What is the Ideal dB Level?

[T Prod.]

What laurend wrote doesn't make sense. He wrote,
The poll asks "What Is The Ideal Decibel Level For Audio Mixdown". Despite the bit depth 0dBFS is not an option for a mixdown. I think what is trying to be expressed here is that due to the limitations of 16 bit audio, we should push the meters as much as possible, as our available headroom is much less than with 24 bit audio.

yeah I didn't pay enough attention to his post, now I see that's rubbish, but the bit depth thing is not.

 

[deanlofts]

I don't even understand what the difference is between dbFS and dbSPL. I haven't looked it up though.I like to have my mix anywhere between -5.0 and -3.0 on the Cubase 5 peak meter. I then will hit it with a limiter and raise the volume to an acceptable level. I assume I am working in dbFS?

 

[Waltz Mastering]

FS = Full scale
SPL = Sound pressure level
VU = volume units

 

[nyteryda]

i keep my peak at -12. it just sounds more crisp when u start to build it up. i have experimented with different levels and -12 works best for me.
i use to peak at -6 but i found that it loses dynamics when you compress.

 

[PSYTRACKED]

Does anybody use the K-System for monitoring metering and levelling? I use K-20 for most of my mixes, K-14 for hip hop, pop and rock records.

I use a lower reference. I use 80dB spl instead of 86dB spl.

---------- Post added at 07:23 PM ---------- Previous post was at 07:08 PM ----------

More and more I am realizing the reality behind mastering. And though the numbers are correct, there is still something missing. Mixing and mastering cannot be numbers alone, although I thought this way for many years. I am beginning to use my ears more and more. And although I am a convinced K-System user, I will admit that I have pushed the K-System to areas that I doubt Bob would approve, +9 in K-14 for example. But it made good musical/business sense IMHO.

Very good points. It is up to the artist / producer to decide what they want for their production. And the numbers can be misleading. I use the TT DR meter. Both realtime and offline and it revealed some things to me that has un-done alot of my original beliefs in the K-system.

I loaded up one of my favourite folk songs and it only has a dynamic range of 8dB. For folk music this is considered to crushed. But it isn't. The songs sounds great. Meters can't tell you if it sounds good or bad. Mixing by numbers? If only it was that simple.

 

[leopoldolopes]

To me there is no fixed number... just the right headroom... which is always with no exception between -3dbfs and -10dbs... and as a additional rule I always try to go between this two, right about -6!
Cheers all!

---------- Post added at 12:58 PM ---------- Previous post was at 12:55 PM ----------

i keep my peak at -12. it just sounds more crisp when u start to build it up. i have experimented with different levels and -12 works best for me.
i use to peak at -6 but i found that it loses dynamics when you compress.

I think that you should try harder on your compressing settings... since I found that is a myth in your head, as I said maybe you are missing something when putting your fingers on compressors!

 

[HakimCallier]

To me there is no fixed number...

I think the post was really geared toward FP members and lurkers new to mixing.

With that said, I do agree with you. Of course we have our standard limits for good audio. But as a professional, our job is to find the best sounding level within those limits and perhaps even pushing those limits when necessary.

 

[leopoldolopes]

our job is to find the best sounding level within those limits and perhaps even pushing those limits when necessary.

I agree that you must try to find the best always... but limits on the mixing process are very clear... to me and of course to you to! As I said... I try to find what better suits the project which I'm working on... but always keeping in mind that are some rules in the "mixing" process that you must keep, always protecting the mastering job and who's going to preform it! I won't find any loss of dynamics or whatsoever if you bounce it/export a song or a track between -3 or -10... just following your working flow and trying to get it right never surpassing those limiting rules, which will ruin the work you've done!

 

[Mr5ambo29]

Here's a guide that explains a lot about DBFS, SPL, etc.. http://www.jimprice.com/prosound/db.htm#dBFS

 

[Chi kitory]

hmm i dont know enough about this, but reading the answers from people gives me a good idea. Good Poll mate. I usually just keep at 0, cause thats standard.

 

[Musikal1]

I try to keep it around -3 havent went lower.

 

[jlgrimes11]

I would like to know if anyone here knows the ideal signal level for your stereo mixdown? I am referring to dBFS (PPM). I find so many producers don't know what level their tracks should be peaking to give the mastering engineer the optimal headroom to work, so I thought it might be a good idea to discuss it here.

Doesn't matter much actually (unless you are clipping) in the digital 32 or 64 bit mixing engine world.


Probably anywhere between -12 to -6 db below 0 would be ideal.


Keep in mind though that for every 6 db below 0 is about 1 bit loss of resolution, but don't get too concerned because you would have to be recording extremely low on purpose to get back down to 16 bit audio.


As a best practice don't clip leave at least 6 db of headroom (but I wouldn't even fret too much if you get some peaks above that).

Do whatever sound good.

 

[SilentVoice]

-6dBFS is the industry standard for common head-room...anything higher than that is cuttin' it pretty close...best bet is to have it -6dBFS.
Anybody said other wise has no idea what they are talking about.

 

[wlouch]

Just want to say that recording at that level is not going to make your songs punchy. Bits are there to be used for headroom purposes as well. I used to think you had to use all the bits, otherwise you lose resolution, bits are used for dyanmic range and headroom on top of that range.

I would much rather recording everything between -18 and -12dBFS (in the digital domain) Recording much quieter than this in digital isn't a problem either. Signal to noise is much less of an issue.

When mixing, I have better control over dynamics, everything isn't just a slightly different volume of loud like it would be recording and mixing at anywhere near 0dB. I would ideally want my completely finished piece to peak at -6dBFS and no louder, even after mastering.

Recording in digital, reducing the levels you mix and record at will reduce the elements of digital that people frown upon, i.e. harshness, distortion etc, this doesnt come in to the picture if you keep volumes down when mixing and recording stages.

For example, having a kick peaking at -9dB as the loudest element in your track, IF it is a drum beat powered track this would give much more punch to your records mixing everything in around this. -9dB is just an example, not a rule.

You would have quieter instruments around it, but it would sit better in a mix like this. and then that drum track can really punch through the composition.

If it sounds good quiet, it will only sound better louder. People can turn your mixes up with their volumes, dont do it for them as this is normally done with compression, THE OPPOSITE of adding punch, you will only get a muddier, less defined sound from doing so, with far less dynamic range.



Mix for balance, clarity and punch.


rant over.

 

[laurend]

I would much rather recording everything between -18 and -12dBFS (in the digital domain) Recording much quieter than this in digital isn't a problem either. Signal to noise is much less of an issue.

I totally agree on that point for 24 bit. 16 bit files may require a slightly higher limit to preserve definition. let's say from -6 to -3 dBFS.

When mixing, I have better control over dynamics, everything isn't just a slightly different volume of loud like it would be recording and mixing at anywhere near 0dB. I would ideally want my completely finished piece to peak at -6dBFS and no louder, even after mastering.

-6dBFS is very conservative for a mastered material. This level is perfect to avoid any clipping during heavy mp3 compression, but for CD production, -0.3 dB is OK since this ususally prevents any inter sample clipping. Loud can be good when mastered.

 

[wlouch]

The only reason I use -6dB is because my signal chain includes analogue and digital stuff, so when I record I have to bare in mind the optimum workings for each device.

If you know your equipment well, then yeah, 0.3dBFS isn't a problem, as long as it is all mixed for clarity first, mastering really can just make you go "ahhh, its done" sounded great before hand, but sounds perfect afterwards.

I am glad someone agrees. I find a lot of people (even professionals) are confused with the different dB mediums for each domain, whether it analogue or digital, there is a mist of confusion surrounding this subject, and since digital has come out, its being misused, purely because more people can afford to buy it, people with lesser knowledge.

 

[laurend]

Here's a website that should be usefull for everebody involved in audio and dB discutions.

 

[wlouch]

Nice website, i do feel having education in music and engineering will be needed to understand that fully.

...or just very dedicated people.

 

[T Prod.]

The lower level you record, the less bits you have for the same signal. It's a fact. BUT. If you work at least at 16bit, it won't be a problem if you don't go under -30dB.

It's not really about resolution and bits, it's about noise.

Don't clip, and make sure you have a good noise-to-signal ratio. You have to know where to get that level from.

for example: if you record an mpc or a synth through a mixer or a channel strip, you have to get the level from the synth/sampler itself, not from the preamp on the mixer. that'd give you much more noise.

anyway, there aren't really any relation between punchy mixes and recording levels.

 

[laurend]

Every bit worths 6 dB in the signal to noise ratio. - 30 dBFS means you're losing 5 bit on 16 in the signal description. Then the theorical signal to noise ratio is 11 (remaining bit) x 6 = 66 dB. which is poor. Bit resolution and noise are related.

If you're talking about -30 dBu (RMS), that's correct because peaks (dBFS) will be 3 or 4 bit higher.