Youlean Loudness Meter 2 - V2.5.10 - November 10, 2024 update

[Compyfox]

Er... just a quick Q
Why do customers not get update mails?

[Youlean]

Er... just a quick Q
Why do customers not get update mails?

You will need to subscribe to the newsletter first.
I have not implemented automatic opt in yet, and this can be tricky because of GDPR...

You can do that here for example:
https://youlean.co/newsletter/

[Compyfox]

Blasted GDPR - especially if you use a third party engine for newsletters. But okay, I'll (re)sign up


EDIT:
Tried, page doesn't "forward" to the next step like "you've been signed up" or "confirm" or something. The submit button shows an animation, and that's it. Tried in three browsers. No dice.

[Halonmusic]

Thanks for this Youlean. Btw im new when it comes to EBU, i use VU meters. What is the difference?

[plexuss]

Thanks for this Youlean. Btw im new when it comes to EBU, i use VU meters. What is the difference?

Meters like VU, RMS, peak measure the physical parameters of the digtially represented analogue signal. They are "technical" measurements. EBU and the like measure based on how the sound is percieved by humans. Typically. Using accepted generalized models of human hearing. This way when you take a measurement you are doing so with respect to how the sound is percieved, rather than it's technical make-up. As you may know there are a lot of complex non-linearities between the technical measurements and the way the audio is percieved (eg. Fletcher/Muson, Equal Loudness, etc).

The result of using loudness meters is that you are mixing and adjusting now based on things like how loud it will sound, percieved dynamic range, etc. It's easier to match intended and expected results with perceptual measurements. It's also prudent to continue using technical measurements to also ensure technical quality (eg. no clipping, headroom, noise floor, etc).

[Halonmusic]

Thanks for this Youlean. Btw im new when it comes to EBU, i use VU meters. What is the difference?

Meters like VU, RMS, peak measure the physical parameters of the digtially represented analogue signal. They are "technical" measurements. EBU and the like measure based on how the sound is percieved by humans. Typically. Using accepted generalized models of human hearing. This way when you take a measurement you are doing so with respect to how the sound is percieved, rather than it's technical make-up. As you may know there are a lot of complex non-linearities between the technical measurements and the way the audio is percieved (eg. Fletcher/Muson, Equal Loudness, etc).

The result of using loudness meters is that you are mixing and adjusting now based on things like how loud it will sound, percieved dynamic range, etc. It's easier to match intended and expected results with perceptual measurements. It's also prudent to continue using technical measurements to also ensure technical quality (eg. no clipping, headroom, noise floor, etc).

Thanks for the explanation

[Compyfox]

More detailed, yet simpler to understand.


Maximum signal strength measurement:

dBTP (SPPM) - Digital peak meter: measures the maximum (digital) signal strength, oversampled, sample accurate, can measure digital overs. Absolute scale

SPPM - Sample Accurate Digital Peak Program Meter: measures the maximum (digital) signal strength. Absolute scale

PPM - Peak Program Meter: measures the maximum signal strength, due to inertia (response time of the meter, 5ms or 10ms), this meter needs a "ceiling"/headroom. Relative scale



Average signal strength measurement (non-weighted):

VU - volume unit (VU) meter: measures the average signal strength, due to inertia (response time of the meter usually around 300ms), this meter can only pick up "average" signal strength and does not respond well to signal peaks. Another downside of this meter is it's stronger response to more bass intensive material. Relative scale, usually calibrated to an alignment level (eg. EBU R68/the infamous -18dBFS or SMPTE RP 0155/-20dBFS equaling to 0 VU on the scale)

RMS - Root Mean Square meter (realtime, average): the realtime RMS meter measures the average signal strength, the response time of the meter usually around 600ms (Dorrough 40A hardware specs, according to the patent "roughly twice as slow as a VU"), this meter can only pick up "average" signal strength and does not respond well to signal peaks. Another downside of this meter is it's stronger response to more bass intensive material. Absolute scale. WARNING: each RMS realtime meter does it's on thing, except the K-System meter or the actual Dorrough range. Check in with the (hopefully detailed) manual of the meter you're using, as realtime RMS meters can range from 300ms to 600ms "analysis window". Most of these meters use the AES-17 +3dB compensation (example: a -20dBFS test signal shows -20dBFS on the meter, not -23dBFS)

BONUS INFO:
The K-System (v1) meter is built upon the RMS realtime meter (Dorrough 40A) with a shifted reference point (in this case: -20dBFS = 0 reference, -14dBFS = 0 reference, -12dBFS = 0 reference). Not only that, it uses the AES-17 +3dB compensation (a test signal shows both the same value on the dBFS and RMS readout). Relative scale on most "K-System" type meters, absolute scale on Dorrough meters.



Average signal strength measurement (weighted) / "LUFS meter":

ITU-R BS.1770-x (as main example):
For average signal strength measurement, all incoming channels are being summed down to "mono", signal is then being run through a so called "weighting filter" (in this case: k-weighted, HPF at around roughly 100-120Hz, and a high shelf starting around 1kHz) before the signal will be analyzed through several mechanics. The weighting filter fixes the offsets due to stronger bass frequencies, and puts and emphasis on the frequencies that can hurt our ears. For sake if simplicity, we're talking ITU-R BS.1770-1, which doesn't introduce gating of signals.

MLk measurement (momentary loudness, k-weighted): analysis window being 400ms
SLk measurement (short term loudness, k-weighted): analysis window being 3s
ILk measurement (Integrated Loudness, k-weighted): a mathematical average of the measured signal (depending on the used standard, usually based upon MLk) from start of the analysis to the end of the analysis. Also known as "program stream measurement (in the old days, that was Leq(m), or "Loudness measured, filtered, over time")

EBU R-128, ATSC A/85, TR-B32 and OP-59 are technically the same as ITU-R BS.1770-x. EBU R-128 was the "forerunner". The main differences are usually in the target level (ILk, mostly -23LUFS or -24LUFS for TV/Radio Broadcasts, Netflix is trying to adapt -28LUFS IIRC, "web"/iTunes/Spotify is aiming for -16LUFS ILk), the allowed offsets (+-1LU or +-2LU shift from target ILk), if you have an absolute or relative scale for the average signal strength measurement (absolute = LUFS/LUKS, relative = LU), and what the allowed ceiling for the maximum digital signal strength are in dBTP (-1dBFS or -2dBTP). Though this particular standard ("recommendation" for some of the mentioned ones) is in constant motion/fluctuation.

Maximum digital signal strength is not being summed in any form.
Everything else like LRA and DR, is bonus information.




Examples in the real world:

Recording and Mixing:
Usually a combination of either a SPPM and VU, or SPPM and Realtime RMS meter

Mastering:
Can either use K-System metering (Dorrough 40A) or ITU-R BS.1770-x/similar type meter

Broadcast (Radio, TV, Cinema):
ITU-R BS.1770-x/similar type meter



If you activated "signatures" in your forum account, check out my KVR Marks - I've talked plenty about all of this on KVR Audio.

[Halonmusic]

Thanks Compyfox

[Compyfox]

Glad I could help. I should add this to my KVR Marks

[Halonmusic]

[Youlean]

Blasted GDPR - especially if you use a third party engine for newsletters. But okay, I'll (re)sign up


EDIT:
Tried, page doesn't "forward" to the next step like "you've been signed up" or "confirm" or something. The submit button shows an animation, and that's it. Tried in three browsers. No dice.

This is really strange. Here it worked fine... Not sure what could be wrong. I will investigate more.

[Youlean]

Thanks guys for the explanations!
You really know this thing well!

[Youlean]

I have just made a breakthrough on mac and now the gui is 2x more smoother, which will be much nicer on retina displays.

Have to do more testing with button glitch, but this will be solved too.

[Hermetech Mastering]

Thanks for all your work!

[Youlean]

No problem!