What is difference between Volume and using the Loudness check box in Final Cut Pro X?

I'm confused about the difference between using the volume control on audio– either raising or lowering by DB, or clicking the "loudness" setting on the Audio Enhancements pane. Can anyone please illuminate this for me?

I always record audio with lower levels to keep it from overloading or clipping and then manually raise the levels afterwards and wonder what is the difference between both approaches?

Thanks in advance.

Loudness is not about volume but about adding dynamics (treble, bass) to the audio. Volume is about getting the audio louder without adjusting the dynamic range.

Thanks! That helps a lot. I wish FCP X had a "normalize" audio level feature like the old FCP 6. I guess manually raising the volume is the best approach for me.

In FCPX you find an audio limiter. Search in the effects for 'limiter'.

Hi ZZ
Just to add to the discussion and to address an implied question…
Here is what Apple says about the "Loudness" setting… “Change loudness settings: Drag the Amount and Uniformity percentage sliders in the Loudness section. The Amount slider increases or decreases the overall loudness (compression) of the clip, and the Uniformity slider increases or decreases the dynamic range affected.”

Excerpt From: Apple Inc. “Final Cut Pro X User Guide.” iBooks. itun.es/us/rrsm6.l

I think the implied question and needed discussion is about recording technique.
If you record at low levels and then boost in post, you risk increasing the noise in your audio. All audio equipment has a Signal to Noise Ration (SNR) and an associated "Noise Floor". These are in essence fixed.
Signal = Good; Noise = Bad.
Creating the best SNR starts with equipment selection especially mics, mic placement (closer to source the better) and recording as "hot" as you can without clipping. Clipping in digital recording, VERY BAD. :evil:
In the good old days of analog recording, we used to want to saturate into the red for studio tape recording, but any true clipping in a digital recording means a big Pop and lost wave forms. Many digital meters have a little actual headroom but you don't want to depend on that.
Being able to live monitor the recording is always a good idea if possible.
So if you record too softly and then boost in post, your Noise is also boosted. If you record as hot as you can without clipping you will have a better initial SNR and the noise will not be noticed.

Here is Apple's take on SNR… Signal-to-Noise Ratio
Every electrical system produces a certain amount of low-level electrical activity called noise. The noise floor is the level of noise inherent in a system. It is nearly impossible to eliminate all the noise in an electrical system, but you don’t have to worry about the noise if you record your signals significantly higher than the noise floor. If you record audio too low, you raise the volume to hear it, which also raises the volume of the noise floor, causing a noticeable hiss.

The more a signal is amplified, the louder the noise becomes. Therefore, it is important to record most audio around the nominal (ideal) level of the device, which is labeled 0 dB on an analog audio meter.

The signal-to-noise ratio, typically measured in dB, is the difference between the nominal recording level and the noise floor of the device. For example, the signal-to-noise ratio of an analog tape deck may be 60 dB, which means the inherent noise in the system is 60 dB lower than the ideal recording level.

Here's is Rode's… www.rode.com/blog/all/what-is-signal-to-noise-ratio

Thousands of more articles out there, but you get the idea.

Anyway just my 2 cents worth and my interpretation of the implied question.

Hope this Helps, Greg

Basically the "Loudness" feature, as described above, is a simplified Compressor with gain added.

If you've pushed your audio to +12dB and it's still too low there is an Audio Filter Gain you can add to the clip.

Of course all this pushes up the noise floor so you'd probably want to follow that with a good noise reduction filter.

Normalize, as I recall, would simply bring up the audio based on the peak. The problem with normalize is that if there were momentary high peaks it wouldn't help the underlying low audio. Just my opinion but it was pretty useless. At the very least if I want things louder I'm either going to need to Limit the peaks to prevent distortion or Compress above a threshold.

Once you understand the concepts, using a Compressor filter will give you better control than the built in Loudness function. But in many cases the simple Loudness function may well suffice.

Apple's own description of their Loudness function seems awkward to me. Basically it's doing a Compression ratio, Gain, Threshold for it to kick in. I'd think. Compression itself actually decreases the dynamic range (makes quiet and loud sounds closer together) and also decreases the the loudness. Then you need to add Gain to bring that back up. The Threshold determines at what sound level things start getting compressed.

I'll try to put it another way:
Threshold - the sound level where the squishing begins. Let's say everything above -18dB
Compression - how much things are getting squished together. It's usually a ratio such as 3:1.
Gain - now that the dynamic range is squished so that there's not much difference between quiet and loud, you're bringing that back up (making it louder).

So FCPX Loudness is combining those functions while hiding the front end settings.
My guess is Uniformity is Compression Ratio and Amount is Gain or maybe Threshold and Gain.

Normalize dynamically adjusts audio so both peaks and lows are at their loudest without clipping. Very popular with music recording engineers. All pop songs are normalized when completed to get the most volume out of the file possible. I miss it in FCPX, is very helpful in many situations.

Compressor:
help.apple.com/finalcutpro-logiceffects/...0.0.6/#/lgex931a04b0

Gain:
help.apple.com/finalcutpro-logiceffects/...0.0.6/#/lgex931a00d2

Limiter:
help.apple.com/finalcutpro-logiceffects/...0.0.6/#/lgex931a007e

Normalization in FCP legacy simply looked at the peak and brought up the entire gain until the peak hit the targeted level. For me at least, I didn't find that useful because the momentary peaks of one clip could be much higher than another so relative loudness between clips would still be disparate.

From the FCP 7 User Manual
How Normalization Gain Works in Final Cut Pro
Audio normalization works by scanning audio for the peak (loudest) sample level and then applying a Gain filter that brings the peak level to the level you request. By default this value is 0 dBFS, the highest level possible before clipping occurs. The Gain filter raises the overall audio level.

Normalize would be useful perhaps after applying Compression so that the relative loudness between clips were similar. Just my opinion from the audio work that I've done.

I do think people mistakenly believed that Normalize would bring clips up to a similar relative loudness. If peaks were similarly related to the clip's relative loudness that would be the case but often they are not. All the peaks could be at 0dB but some clips with momentary hight peaks would be quieter than other clips in which the peaks are much closer to the relative loudness.

It might be useful if Apple added a Normalize filter but I'm personally I'm glad it's not a button or built in to the Audio Enhancements. Loudness is more useful IMHO.

EDIT
From the FCPX Logic FX reference
Adaptive Limiter
“Adaptive Limiter is typically used on the final mix, where it can be placed after a compressor, such as Multipressor, and before a final gain control, resulting in a mix of maximum loudness. Adaptive Limiter can produce a louder-sounding mix than can be achieved by normalizing the signal”

help.apple.com/finalcutpro-logiceffects/...Revision-DYN-1007877

So this may reflect Apple's position on this in that it was something applied after compression and that the Adaptive Limiter is more useful in that regard.

Well this thread is now heading off into some technical areas. Perhaps more than the OP expected.
I think there is some general confusion about the definition of "Normalization". In fact there may not be a single definition but at least 2 or 3 or more. Peak normalization, average and perhaps a subset RMS. Not to mention Broadcast versions.
What Craig said about Peak normalization I believe is correct. If you apply a "Peak" normalizer on a set of clips or tracks in a DAW, the program will analyze the tracks for the highest (peak) volume in each track and then simply apply a volume change to each track to bring the peak volume to the target level. This form of normalization does not change the dynamics of any individual track. It is exactly the same as moving the fader on a DAW.
With a "Average" loudness normalizer, the program attempts to find the overall middle of the audio volume. RMS is another way to do this. Some plugin's simply then push each track's middle up or down and some actually apply some compression or a limiter to avoid clipping. If using the simple method, dynamics are not affected and may result in clipping, but obviously, if the plugin applies compression or a limiter then the dynamics are affected.
Personally, for a final mix, I like to EQ>Compress>Limit all while watching the faders, then I might, but normally don't apply a normalizer for Audio CDs just so the tracks play the same. Prefer my ears and eyes.
Finally, there is the issue of using a broadcast version of a normalizer, something like an EBU-R 128 for compliance with Broadcast Standards. Whole 'nother' can of worms. These can be applied in final or via the broadcaster themselves. Both U.S and European standards need apply! :whistle:

Finally, there is no such thing as only one way of doing something.

Hope this Helps, Greg

When recording music, and doing our broadcast audio work, normalization is almost always used very last, after gains, limiters, exciters, compressors. Most all recording studios I've been in do this. Just saying. It is very useful.

normalization is almost always used very last, after gains, limiters, exciters, compressors.

Absolutely useful after compression or otherwise at the end of the chain. But newbies often didn't know or understand that workflow and they thought it was a thing to "even out" loudness by itself as it seemed to me by the original poster's wish for "normalize" after asking about loudness. That's why I'm glad it's not there as an Audio Enhancement.

Apple's recommending Adaptive Limiter as a better alternative (in their opinion).

Unfortunately Normalising isn't a terrible get technique for getting sounds louder.

First thin is you really need to record levels loud enough in the first place without clipping it. Unfortunately many onboard camera preamps tend to get noisy as you push them, so that becomes problem. So in an ideal world you would be looking for a more sensitive mic, a third party recorder/preamp, and getting the mic as closely as possible to the source without it appearing in shot.

Thats at the recording end and is a little bit of an idealistic look, you just do what you can.

The second problem is that Normalisation will just rain levels until the highest bit of peak information hits 0dBFS or thereabouts. On the face of it this may seem a good thing, but all sounds aren't made equal, and this could lead to a great deal of audio imbalances which you will only have to fight with later.

Take for example someone knocking on a door and a running car engine...

Knocks are short, percussive sounds, they are there one sec, gone the next. They have a lot of peak energy and you could normalise it to make it louder, and it definitely would help, but it could still sound thin, artificial and lack presence. But what can you do about it?? Well you can't turn it up because you're already at 0dB and that just means nasty digital clipping. Some make the mistake of turning everything else down, but their feature ends up much more quite than it should be.

Now when you look at the car, thats constant sustained sound, with rumble lots of bass energy, quite a bit of tremble energy. If you Normalise that you could risk blowing the s&#t out of your speakers!! All the other audio would struggle to compete.

So with Normalisation you potentially end up with two issues, you have maximised sounds that still sound quiet, or drowns everything else out.

You've probably figured by now, that this is not how the pros solve the problem...

Unfortunately the solution requires a bit of audio education and quite a bit of practice, but why you really need to look at is Compression & Limiting, as some have already touched on.

There are other tools that make the process a little easier. Waves have a couple that come to mind, Vocal Rider and MV2. MV2 especially is ridiculously easy to use and allows to to independently bring up quite parts, while bringing down the loudest. As always, be careful with what you do as you can so easily overcook it and have all your audio scream at you.

Getting back to compression and limiting, you really want a limiter governing what your highest parts, or peaks, are doing, and a compressor to govern the overall loudness and energy level or the sound. A limiter is really just a specialised compressor, so you could just use any compressor with a fat enough, sub 1ms attack and a ratio of 20:1 or more.

However FCPX does have Compressor and Limiter which should do a pretty good job, if you don't go the simple way of getting something like the MV2 to do the job.

If you need to know more about how to set them up, just post a reply here. I just think i've already rambled on long enough.