Hi all – I expect plenty of newbie naivete is showing here, but can you help me make my exported WAV file sound louder? The reason I believe there is unrealized opportunity is because I have a friend doing a ‘remix’ of my project using some of the same raw tracks, and his exported WAV file has a bigger sound than mine. I think he is using Garargeband and/or Audacity.
I’m running Ubuntu Studio 19.10, Ardour v 5.12. I’m using Calf LV2 compressor as a plug in on the master channel. I’ve set ratio to infinite to get a feel for the range of possibility. Even so I can’t seem to turn up the makeup gain very much before the master channel starts creeping over 0db. In any case, I then do Session/Export to Audio, and choose the standard CD/Red Book format (normalize peak; WAV, 16 bit 44.1KHz).
Any advice? Sonic accuracy is not a high priority.
Any comments/directions/suggestions would be much appreciated.
I think a mix of compression and limiting will do the trick. But, before that, double check what your RMS levels or LUFS levels are before assuming you need any plugins to help. Depending on where the song will end up, it will get turned down if it exceeds a certain LUFS value. Aiming for -16 LUFS is a safe bet these days.
That said, by “bigger” sound you might be hearing the effects of compression, limiting and EQing (and maybe even saturation). There are plenty of tutorials on the web for how to use each individual type of plugin and even for combining them.
I would not go down the route of sonic accuracy not being a high priority unless you are deliberately adding lo-fi effects for that purpose (e.g. bit-crushers). You can maintain good/excellent quality with the usual compression/limiting if you apply good gain-staging (also plenty of tutorials on this) coupled with sensible plugin values.
There’s no magic recipe for songs as it all depends what your source material is and I can’t stress enough that it will come down to using your ears to make the best judgment. That said, there are, as I mentioned, plenty of tutorials that would give you an excellent starting point. I also highly recommend the Mike Senior books on recording and mixing as they are beginner-friendly and aim to give you the essential information quickly.
EDIT: Stay away from Calf plugins (no end of instability and sound-quality problems for many folks) and try the set by LSP (mentioned relatively recently given a new version). They are excellent, as are the x42-plugins by Robin Gareus. The Ardour manual also maintains a list of “good” plugins.
Yes, well I am aware of the loudness wars. For better or worse I’m in a very small loudness war already. I’m not trying to sound huge, just not anemic relative to my friend. In any case could anyone direct me to more targeted tutorial on the web? Also is there a tool I could use that could help me identify which track has some dynamics that I could tame down with a limiter and/or compressor plug in?
Any of the LUFS meters will be good for overall volume. The more useful ones will also give you read-outs on dynamics so you can tell how squashed something is. This is the one I’ve been using recently: https://www.klangfreund.com/multimeter/. You will see it gives you plenty of dynamic readings including PSR, PLR and LRA (descriptions of these are on that product page).
Makeup gain is applied after the compression - you can think of it like a fader after the compression plugin. Compression generally brings the level down so you need makeup gain to bring it back up again. You’re looking for a limiter (or compressor with an infinite ratio) and then you’re going to be adjusting the threshold (lowering it) and then bringing the level back up with makeup gain.
Any YouTube video on master limiting should show you the basics.
Since you’re using Calf plugins here is my answer of what I do:
I mix all my instrument buses pretty low. I like to keep everything around -15.0 dB
For my music (folk and surf styles), that will make the master bus level average around -6.0 dB with everything playing at once at the choruses, etc.
Then when I’m all done mixing/mastering the song I :
Put a Calf Limiter plugin on my Master bus.
Set the Output Gain dial to -1.0 dB
Set the main “Limit” dial to something like -5.0 dBFS
Then I start raising the Input Gain slowly, watching the input and output levels.
I want the Output level to be pretty close to -1 dB at the loudest parts of the song, but i don’t want verses to sound quiet and dinky either. So raise the input slowly while checking the levels. Generally, I want my verses more chill, a bit quieter, and the choruses punchy. Too much compression and the song may not sound dynamic anymore.
The -5 dBFS is telling the plugin to compress the sound, (raising the quiets, and squishing the louds). You don’t want too much compressed though ,you still want your song dynamic and your overall volume is part of that. So some songs will only need -2 dBFS of compression, some may need more! It depends on your song, genre, and your ears! Doing all this will raise the overall volume of the song and let you control it carefully so you don’t kill other aspects of the song.
Hope this helps, And remember there is no one-size-fits-all here, that’s why i’m just sharing my general settings.
which are free (as in freedom), Open Source and multi-platform.
Another renowned option is this one:
Youlean L. M. is closed-source and commercial (there’s also an usable free -as in free beer- version, though). No native Linux version is available, but the windows version runs without problems under Linux with wine (via LinVST).
For Linux, a good native alternative can be “lufsmeter” (based on klangfreund’s code), readily available from kxstudio repositories:
@BonesInTheWalls , I’ve pretty much followed your advice and still not entirely happy with my results. Could be most of my problem is due to not optimally exporting for my needs. One area of confusion for me is the export option of normalizing. I can readily understand how failing to normalize to peak would make my WAV file quieter than my friend’s. What I don’t understand is normalizing loudness. This is something different from normalizing peak, produces something presumably different. But this different thing will presumably have peaks lower than 0dBFS or higher than 0dBFS, either of which would be a problem for me, am I wrong? either too quiet, or clipping?
Hey, new member here - Full disclosure I actually haven’t used Ardour much except years ago with a previous version. However, I have had tons of experience using other DAWs, and have hundreds of hours in a studio as an executive producer (now, am I a good executive producer? Jury’s still out LOL).
On to my suggestion: In many cases, engineers will boost the audio signal simply by duplicating tracks, instead of doing any processing on those tracks. This might be a little dependent on the DAW being used, and I don’t know exactly how Ardour will handle this. BUT, I can assure you, in practice, duplicating a track is usually a surefire way to increase the volume of that recording without bumping up against clipping, distortion, etc.
Usually, this is just used to boost a single instrument - most notably vocals - prior to adding effects to one of the duplicated tracks. But I don’t see why it wouldn’t also work in your situation.
Obviously your mileage will vary. My fingers are crossed that I’ve actually told you something helpful!
Thanks @irelandm I’m not sure I’m looking to give any particular track a boost. If I just turn up the volume knob on my audio player my overall mix sounds greaat. It’s just that I have to turn the knob up in a way that I don’t when I play my friend’s WAV file, and we’re sharing e.g. the same very loud sounding (probably compressed) drum track.
Gotcha, so just for funzies, try this then: Create a new project, import your completed mix WAV that you already have as a new stereo track. Then duplicate that track onto a second track. From what I understand, there’s no limit to how many times you do this, just that eventually it becomes impractical. I am sure someone out there has done the math, but it’s probably similar to each additional track adding a few decibels. One of the key benefits to doing it this way is that you avoid volume clipping which happens on individual tracks when you just use amplification filters/effects. Again, your mileage may vary … I haven’t yet tried this with Ardour. It’s entirely possible that it does nothing, LOL! But I can say that it does work in Protools and in Audacity. It’s definitely one of the tricks one can use to get that big sound.
Well, I’d be an idiot to argue with you, especially considering my limited experience with Ardour specifically. Please keep in mind, though, that my suggestion was entirely intended to be a kindness with the goal of giving a self-professed new user a really simple option which I have had much success with in the past using other DAWs. To me, this is a case of no harm, no foul. As to whether it has no basis in any part of digital audio, well … again, I’d be an idiot to argue with you. Maybe it’s not precisely to “avoid volume clipping” as I said. Maybe to my terribly inexperienced ears it just sounds better and is entirely psychosomatic. Your mileage may vary :
Either way, I wish the OP success in his efforts.
Just to be clear, replicating the tracks will increase the volume. But the result is absolutely identical with doubling the gain multiplier. So it’s not that your suggestion won’t work, it’s just an incredibly inefficient way to do this.
Fair 'nuff. And from a technical perspective, it does make sense. I can honestly say, however, that I will absolutely continue to use the track duplication method going forward. While it may be wrong (especially when it comes to Ardour in particular), it satisfies my own ears, whether psychosomatic or not. And my own modest PCs have no trouble with it.
A modern PC can generally handle 100+ tracks without breaking much of a sweat unless there is a lot of plugin DSP going on. II’ve done 400 on mine without any real effort.
So yes, there’s a lot of headroom for this if you have a small session. But in any DAW, replicating tracks is more or less guaranteed to double the CPU load associated with the initial track (not quite, because disk data will likely cached).
If you have done what I’ve explained and it’s still quiet, there are pretty much only two possibilities I can imagine. 1. Your Limiter is in the wrong spot, or still has the wrong settings, or 2. Your Master bus fader isn’t at unity (0db), so it’s cutting your volume at export.
First, make sure the limiter on your master bus is the VERY LAST plugin (bottom most). Also make sure your master fader is set to unity (0 db). So if you followed my instructions, your limiter is showing audio output as close to -1 or -2 db. If it’s not showing that at its output meter, then increase the input dial more until you’re closer to that point.
On your master fader’s meters you can see exactly how loud your track will be at export. If it’s still bobbing around way below -1 to 2-, then simply turn the limiter input dial up more, or it’s limit dial more (compressing it further, so be careful with that dial).
My second guess of your issue, is that you simply have your master fader dial turned down, so after all your hard work getting the limiter perfect on its meters… The master fader is then cutting all that volume you’ve added immediately back down again. Does that make sense? The limiter makes it louder but then your fader turns it back down again?
-Mix your track
-Set master fader to 0 (basically you never need to touch your master fader)
-Add a limiter as your last plugin
-Push the compression and input volume in the limiter as high as you’d like.
-Take a final glance at your master fader’s average levels to make sure they’re loud enough (this will vary by genre/song).
-If they’re not, go back into your limiter and turn it louder.
-Export the tune.
Basically if it sounds good I’m your headphones, and the mixing is done, then don’t worry about normalizing. You’re already way past that point. It’s just time to get your limiter set right, and export.
I hear a lot of complaints about Calf plugins but I’ve used them with Ardour for years. I’ve never experienced any unreliability (apart from when an outdated LADSPA plugin that came bundled with Ubuntu Studio broke everything until I uninstalled it).
Have you played with multiband compression at all? It can be much more subtle than just compressing all frequencies equally, and can help to fatten up mixes without being too obvious. I use the Calf Multiband Compressor a lot.