Right-click each audio event and select Switch | Normalise. Unfortunately you cannot multi-select events - you have to do each one individually.

If the nature of the clips is very different you may need to do some different compression on each to even things out more.

geoff

Geoff, you Can multi-select the audio events and normalise them all at once.

Though, without any compression the highest peak will determine the rest of the audio track.
Adding compression can balance out the highest peak and lift the lower peaks as well.


VMP

Normalize in Vegas' Event context menu is peak normalization, and is not at all what you want. Has no correlation to relative loudness (loudness is NOT volume).

An OK approach is to normalize each media to say, -12dB RMS with Limiting in Sound Forge (Process->Normalize). This approach does change the media dynamics to better match, but still requires additional tweaking to correct subjective differences.

If you like the individual media dynamics, but they just don't work together on the timeline, the better solution is true loudness normalization by monitoring with Toneboosters TB_EBU plugin pack. This is the broadcast industry approach, which does not alter media dynamics, but matches clip Loudness (as defined by BT1770). This is very effective, and usually needs no additional volume tweaks.

Loudness target levels are -24 LKFS for US broadcast (ATSC A/85), and the de facto internet target is 6dB louder (-18 LKFS), using Apple's benchmarks.

Now that you have the tools, experiment!

My experience with audio tracks from various sources is like yours. Levels are all over the place :(
What I do is simply adjust the individual track volumes so that they all match level-wise. I'll apply a volume envelope if I need to in specific spots but that's rare.

Mike

Mike, that's exactly what BT1770 loudness monitoring does, but scientifically, tied to human hearing response.
Volume levels and Loudness are not directly related in any meaningful way.

For music only, it's really pretty simple to get close to where you want to be. In Sound Forge, normalize each track to -27db. Here's an example of tracks from a current project. You can see they're really all over the place.



For each track, I've opened a copy in Sound Forge, normalized to -27 db, and then saved which adds the file as a take. The results look like this.



I've also added the LUFS values of each track using the Loudness Meters within V13 which track very closely the values from the TBU Loudness plugin mentioned by Musicvid10. To compute those values, you can listen to the entire track, or simply render render each track as a Wav file making sure the "Save loudness......" is ticked. Then open the loudness log, go to the end and read the LUFS value. It sure is a lot faster than having to "play" each track. As you can see, the loudness of each track is pretty close to the desired value of 24. You could then adjust the volume of each track to get it to be exactly 24, but I suspect the difference would not be perceptible. The final step would be to add a track compressor and then tweak with a volume envelope. The first step, however, and easiest--just normalize to -27db.

wwaag

I just discovered something even quicker to generate LUFS values. Under Tools, select Generate Loudness Log. It creates the log file without rendering. If you have a loop region, it will generate the log file only for that region. Pretty slick.

Have you considered creating a new audio bus, assigning the outputs from all your source tracks to the new bus and adding your fX to the bus?

Have you considered creating a new audio bus, assigning the outputs from all your source tracks to the new bus and adding your fX to the bus?

I presume this is in response to the OP's original question and not my example above. If you normalize each selection (or even if you don"t), it's even easier to put all your music on one track, which is what I do. Then a single Track Compressor Fx and a single velocity envelope.

wwaag

Have you considered just listening? ..and adjusting the levels from one segment to the next. Then adjust the master (or sub-master) for overall level. Some gentle compression and peak limiting on the master buss would keep levels in check further.
Sorry, it's not instant gratification, but there is a free VST plug-in that 'kind of' is though. Nothing to set, render.. and your good to go. I don't recall the name at the moment or if it works w/ Vegas. It does with SF. Levelizer or something like that.

@ wwaag

Yes, to the former.

@rraud

"[I]Have you considered just listening?[/I]"

Best advice I've heard all week.

"Some gentle compression..."

Would this be considered "gentle compression"?

wwaag,

Thanks for response to my question, all comments have been enlightening, I'm learning things about VP I didn't know.

I've created loudness logs as wel as viewed the VP Loudness meters. I see Momentary, Short, Integrated, and Loudness range values. Can I assume that your screen shot showing the LUFS values that you superimposed (24.18, 23.38, 23.81, etc.) are the INTEGRATED values. I think so, but want to make sure.

Thanks,

A few folks have asked, "Have you tried listening ...", if that question is directed to me (the OP), that is exactly what I have been doing for a decade of using VP, but it is time consuming, error prone, and subjective. I end of tweaking and tweaking and tweaking.

I asked my question thinking there has got to be a better way, and I believe there is. If nothing else, I learned about LUFS values. I now have an objective number to use rather than my own subjective judgement.

This is all good stuff and I thank everyone for their responses.

are the INTEGRATED values.

Yes. I copied them from directly from the Log files that were created.

"Have you considered just listening?"

Obviously, you have to listen, at some point, but I I have to agree with the OP that that's the approach we've been doing for years and probably wasting countless hours. The whole point of normalizing and compressing and using the loudness numbers is to simply get you "in the ball park". At least you know that you're numbers are "legal" and many times they will be "good enough". If not, then you can tweak and fine tune by listening and adjusting your volume envelopes until you're satisfied or simply sick of it, whichever comes first.

wwaag

My Toneboosters suggestion is for folks with older Sony software.
-27 LUFS (-28 LKFS) is awfully quiet, though. Yes, the suggestions are Integrated values, but they are not dB scale, just increments.
Gotta watch those True Peak values, as well.

Oh yes, one CAN multi-select and normalise ! Tried it the other day and it wouldn't ??!!??

One drawback of normalising everything , then fading things down again, is that (especially if a lot of tracks) is that quantisation noise will be brought up and add to background fud. Not such a problem with 24 bits, unless an extreme amount of gain is applied for the normalisation process.

That is unless V is very clever and works out the difference between the start level and the final destination level, and just does one operation - instead of doing the normalise thing followed by the level-fader thing. Anybody know for sure ?

geoff

I’ve spent the last hour or more experimenting with loudness in Vegas and Sound Forge. I am new to all this, but I’m not getting what others have suggested.

• Generating a loudness log of the original soundtrack, I get -7 LUFS
• Generating a loudness log after making a copy in Sound Forge, using Process/Normalize with Average RMS level set to -27dB, I get -12 LUFS
• Generating a loudness log after making a copy in Sound Forge, using Process/Normalize with Peak level set to -27dB, I get -39 LUFS

In summary, I get “Integrated LUFS” values of -7, -12, and -39 “all negative values, is the desired target really +24 (positive)? Even if it is -24 (minus), I’m nowhere near that either. Got any suggestions?

Um., surely "+24" is not possible. It starts at zero and goes negative from there. Yes - you are right - it should be "-24" ;-)

Sorry , not much help in trying to attain that magical figure.

geoff

Rich,
Sorry about the +- confusion. I failed to add the minus signs in front of the LUFS values. The -39 LUFS when normalized at -27db being the peak value seems reasonable. The -12 LUFS for average does not. Suggest you try again from the original soundtrack that you got a -7LUFS from the log. Then make a copy and open in sound forge (if you do this frequently like I do, you may want to make a keyboard shortcut to speed things along). Before applying the -27db normalization, first do a scan. The RMS value should be around -10db. Then apply the -27db Average RMS normalization. Then scan again, just to ensure that average RMS is -27. Save, return to Vegas and do the loudness scan again making sure that you have created a loop region around that event only and that all other tracks are muted. It should be around -24.

wwaag

One other thing, when you copy to Sound Forge, make sure that your audio effects (both track and event) are disabled. Otherwise, they will be applied before normalization in SF.

IT WORKS!

Thanks wwaag. I normalized a few audio clips to the industry standard -24 LUFS using your procedures. The resulting clips are much lower in loudness, but at least now they are all of equally low loudness.

Now with my -24 LUFS audio clips, Vegas Loudness Meters (EBU R128) , the True peak meters peak at -15 dB, which is pretty low. Is that normal? I know you don't want to clip at 0.0 dB, but -15 dB is really low.

Glad it's working for you. One thing to note. Apparently, EBU R128 is the European standard whereas ATSC A/85 is the US standard. Unless you're submitting your work for broadcast, I suspect it really doesn't make a difference which one you use, just so long as it's the same. I may be mistaken, but even -15db exceeds certain broadcast requirements. Some years ago, I got turned on to using these Loudness meters by Musicvid10 for the very same reason of your initial question, so perhaps he could chime in with a definitive answer.

wwaag

wwaas,

Thanks once again for the American vs. European standard info.

I hate asking because I should be able to find it myself, but where is the "setting" that super imposes the TEXT on the Vegas event that you show on your Vegas screen shot above. I'm referring to the name of the clip (i.e., Take 2, etc.)

I found this a few years ago, but turned it off and can't find it now.

Rich

Ctrl-Shift-I will toggle it on and off.