TBProAudio releases mvMeter - Multivariable Meter including RMS, EBUR128, VU and PPM measurement for

[Compyfox]

I'm on to it. Give me 1-2 hours, then I can tell you.


But according to GS on v1.01

Well, PPM is not 2.5dB off, it is just calibrated to -12dBFS.

... *sigh*

Let's do a quick math here...
If 7 on the PPM equals -12dBFS sine test signal, that means that 4/Test on the scale equals -15dBFS.


I sent in a -18dBFS sine signal into the PPM mode and it showed me 2,5 on the BBC side, and -6 on the EBU side. Granted, it's was not 2dB or 2,5dB off (apologies for that)... but 2,5PPM/BBC and 6PPM/EBU... The "calibration" was therefore wrong to begin with (-18dBFS is a standardized test signal you know).

A better explanation in the manual would have been TEST/4 = -12dBFS (despite the fact that this doesn't make any sense at all) rather than 7 = 0dBFS.

BTW: all meters in mvMeter have a well defined calibration, the manual helps here;)

*snicker* Sure...





Be back in 2 hours max...

[TB-ProAudio]

Thanks for the update!

I don't really know most of the "wrong stuff" going on for this plugin, but does this new version fix all those bugs that Compyfox talked about? Or just some of them?

Some points Compyfox mentioned are either fixed (VU timing, EBU readout) or described in more detail (meter ballistics and calibration).

[Compyfox]

As promised... my report


For the TL;DR people:

Question: Has the meter been fixed?
Answer: partially

Question: Would you recommend/use it?
Answer: no / only if it's accurate and I have no other choice of selection





IN-DEPTH ANALYSIS

Digital Peak:
Seems to be accurate, but suffers the same graphical issues as the RMS meter





PPM:
According to the manual, the PPM is calibrated with -12dBFS Sine, resulting with a -18dBFS test signal in 2,5PPM/BBC and -6PPM/EBU. The correct value with a -18dBFS/1kHz Sine test signal should be 4PPM/BBC and TEST/EBU however.

I then proceeded to test with a 10ms Sine burst followed by 500ms of silence (in a constant loop) to test if the signal reaches 85-90% of the maximum peak within the given time frame (10ms). I used VUMT2 Deluxe for comparison again (calibrated to -12dBFS, like mvMeter) and ignored the "fallback time", which is really only for more more easier-on-the-eyes visuals.

The readouts:
mvMeter BBC: 2,3
VUMT2 BBC: 3.13

muMeter EBU: -6,5
VUMT2 EBU: -3,5


This gives me the impression that the ballistics are not 10ms. So I lowered the rise time in VUMT2 to find the correct values.

The readouts:
mvMeter BBC (supposed 10ms to reach 85-90% of the max value): 2,3
VUMT2 BBC (roughly 35ms): 2,4

muMeter EBU (supposed 10ms to reach 85-90% of the max value): -6,5
VUMT2 EBU (roughly 30ms): -6,5

Also checked this with zplane's PPMulator - and I can confirm mvMeter's offsets in there as well





VU Meter:
Static test tone - correct
300ms sine burst/silence - feels offset, but shows the correct values.
Full spectrum material (or anything other than sine test tones) - offset of +-0,5dB to +-1dB

I assume that there was no "needle smoothing" coded, since the behavior is jumpy.





EBU R-128
Works like it should with the known EBU R-128 test sequence
offset +-0,1LU within acceptable range





RMS Meter:
According to single-frequency test tones (bursts), it "reaches" the desired position (-20/-14/-12) in time. But once you use a full spectrum signal (regular music), the meter is off by 2dB. This is due to the ballistics being 300ms and not 600ms (Dorrough 40A specs).

On top of hat, this mode feels like as if it's dropping frames (lagging needle, Wavelab 8 ). Which is getting worse in Wavelab 9. I take it the graphics engine hasn't been optimized (needle redraw due to the scaling).

The scale is also a "relative scale", not an "absolute" one.
Apparently, the "RMS Meter" doesn't use a weighting filter anymore.


QUOTE FROM MANUAL:

RMS: AVG RMS 300ms based on Bob Katz's K-System, several calibration references

Wrong - Bob Katz's K-System (v1) is based upon the Dorrough 40A with a shifted reference level. The Dorrough "idea"/patent states that it's "twice as long measurement time as a VU meter", meaning 600ms. If you want to copy my commentary and sell it as your own - do it right, please.

EDIT:
There "are" RMS meters at 300ms time window - but this is not the "Bob Katz' specification" and kind of obsolete, since a VU also already uses 300ms.





Additionally:
- I did not test the TRIM mode if this causes issues while rendering
- the criticism of the nearly 1:1 copied skins from Klanghelm still remains
- furthermore, the Meter/Scale specification chart on page 4 of the manual is nearly a 1:1 copy from Klanghelm's VUMT2 yet again

You can't convince me of any coincidence anymore...





FOOTNOTE:
Testing all odds/ends eats up a lot of time. Since I'm not being paid for this amount of work, I really have to consider investing another 2-3 hours for further testing the "accuracy" of this tool with a possible (mandatory!) future update. I think this is why VUMT2 was so long in the making... Looks like Klanghelm's beta testers were really busy...

[provoc]

Found a nasty bug in reaper where when you click on the gain to drag but it jumps straight away to +24dB!

[Chris-S]

Thanks for this freebee.

What is the difference between "stereo" and mid channel?
@provoc: Can't reproduce your bug in Reaper.

[Eleventh]

Found a nasty bug in reaper where when you click on the gain to drag but it jumps straight away to +24dB!

Did you shift+click it and engage the auto-gain feature? Sometimes accidentally do that too... I think a separate button for that feature would be less error-prone.

[heavymetalmixer]

@Compyfox What's you opinion on these meter plugins and how they are compared to the mvMeter?

1) HOFA 4U meter
2) dpMeter2
3) Sonalksis FreeG
4) Melda MLoudnessAnalyzer
5) Audiocation Loudness meter
6) Algorithmix Gmb H TT-DR Meter
7) Sleepy-Time DSP MonoChannel/StereoChannel
8 ) Voxengo SPAN
9) Youlean Loudness Meter


@TB-ProAudio
Does the autogain feature means that mvMeter was made to replace the AutoGainStage JSFX?

[Compyfox]

@Compyfox What's you opinion on these meter plugins and how they are compared to the mvMeter?

1) HOFA 4U meter
2) dpMeter2
3) Sonalksis FreeG
4) Melda MLoudnessAnalyzer
5) Audiocation Loudness meter
6) Algorithmix Gmb H TT-DR Meter
7) Sleepy-Time DSP MonoChannel/StereoChannel
8 ) Voxengo SPAN
9) Youlean Loudness Meter

You're trying to compare Apples and Oranges

1) declared as "EBU R-128 meter" bargraph
2) ITU-R BS.1770-x / RMS (summed) bargraph meter
3) Digital Peak and IIRC RMS meter (is that really still available?!)
4) EBU R-128 bargraph meter
5) EBU R-128 meter, numeric readout (discontinued IIRC)
6) Algorithmix never created the TT-DR Meter, it was created by Brainworx and later re-rereleased as bx_meter - RMS, K-System (v1), Digital, Brainwors re-release with weighting filter, bargraph meter
7) VU Meter only (sadly discontinued)
8) Digital and RMS meter, FFT analysis, bargraph meter
9) ITU-R BS.1770-x meter at it's core, presets like ATSC/85 and EBU R-128 possible


If you want to compare mvMeter with anything, then it's PSP Triple Meter, zplane PPMulator XL and especially Klanghelm VUMT2 (Deluxe). All of them are needle meters. mvMeter's... concept... is an obvious clone of the last mentioned metering tool in this paragraph.

@TB-ProAudio
Does the autogain feature means that mvMeter was made to replace the AutoGainStage JSFX?

There is no "auto-gain feature" in mvMeter

[TB-ProAudio]

Found a nasty bug in reaper where when you click on the gain to drag but it jumps straight away to +24dB!

We will look into this, should not happen. Is it reproducibly? Thanks

[TB-ProAudio]

Thanks for this freebee.

What is the difference between "stereo" and mid channel?
@provoc: Can't reproduce your bug in Reaper.

"stereo" is either summed (EBU) or averaged loudness of the 2 stereo channels.

"Mid" is just the mid signal only of the stereo signal, fed to both internal channels, in opposite to "Side" signal.

[TB-ProAudio]

Found a nasty bug in reaper where when you click on the gain to drag but it jumps straight away to +24dB!

Did you shift+click it and engage the auto-gain feature? Sometimes accidentally do that too... I think a separate button for that feature would be less error-prone.

Thank you, we will think about this.

[TB-ProAudio]

Does the autogain feature means that mvMeter was made to replace the AutoGainStage JSFX?

No, this function matches max. readout level with 0 level of the respective meter mode, only once with the mouse click.
AutoGainStage JSFX or AB_LM does this constantly.

[b4serenity]

Compyfox knows his stuff people. Why he gets any gruff about it is beyond me. If you have problems with his scientific data then argue against it with better data. Calling him names only shows a persons lack of maturity. I, for one, thank him for spending his time to help us all. Just how much time have you spent unselfishly spending hours of time just to help the KVR community?

[provoc]

Found a nasty bug in reaper where when you click on the gain to drag but it jumps straight away to +24dB!

We will look into this, should not happen. Is it reproducibly? Thanks

Hmm tried it again this morning and it seems to be ok today .. I did press shift click at one stage when I was testing so maybe somehow the shift function got stuck? Will post here if it happens again and I work out how to reproduce.

I did however note that if you press shift and drag it jumps straight to +24db even with a loud signal going through (at least for me here on latest reaper pre release x32) which is not safe and should probably be changed .. This is easy to do since plenty of plugins use shift for fine movement so I suspect it will be a habit for many!

Anyway thanks for your efforts and all the best with your ongoing development

[heavymetalmixer]

Does the autogain feature means that mvMeter was made to replace the AutoGainStage JSFX?

No, this function matches max. readout level with 0 level of the respective meter mode, only once with the mouse click.
AutoGainStage JSFX or AB_LM does this constantly.

Cool then, I'll keep using both. Is there any mvMeter update coming soon?