something I'm not sure and also even if I'm given a specific db, I'm not sure how to achieve.
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

All normal...

For measuring the headroom you can insert EA Inspector or Voxengo Span in the master bus. These measure the peak and average (RMS) levels of your mix. If the peaks don't go all the way up to -0.1dB(fs) you can gain the master fader a bit.

Then go back to Uncle Bob (Katz). His K-system is very simple, but you have to make some decisions on what loudness you want to achieve. There's K-20, K-17, K-14 and K-12. The numbers are simply the RMS levels. K-20 is good for classical orchestras, K-12 for rock and house. The main tool for reducing headroom is the mastering limiter, and compressors/limiters on tracks that cause the peaks (mainly drums and vocals)

If your RMS levels are somewhere between -12dB and -14dB then it's a pretty good level I'd say. Some commercial tracks push it up above -10dB, but that's at the cost of fidelity.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

BertKoor wrote:All normal...

For measuring the headroom you can insert EA Inspector or Voxengo Span in the master bus. These measure the peak and average (RMS) levels of your mix. If the peaks don't go all the way up to -0.1dB(fs) you can gain the master fader a bit.

Then go back to Uncle Bob (Katz). His K-system is very simple, but you have to make some decisions on what loudness you want to achieve. There's K-20, K-17, K-14 and K-12. The numbers are simply the RMS levels. K-20 is good for classical orchestras, K-12 for rock and house. The main tool for reducing headroom is the mastering limiter, and compressors/limiters on tracks that cause the peaks (mainly drums and vocals)
THANKS BertKoor. In my mix, well, since it's rock, I think it's logical to aim for K-12. I pressume this is used during mixing(not mastering).

Anyway, I dive into Voxengo Span very quickily. Read the help file. However, I'm not sure I understand the 'RMS' and 'PRMS' indicators. When I set it to 'K-12', a snippet of my mix ends up with the 'RMS' showing about -2 plus on average and the 'PRMS' on average to be about +2. What does this mean, really?

Here's what the readme file says abotu those indicators: -
"...that RMS counters show dB values: A/B, where B = A + 3 dB. PRMS (peak RMS)counter has 50 ms averaging time." That's all :shock:
BertKoor wrote: If your RMS levels are somewhere between -12dB and -14dB then it's a pretty good level I'd say. Some commercial tracks push it up above -10dB, but that's at the cost of fidelity.
This sort of RMS is post mastering, right?? For me, it is quite a challenge to reach that range in my own amateur mastering...so I doubt I ever achieve that with just post mix.
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

You have to calibrate your monitors if you are going to use Bob's k-system. Just setting Span to k-12 is rather pointless. Read this:
http://www.digido.com/portal/pmodule_id ... page_id=59

Post

jess123 wrote:since it's rock, I think it's logical to aim for K-12. I pressume this is used during mixing(not mastering).
K-12 is nice indeed, but this is the aim after mastering. During mixing it's not that important, especially if someone else is going to do the mastering. An RMS level of -20 can do fine, but usually in rock the level in mixing is already higher.

I'd try to insert a very simple mastering limiter on the master bus, e.g. the Kjaerhus Classic Mastering Limiter: http://www.kvraudio.com/get/650.html
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

As one who uses the K-System in productions, I really appreciate this "system" to percieve the dynamics of the original track. It really is all about the "mixing" now, not squashing the shit out of the tracks.

The "limiter" in the master bus is a nice feature, but if you produce your music all below 0dB till the mastering takes place, you have absolutely no dynamics loss. So it's up to you if you use one or not.


For the K-System measuring, you really have to set your meter up. Inspector Free doesn't work for this unfortunately, but if you use PSP Vintage Meter, set the integration time to 600ms and the 0dB Point to -12dB or -14dB, you're in the corresponding system. If you look at the needle now, everything that hovers around 0dB if K-12 or K-14. You're allowed however to use the so called "amber zone" of 2-4dB for mezzoforte parts. How long is your thing, according to the chats I had with Mr. Katz. But if you go in the reds (+4dB and over) you're not in the system anymore.


Read yourself through the articles. They're hard to get at first, but if you setup your speakers and your metering tools, you'll realise that most pop productions simply sound crap now. And btw... most pop stuff is at -8dB RMS to -6dB RMS. That'd be K-12 with fully used Red-Zone all the way from beginning to the end of a track.


The last production I made was in K-12 (Chrono Symphonic). I had RMS levels from -20dB to -8dB (K-12 with amberzone). Peaks were barely cut off and sounded totally awesome after watching a very dynamic video DVD with the same volume settings while listening to the production I engineered.


It's up to you what you prefer. Loud and squashed, or loud and dynamic, even though people will tell you "turn up the gain!". I switched, and I didn't regret it.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Thanks gentlemen,

Seems like a lot of reading. Man, this Bob Katz stuff is really technical. As long as I know, I'm below clipping, I'm all good and that's the comfort I get from these postings. But sometimes, I wish I know how much RMS I'm in in almost realtime during mixing, that's all. I just need to dissect what this RMS and PRMS from Voxengo SPAN really means. Maybe I hop on to the Voxengo forum on that.
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

Well just put the PSP Vintage Meter in the master bus, set it up to int. time of 600ms, then 0-point to -14dBFS and you know where you are while producing.

The meter needs that integration time to measure the correct loudness what "we" feel, too. Everything else is too fast and is off in terms of measuring. This is why I said Inspector Free isn't suitable for that (you can't setup the response time of the RMS meter).

HOWEVER...as long as you use the limiter in your sequencer only for cutting off spikes, or not using a limiter at all and keep below the 0dB peak level, you should always be below -14dB to -12dB RMS of the K-System.

THIS is the advantage of this system, you can't get too hot right while producing. That's what mastering is for, you set the final "loudness level" in this step, not while producing. And Mastering Engineers will love you for that. At least those who still know that a song can be dynamic, and not squashed to death.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Compyfox wrote:Well just put the PSP Vintage Meter in the master bus, set it up to int. time of 600ms, then 0-point to -14dBFS and you know where you are while producing.

The meter needs that integration time to measure the correct loudness what "we" feel, too. Everything else is too fast and is off in terms of measuring. This is why I said Inspector Free isn't suitable for that (you can't setup the response time of the RMS meter).

HOWEVER...as long as you use the limiter in your sequencer only for cutting off spikes, or not using a limiter at all and keep below the 0dB peak level, you should always be below -14dB to -12dB RMS of the K-System.

THIS is the advantage of this system, you can't get too hot right while producing. That's what mastering is for, you set the final "loudness level" in this step, not while producing. And Mastering Engineers will love you for that. At least those who still know that a song can be dynamic, and not squashed to death.
Heh CompyFox, I went to your nice website. You've got good studio env and nice reference.

Vintage Meter settings: '....then 0-point to -14dBFS'
Do you mean by turning the knob on 'OVU refer. level' to '-14.0dBFS'?

I re-read the Bob Katz stuff....and I think I know a little bit more now. I should be in the K-14 level....that's why you recommended me to tweak it to '-14.0dBFS' ?
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

I haven't used the Vintage Meter in a while since I'm using InspectorXL (before it was raped by RND), but the setup is fairly simple. Just let me take a look at the screens from PSP... ah yeah.

Set it up this way:
VU integr. time: 600ms
0VU refer. level: -20dBFS for K-20, -14dBFS for K-14 and -12dBFS for K-12

The rest can be left to standard. Now how to read it:
If the switch on the main pannel is on VU, and the dials are on +6dB (you can switch them by clicking on the dial fields from +3dB to +6dB), then everything that hovers around or "below" the 0dB point is in the proper K-System. If you go 4dB louder, that's for mezzoforte parts and spikes, everything over +4dB to +6dB is the red zone and you shouldn't even reach that.


Keep in mind though... if you decide to use up the full amber zone of K-12 (in mezzoforte parts), you're still in K-12, but the RMS level is -8dB. Just do the math. As long as you stick to the rules (overall level at 0dB, mezzoforte parts not louder than +3dB, which is a reccomendation by myself), you're fine and good to go.

The K-14 recommendation is from Katz himself, too. He suggests that K-20 is for very dynamic stuff (orchestra, etc), while K-14 is for most pop productions and K-12 is for Broadcast. There're still no set rules about the usage of the amber zone, so everyone is on their own, but if you stick to those little rules, you'll realise your sounds is way better than before.


BTW: Thanks for visiting my page and giving feedback. Unfortunately I can't give you a better example of the usage of the K-System with my productions, as they're fairly old (though I'm sure I fixed them for K-12, which is my upper limit). However if you want, go to my torrent section and listen to the free available Chrono Symphonic soundtrack project - this was done completely in K-12, and that shines through.


Hope I could help.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post

Compyfox wrote:I haven't used the Vintage Meter in a while since I'm using InspectorXL (before it was raped by RND), but the setup is fairly simple. Just let me take a look at the screens from PSP... ah yeah.

Set it up this way:
VU integr. time: 600ms
0VU refer. level: -20dBFS for K-20, -14dBFS for K-14 and -12dBFS for K-12

The rest can be left to standard. Now how to read it:
If the switch on the main pannel is on VU, and the dials are on +6dB (you can switch them by clicking on the dial fields from +3dB to +6dB), then everything that hovers around or "below" the 0dB point is in the proper K-System. If you go 4dB louder, that's for mezzoforte parts and spikes, everything over +4dB to +6dB is the red zone and you shouldn't even reach that.
All this is really for mastering stage consideration, huh? If I do use this for mixing, I think I'll stick to K-14. I'll make sure my mixes are within 'normal', thus the master engineer (If I can find one and also afford one) can do the job right.
Compyfox wrote:
Keep in mind though... if you decide to use up the full amber zone of K-12 (in mezzoforte parts), you're still in K-12, but the RMS level is -8dB. Just do the math. As long as you stick to the rules (overall level at 0dB, mezzoforte parts not louder than +3dB, which is a reccomendation by myself), you're fine and good to go.

The K-14 recommendation is from Katz himself, too. He suggests that K-20 is for very dynamic stuff (orchestra, etc), while K-14 is for most pop productions and K-12 is for Broadcast. There're still no set rules about the usage of the amber zone, so everyone is on their own, but if you stick to those little rules, you'll realise your sounds is way better than before.


BTW: Thanks for visiting my page and giving feedback. Unfortunately I can't give you a better example of the usage of the K-System with my productions, as they're fairly old (though I'm sure I fixed them for K-12, which is my upper limit). However if you want, go to my torrent section and listen to the free available Chrono Symphonic soundtrack project - this was done completely in K-12, and that shines through.
My amateurly mastered material are far from that kinda RMS. Was that using quite a lot of the amber zone :) during mastering stage?
Compyfox wrote: Hope I could help.
You did more than help, thanks.
Warm regards,
- Jess
Win XP SP2
Cubase SX 3.1.1

Post

jess123 wrote:All this is really for mastering stage consideration, huh? If I do use this for mixing, I think I'll stick to K-14. I'll make sure my mixes are within 'normal', thus the master engineer (If I can find one and also afford one) can do the job right.
No... not at all. The K-System is a "guide", the meters are for helping putting it into the final form. If you produce your stuff and it's not clipping at all while producing (meaning doesn't reach 0dB peak), then the rest is totally for the mastering stage.

K-20 barely hits the 0dB range anyway, so I'd go with that and put it into the final form (K-14 or K-12) later. It's all about the mix.
jess123 wrote:My amateurly mastered material are far from that kinda RMS. Was that using quite a lot of the amber zone :) during mastering stage?
If you mean "Chono Symphonic", yes. There's everything from -20dB RMS to -8dB RMS (full amber zone) if you go by K-12. Overall loudness is K-12, which means it hovers around 0dB VU on the Vintage Meter from PSP.
[ Mix Challenge ] | [ Studio Page / Twitter ] | [ KVRmarks (see: metering tools) ]

Post Reply

Return to “Production Techniques”