AI Loudness Meter Monitoring in Realtime

[TheMadman_tobyjenner]

About 3 years ago I posted a topic about using a Loudness Meter to get the optimum recording levels or performance levels for our OMs. It was more about using a meter in a DAW or OBS opposed to monitoring what is coming out of your AI in real time.

https://community.justinguitar.com/t/getting-the-right-loudness-levels-and-mix/119812/21

Doing some post show digging around and trying to work out how we could use a Loudness Meter for OM soundchecks, I worked out/remembered how to monitor at a desktop level or directly for an AI or audio source.

So although this does not address the soundcheck situation yet, it should allow folks to get pretty near to the optimum sound levels going into a JGC OM event, without the use of a DAW or video/audio recording interface like OBS.

So I created a quick video guide on how to go about, using Youlean Loudness Meter 2.
The thumbnail is a little teaser

AI Loudness Meter Monitoring in Realtime

Obviously this was monitoring a complete song. If you are using this process when setting and/or rehearsing its the overall level that should be around -14 LUFS.

As a rule I would sing the vocal part on its own and get this to around -15 on its own.
For backing tracks I aim for around - 18 and the guitar part I am playing about -16.
This ensure my guitar is clearly audible over the BT but not drowning it.
I would the fine tune the vocal level against the BT/GTR for balance, again with the overall track level coming in around -14 LUFS.

So there you go, hope you find this useful. It is possible that this should also be possible with other loudness meters that can run in standalone mode. Also this has been done on a Win PC, so may vary on other platforms. But I’m old skool and staying that way, too old to change !

Thx

[CD02]

This is great info, thanks for sharing. Levels are something I’m toiling over at the moment and there is a lot of info on YouTube, not all of it good.

I’ll have a play around with some of this tomorrow.

[TheMadman_tobyjenner]

Craig if you are just uploading to YT these are the levels you should be aiming for. Might be worth your while looking at some of the other articles in this sub-category, as it covers a wide range of set ups.

[DavidP]

Thanks Toby. Exactly what I had in mind when I asked you if you had your meter running. I was on my laptop in my office rather than the PC so didn’t have it available.

I’d encourage all to make use of this as part of prep for an OM. I don’t think it is too difficult to get set up on any set up, from most to least complicated.

[CD02]

@DavidP

I’d encourage all to make use of this as part of prep for an OM

Agreed, I think a lot could improve their overall sound with a little bit of this.

Whilst I have used Youlean I have really only used it on the recorded (Reaper) version and not to set incoming levels at the time on recording so I’ll give this a try.
The problem I have faced using it the way I am is that I struggle to balance a reaching the approx -14 LUFS I with whilst keeping the Peak from clipping. I am using a compressor in Reaper to see if this helps but am now thinking that by setting up the incoming signals using Youlean I may find that this whole process is easier.

[Malz]

This is a little over my head. Though will try to work on it when I find time. Looks useful, thank you sir.

[TheMadman_tobyjenner]

@CD02

whilst keeping the Peak from clipping.

Regardless of setting up for OMs or just recordings, if I am trying to get a base I always put
ReaLimt and Youlean Loudness on the master track (I put both on individual tracks but lets keep it simple).

I set the ReaLimt Brickwall Ceiling to -3 db

This is well under the -2 db True Peak that YT is happy with and will eliminate any clipping above -3db. You can the use the Threshold setting on RL to push the Master level and monitor via YLM and fine tune to around -14.

Once you have got that base setting in a DAW and you were not using the DAW and additional effects in the DAW for say an OM performance, just don’t touch any dials !!
As they should then remain constant. If you are using ReaRoute to OBS to Zoom just monitor the DAW’s master for starters.

You can the do a test recording in Zoom and drop the resultant video file into YLM running it in standalone, which will confirm what Zoom is receiving and I should add possibly “adjusting”. We have noticed Zoom dropping output to -17 LUFS/DB when a input signal of -14 has been presented. We don’t know why. For example video recorded in OBS shows around -14 and good to go. Record in Zoom and same performance shows as -17 ! ! !

But I found out setting up for the last OM, by nudging Zoom Mic Level a fraction and doing some fine adjustments and then checking the output in YLM I pretty much came in bang on target, then left everything alone !

Hope that helps.

[TheMadman_tobyjenner]

This is what you will see from the Master when ReaLimit is managing the clipping

In this example from the Drum track - when creating the OM backing track - I’d set the ceiling to -2 (Master was -3) but the imported audio from Guitar Pro was quite low, so I bumped the Threshold up until I was happy with the track.

That was to get the drum track up to around -18 LUFS-I. Same starting point for the bass track. Vocals will be loudest so aim for -14 and the guitars should slot in between.

Slightly off topic but does cover the set process we are trying to monitor here.

[CD02]

Oh, this is brilliant, thanks Toby for taking the time to share. I’ve never used ReaLimit so there is plenty there for me to trial with some setups. Really appreciate it

[Majik]

@TheMadman_tobyjenner

Obviously this was monitoring a complete song. If you are using this process when setting and/or rehearsing its the overall level that should be around -14 LUFS.

A small note, as many people may not understand what the various numbers mean, and this might be important to how people use these tools.

When measuring “loudness” (which is what “LUFS” aims to do) there’s a few different measurements:

Momentary is an “instantaneous” reading of the loudness (actually, the average over 400ms)

Short term is a measurement of the loudness averaged over the last 3 seconds

Integrated is the measurement over the whole “program”, which usually means an entire song. In the case of Youlean, it means since the meters were last reset.

Youlean Loudness meter shows the Short term, Integrated, and Momentary max.

Youlean Loudness Meter - MeasurementsExplained

How to use these?

For general levels, I would suggest looking at the Short term Loudness and, perhaps, use that to set the levels based on the louder parts of the song.

Ideally, you should rehearse the song all the way through and capture the Integrated level, especially if the song is particularly dynamic (lots of variation in level) as that will give you a view as to whether the whole performance is a good level.

If, for instance, the loud parts of the song are at -14 LUFS or above, but the whole song is well below that, then it may be beneficial to increase the overall level and use a limiter to control the peaks (which has been discussed above).

A reminder: the integrated LUFS measurement is from when the Youlean meter was first started or reset. You shouldn’t leave big gaps at the start or end of the song whilst you are rehearsing as that will impact the reading. Similarly, if you adjust the levels and re-perform the song, you should restart the metering.

Cheers,

Keith

[TheMadman_tobyjenner]

Useful information Keith, thanks for adding.

In the video I made frequent reference to the integrated LUFS value - LUFS-I which I tend to use when checking the overall level of the whole song. That could have been made clearer in the text.

And yes you really need to clear the audio capture each time you measure. To do that hit the red circle with the cross and it will clear the entry, ready for the next pass.

Bottom left side of application window.

[CD02]

I know this is more mix related, but also recently found that a multi track consisting of percussion, guitar and vocal tracks sounded fine to me when I listened in my headphones to the complete Rendered sound but when sent to my phone (even though its a late model with good sound) the percussion seemed lost and barely audible. I am guessing because of the bass response or lack of in the mobile devices.

Anyway, I found that by cranking the backing a bit, what then sounded a bit too much in general worked well on the phone.

This happened because I specifically wanted it to go to a phone.

I’m sure this was mentioned before, I seem to remember seeing something.

[Majik]

It’s well worth checking mixes on a phone and also, if you can, collapse the mix to mono and check how that sounds. Often a mix which sounds wide and full in stereo on good speakers or headphones will sound wrong in mono or on a phone speaker (which lack bass and are, effectively, mono).

It may well be that inverting the phase of one track makes it work better in mono. There’s also plugins which will check for stereo phase issues (I use x42 Stereo Phase Correlation Meter)

Another trick is to play it on a car stereo, as that can also reveal weaknesses in the mix.

Cheers,

Keith

[brianlarsen]

“Can you hear me?”
“I said: CAN YOU HEAR ME?” works fine for me at soundcheck

I have developed a bad habit of drifting into the naughty red zone…
Then again I’m used to having a clip round the ear