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

[TB-ProAudio]

In stereo mode, the average level of both channels is calculated: (l+r)/2.0.

And this is different from Mid?

My understanding is that Mid=Mono=(L+R)/2

EDIT: After some testing I think that in Stereo mode you calculate the max(L,R) and in Mid Mode the average (L+R)/2.

Well, for instance in VU mode:

Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0; // first VU level calculation of each channel and then average level calculation
Lvl_mid = VU((Inp_l + Inp_r)/2.0); // first mid signal calculation and then VU level calculation

There must be a difference as Inp_l/Inp_r can be negative and positive, the result of VU() can only be positive. So if you have equal signals on left and right channel and left signal is phase-shifted against right signal Lvl_mid can even go down to zero, while Lvl_stereo still shows the average level of left and right signal. (all levels before dB calculation)

I hope this makes the difference of Stereo and Mid mode clear

[Chris-S]

Yes, thanks.

[Compyfox]

First and foremost - I said I apologize if I messed up, and I will...
I did take mvMeter for another spin. And looks like the "Offset with mono signals" was a possible user error.


However...

Well, for instance in VU mode:

Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0; // first VU level calculation of each channel and then average level calculation
Lvl_mid = VU((Inp_l + Inp_r)/2.0); // first mid signal calculation and then VU level calculation

There must be a difference as Inp_l/Inp_r can be negative and positive, the result of VU() can only be positive. So if you have equal signals on left and right channel and left signal is phase-shifted against right signal Lvl_mid can even go down to zero, while Lvl_stereo still shows the average level of left and right signal. (all levels before dB calculation)

I hope this makes the difference of Stereo and Mid mode clear

Okay, I invested day just for this, and I think this DEFINITELY needs further clarification - maybe even correction. At bar minimum, further documentation in your manual, or as part of a technical paper. Because you are (to my knowledge) the only person who does that type of "summing" (even in your other meters in RMS mode - though I seem to remember that the RMS mode is summed according to the ITU-R specs), and this results in offset "mystery readouts".


Let me explain why:
A VU meter is not(!) made to show stereo signals on one needle alone. There is the route to either show L, or R, or the MID signal. A stereo signal should ideally be shown with a "stereo meter" (bargraph or dual VU), or with the "Highest value" from any given channel. Since the latter can never happen in the actual hardware world with "one needle" (also electronic wise), we only have access to Summed (Mid), L or R.



What you do, is IMO a confusing math, that doesn't even add up if you do it manually (or with an Excel Script) and then compare it with the readout of your meter. Therefore, offsets especially on stereo signals are inevitable, and throw off users unnecessarily.

This is the very reason, why most "single needle VU's on Stereo Signals" focus on the way simpler "Mid Value" (formula "(L+R)/2") rather than the formula you're currently using for the "Stereo Mode" ( Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0; ).

Actually, the correct value for the MID Signal would be:
=sum ( ((L+R)-3) *0,5 ) or =sum ( ((L+R)-3) /2 )
...same thing really (see Article on Sound on Sound)

The -3dB is added so that there is no increase in signal level after the summing.
Then we only want to have "one" needle (or bargraph if we talk LED bank), so he have to divide by the channel amount (divided by 2)



So let's do some math, like back in school, and see what would actually happen if I (for example) use the values +7,4 for L, and +9,4 for R. (you'll see why in a minute)

First, your "Stereo Sum" math:

Code: Select all

Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0;
Lvl_stereo = (VU(7,4) + VU(9,4))/2;

Now we have two ways to solve this...
Way 1:
Lvl_stereo = (7,4/2 + 9,4/2);
Lvl_stereo = 3,7 + 4,7;
Lvl_stereo = 8,4;

Way 2:
Lvl_stereo = (7,4 + 9,4) /2;
Lvl_stereo = 16,8 / 2;
Lvl_stereo  = 8,4;

The needle max value therefore shows 8,4 VU

So far so good... but what would an actual MID CH needle show? We're using the same values...

Code: Select all

Lvl_mid = ((L+R)-3) /2;
Lvl_mid = ((7,4+9,4)-3) /2;

First, we solve for the brackets

Lvl_mid = ((7,4+9,4)-3) /2;
Lvl_mid = (16,8-3) /2;

Then, this can go two ways...

Way 1:
Lvl_mid = (16,8/2) - (3/2);   or   (16,8*0,5) - (3*0,5);
Lvl_mid = 8,4 - 1,5;
Lvl_mid = 6,9;

Way 2:
Lvl_mid = 13,8 /2;
Lvl_mid = 6,9;


The needle max value therefore shows 6,9 VU

Now why is this important? I think an image says more than thousand words:

Janis Jopling - MS_vs_ LR_vs_Stereo_Sum_Math.jpg

Take note of the "Stereo Mode" meter to the top left



Right... so what would happen if the meter would actually be in the negative on one channel, and on the positive on the other? Let's find out!

This time, we use the values -0,6VU L / +1,4VU R

First, TBProAudio's "Stereo Sum" calculation again

Code: Select all

Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0;
Lvl_stereo = (VU(-0,6) + VU(+1,4))/2;

Let's simplify
Lvl_stereo = (-0,6 + 1,4) /2;
Lvl_stereo = +0,8 / 2;
Lvl_stereo  = +0,4;

The needle max value therefore shows +0,4 VU

And then with the MID CH only...

Code: Select all

Lvl_mid = ((L+R)-3) /2;
Lvl_mid = ((-0,6 + 1,4)-3) /2;

Solved
Lvl_mid = ((-0,6 + 1,4)-3) /2;
Lvl_mid = (+0,8 -3) /2;
Lvl_mid = -2,2 /2;
Lvl_mid = -1,1;


The needle max value therefore shows -1,1 VU

Why did I chose these values?
Well, I took the same mix, gain reduced it by -8dB and went for another spin. I wanted to see what each meter shows, and if there is some "negative value" flipped during the math or not.

Well, considering needle overshots, I'd say an offset of 0,1dB for the Mid Mode is still fine. Not so much for the Stereo Mode again...

Janis Jopling - MS vs LR vs Stereo Sum Math Gain Compensated.jpg

"Is there a spread sheet to confirm the values and simplify things?"
Not necessarily a spread sheet, but an Excel sheet where you can insert values and test away...

You can find that one on my server
EXCEL 2000 format (exported from OpenOffice), should work on both Win and Mac, no macros




"But why go all this hassle?", you might be asking?

Simple... something is still wonky. Or at least the "Stereo Mode" is. And this should be further commented on/explained in the manual (didn't happen so far), else Joe Normalguy uses the meter, thinks "this is accurate, this is how it should be", only to ultimately realize in direct comparison with other meters... "wait... why is my signal off by a couple of dB's?!". If he even comes that far (which is a problem I've been addressing in my last couple of posts).

If you run a mono signal through mvMeter, this is not such a big deal. The offsets are minor. But once we talk "stereo signals" (and may it only be an individual, mutating stereo pad), or full mixes... then it is of bigger concern.



"Less talking, more solutions for a fix!"

There are two possible solutions for this IMO:

1) explain in minute detail why you went this route, how to do the math, and why you think this superior to all other meters out there and time proven metering solutions. Basically - start a thesis, kick off a discussion. This has to(!) be written in the manual IMO. Even though I still think that this solution is not optimal, as VU's and PPM's were NEVER made to show stereo signals. Especially not this type of summing (which I don't even know how to build via electronic spare parts to be honest)

2) add a "dual needle view". Either with two screens side by side, or two needles on one scale. Then let the numeric max readout pick up the needle that jumped the highest. The competition does this with "Display LR" modes on one needle. Yes, that means PPMulator and VUMT.



"Did you just pull this out of thin air, or can you back up these findings?"
Ask, and ye shall receive...

Code: Select all

Testmaterial - Kick, Vocals, Gitarre, Virtual Orchestra Mix (30s)
(Individual Signals)

WAV 44/48kHz und 16/24 bit, input gain -6dB via Melda Production MUtility
for sake of simplicity, highest value for Duo/LR view only

mvMeter doesn't automatically switch into "Mono Mode"
VUMT2 switches into "Mono Mode" if a mono signal is applied, and "force MonoSum", aka MID, if the UI is switched to "Solo Needle View"



KICK 1 (original stereo, Sub Heavy)
mvMeter 2,2VU / 6,3PPM/BBC (Stereo Mode)
VUMT2 VU 2,3VU / 6,32PPM/BBC (Solo/M view)
VUMT2 VU 2,3VU / 6,32PPM/BBC (Duo/LR view)

KICK 1 (Mono L CH, Sub Heavy)
mvMeter 2,2VU / 6,3PPM/BCC (Stereo Mode)
mvMeter 2,2VU / 6,3PPM/BBC (Mono Mode)
VUMT2 2,3VU / 6,32PPM/BBC

KICK 2 (original stereo, Transient Heavy)
mvMeter -7,3VU / 6,0PPM/BBC (Stereo Mode)
VUMT2 VU -7,2VU / 6,03PPM/BBC (Solo/M view)
VUMT2 VU -7,1VU / 6,09PPM/BBC (Duo/LR view)

KICK 2 (Mono L CH, Transient Heavy)
mvMeter -7,4VU / +6PPM/BBC (Stereo Mode)
mvMeter -7,4VU / +6,0PPM/BBC (Mono Mode)
VUMT2 -7,2VU / +6,03PPM/BBC

Vocals (Female, Pop)
mvMeter +0,9VU / 4,9PPM/BBC (Stereo Mode)
mvMeter +0,9VU / 4,9PPM/BBC (Mono Mode)
VUMT2 +0,9VU / 4,87PPM/BBC

Acoustic Guitar (Original Stereo, Narrow Stereo field, Bass heavy)
mvMeter 0,2VU / 4,9PPM/BBC (Stereo Mode)
VUMT2 0,1VU / 5,06PPM/BBC (Solo/M view)
VUMT2 0,3VU (0,3L/0,1R) / 5,09PPM/BBC (5,09L/5,03R) (Duo/LR view)

Acoustic Guitar (Summed to Mono, Bass heavy)
mvMeter +6,1VU / 6,4PPM/BCC (Stereo Mode)
mvMeter +6,1VU / 6,4PPM/BBC (Mono Mode)
VUMT2 +6,1VU / 6,56PPM/BBC

Virtual Orchestra Mix (Wide Stereo!)
mvMeter -1,3VU / 5,2PPM/BBC (Stereo Mode)
mvMeter +0,6VU / 6,1PPM/BBC (Mono Mode L)
mvMeter -1,3VU / 5,2PPM/BBC (Mono Mode R)
VUMT2 -2,9VU / 5,10PPM/BBC (Solo/M view)
VUMT2 +0,6VU (0,6L/-1,3R) / 6,07PPM/BBC (6,07L/5,27R) (Duo/LR view)

Virtual Orchestra Mix (Summed to mono)
mvMeter +3,1VU / 6,5PPM/BBC (Stereo Mode)
VUMT2 +3,1VU / 6,61PPM/BBC





MUSIC (STEREO, input gain -8dB via Melda Production MUtility)
MP3 44/32 (due to Wavelab importing MP3 as 32bit)

AC/DC - You shook me all night long
mvMeter -2,7 VU / 5,3 PPM/BBC
VUMT2 -3,1 VU / 5,29 PPM/BBC (Solo/M View)
VUMT2 -2,7L/-2,0R VU / 5,14L/5,45R PPM/BBC (Duo/LR View)

Coming Back Around (How to Train your Dragon OST)
mvMeter 7,1 VU / 6,6 PPM/BBC
VUMT2 5,2 VU / 6,57 PPM/BBC (Solo/M View)
VUMT2 8,9L/6,2R VU / 7,0L/6,94R PPM/BBC (Duo/LR View)

Michael Jackson - Xcape (Dark Chao Version)
mvMeter 6,1 VU / 6,7 PPM/BBC
VUMT2 5,9 VU / 6,69 PPM/BBC (Solo/M View)
VUMT2 5,7L/6,6R VU / 6,69L/6,7R PPM/BBC (Duo/LR View)

Janis Joplin - Piece of my Heart
mvMeter -0,6 VU / 5,0 PPM/BBC
VUMT2 -1,2 VU / 4,97 PPM/BBC (Solo/M View)
VUMT2 -0,6L/1,4R VU / 4,94L/5,25R PPM/BBC (Duo/LR View)

"Right... now that's a lot of wood again to work through..."
Oh, sorry... I did forget something...



I also did this test to see if mvMeter has a Needle Overshot of 1% bar minimum (as this is how a real needle meter would/should work), and these are my findings:

mvMeter:
- VU Mode: Needle Overshot existent (>300ms Sine/random length of silence)
- Numeric Overshot: 0,1
- Testfile: 44/16 mono and stereo
- "Smoothing": feels chopped, especially with the PPM
- Overload LED goes off with EVERYTHING that goes beyond the reference level (so no "clipping" warning indication)

VUMT2 Deluxe:
- VU Mode: Needle Overshot existent (>300ms Sine/random length of silence)
- Numeric Overshot: 0,0 to 0,2
- Testfile: 44/16 mono and stereo
- "Smoothing": smooth needle, doesn't feel chopped (no matter which mode)




"Sorry - but I still don't get it... this is way over my head"
Trust me, I can relate. Let's just dumb it down...

Do I recommend this meter?

Well... this might be an interesting freebie. But there are still too many strange things happening (which confuses Joe Normalguy), lack of proper documentation, offsets in the Stereo Mode (with a summing matrix that doesn't make sense IMO), EBU R-128 scale is not good to read (unusable)...

Add to that my commentary on the
- "auto resets values" even if "Max Value mode" is active
- random usage bugs
- and my criticism on the "copied" UI/manual

Not to mention the selective radio silence by the developer regarding questions (I'd still love to know the development time), or commentary on how to improve this thing...


I'd say still "no, I don't recommend it".
I recommend to invest both your time and money into something else...




"Then why are you still in this thread?!"
Because the topic metering interests me, and I do to try to educate current and interested/possible future users with such posts...

If this helps the developer in the process, even better. But I think at this stage, I'm really way past my point. If there is another update (which I sure hope there will be), you now have ENOUGH information to do all these tests yourself. And if you're after accuracy in metering tools, I recommend you to do so. It's time consuming, but nothing is worse than using faulty tools.

I will now sit back, and mostly watch from the sidelines...

[frizzbee]

Actually, the correct value for the MID Signal would be:
=sum ( ((L+R)-3) *0,5 ) or =sum ( ((L+R)-3) /2 )
...same thing really (see Article on Sound on Sound)

actually this is not right.

mid = (l + r) / 2, side = (l - r) / 2

just try it with a (normalized) value of 1 for l & r (same signal at both channels = mid/mono, no stereo):
mid = (1 + 1) / 2 = 1
side = (1 - 1) / 2 = 0

now for l = 1 & r = -1 (antiphase = complete stereo, no mono):
mid = (1 - 1) / 2 = 0
side = (1 + 1) / 2 = 1

and these are the expected results.

the -3 (described in the article) at each sum is not false. but it is just to get the same results when decoding back to l & r (you have to lower the l & r then again). if you divide by 2 once (which is also mathematically correct) you don't need to do anything when decoding back, just summing. also when using the -3dB solution you might not get back to the exact values, because you are multiplying by a irrational number:

l = 1, r = 1:
mid = (1 + 1) * 0.707946... (-> this is the same as lowering the sum by 3dB & its an irrational number!) = 1.415892
side = (1 - 1) * 0.707946... = 0
back to l = (m + s) - 3dB = (1.415892... + 0) * 0.707946... = 1.002375...
bacl to r = (m - s) - 3dB = (1.415892... - 0) * 0.707946... = 1.002375...

1.002375 is not 1. this version also increased the mid signal value to a higher value than the input (may be right or may be not!?).

but what you are doing is a mix of both and this is wrong. just sayin

[subterfuge]

"Then why are you still in this thread?!"
Because the topic metering interests me, and I do to try to educate current and interested/possible future users with such posts...

I for one am glad you are still in this Compyfox, even if in sidelines now. Good information, kudos for extensive testing too.

[Compyfox]

1.002375 is not 1. this version also increased the mid signal value to a higher value than the input (may be right or may be not!?).

but what you are doing is a mix of both and this is wrong. just sayin

Honestly, and "just saying" - If I did mess up somewhere on this behalf, I'm terribly sorry. I've been at this the whole darn day, and at this point I only see numbers floating around my head.

If I go by the simple math (L+R)+2, then I get the same values that the OP (TBProAudio) implies with his Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0;

I mean, look at the formulas and compare it:
Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0;
Lvl_mid = VU((Inp_l + Inp_r)/2.0);

Where... is the difference other than the dev mentioning "yeah, one value has to be flipped to be positive". But even then... is it even correct? Was this not oversimplified then? This is one of the reasons why I MEASURED a track with both positive and negative VU values on purpose (Janis Joplin).


I see absolutely no difference here, unless thers is some "magic" going that we don't know (or is highly technical coding speech that 99% of us won't understand anyway - myself included, as I'm no programmer), and is not "understandably explained". In fact, your math with the 0.707946... is a bit beyond me. I mean... where does this "irrational number" come from all of a sudden? What is the source? What should it represent?

You say -3dB, but where is this based on? What "range" are we talking signal strength wise (from -xyz to 0)?
Is this a programming terms? Has this been taken from a JS script?


Also please keep in mind, my school math is a leftover from late 90ies, so I've been years past this. I am (as mentioned) also no developer. I can only repeat myself - I've been at this for the whole day and only the -3 add-on to the "M = L+R" formula comes even close to what we see both on mvMeter and VUMT2 in "Mid Mode". mvMeter's Stereo Mode is nonsensical. If we talk "pure math" that is (there is so much more involved).


So... if I go by "(L+R)/2" (plain MID math, then divided by applied channel amount) alone, I get the very same results as "((L VU max) + (R VU max))/2" (the TBPro math). And that can't be correct.

It is also to note that a signal that is being summed, can get and additional signal boost of up to 3dB (average, depending on the stereo width, what instruments were used, etc - max peak can be way higher), so we need to take that into consideration. It's not just for "reverting a matrix".

In fact... take the values I've posted, run them through the EXCEL sheet I provided (the lower section, doing math for the MID incl. 3dB compensation and division by two channels) the values do line up - with a certain "offset" (depending on how the needle responded, etc - something we can't CONFIRM with plain math, unless we're the developers that can solve the "riddle of the offsets" due to the needle inertia).


Let's take a look at the AC/DC track:
VUMT2 in "Duo Needle View" mode shows -2,7VU L and -2,0VU R

mvMeter in Stereo Mode (single needle) shows: -2,7VU
VUMT2 in "Mono Sum mode" (single needle) shows -3,1VU

So who is ultimately correct?



By the TBProAudio math (and only by the math!) we'd be at this:

Code: Select all

Lvl_stereo = (VU(Inp_l) + VU(Inp_r))/2.0;
Lvl_stereo = (VU(-2,7) + VU(-2,0))/2;

Let's simplify
Lvl_stereo = (-2,7 + -2,0) /2;
Lvl_stereo = -4,7 / 2;
Lvl_stereo  = -2,35;

The needle max value therefore shows -2,35 VU

By not using the -3dB offset (which should result in a higher readout to begin with), we're at

Code: Select all

Lvl_stereo = (L + R)/2.0;
Lvl_stereo = (-2,7 + -2,0)/2;

Let's simplify
Lvl_stereo = (-2,7 -2,0) /2; (because + and - equals -)
Lvl_stereo = -4,7 / 2;
Lvl_stereo  = -2,35;

The needle max value therefore shows -2,35 VU

THAT MEANS THAT BOTH THE MID CALC AND THE TBPROAUDIO CALC IS THE SAME!
And that can't be right - IMO!

With the 3dB offset (again, plain math only)

Code: Select all

Lvl_mid = ((L+R)-3) /2;
Lvl_mid = ((-2,7 -2,0)-3) /2; (because + and - equals -)

Solved
Lvl_mid = (-4,7 -3) /2;
Lvl_mid = -7,7 /2;
Lvl_mid = -3,85;


The needle max value therefore shows -3,85 VU

The "alleged" MID calculation by the dev, also the M = L+R one without the 3dB compensation results in -2,35. Which in turn results in the following offsets:
mvMeter (-2,7VU) = 0,35dB
VUMT2 (-3,1VU) = 0,65dB

If we go by the MID calculation with the 3dB compensation, and ignore the possible "integer points" (everything post the third decimal)
mvMeter (-2,7VU) = 1,15dB
VUMT2 (-3,1VU) = 0,75dB

Offsets on a needle VU up to 1dB are considered "normal" (we talk about electro mechanics, not digital calculation), offsets of 0,5dB are better. So far, we've also not taken the 1% overshot into account either.


So who is ultimately correct?
Who interprets the measurements right?





Yes, you pretty much read EVERYWHERE that a "MID" signal is basically "M = L+R"

In fact:
https://de.wikipedia.org/wiki/Mid/Side-Stereofonie
http://www.audio-times.com/2015/03/10/m ... rocessing/
https://www.gearslutz.com/board/so-much ... oding.html
https://theproaudiofiles.com/mid-side-processing/

The latter one has an interesting paragraph adding to this topic as well:

When the left and right channels are added together, the amplitude of a center signal is 1.414*original mono amplitude (~3 dB louder).

Therefore, the “mid” channel in MS decomposition can have a higher amplitude for center information than side information.

I did "try" this earlier today with "summing" stereo signals to mono (mix to mono, see test files). I can indeed confirm a strong signal boost, depending on how "wide" the stereo field was (quote from earlier post, stereo signals, summed/mixed to mono):

Code: Select all

Acoustic Guitar (Original Stereo, Narrow Stereo field, Bass heavy)
mvMeter 0,2VU / 4,9PPM/BBC (Stereo Mode)
VUMT2 0,1VU / 5,06PPM/BBC (Solo/M view)
VUMT2 0,3VU (0,3L/0,1R) / 5,09PPM/BBC (5,09L/5,03R) (Duo/LR view)

Acoustic Guitar (Summed to Mono, Bass heavy)
mvMeter +6,1VU / 6,4PPM/BCC (Stereo Mode)
mvMeter +6,1VU / 6,4PPM/BBC (Mono Mode)
VUMT2 +6,1VU / 6,56PPM/BBC

Virtual Orchestra Mix (Wide Stereo!)
mvMeter -1,3VU / 5,2PPM/BBC (Stereo Mode)
mvMeter +0,6VU / 6,1PPM/BBC (Mono Mode L)
mvMeter -1,3VU / 5,2PPM/BBC (Mono Mode R)
VUMT2 -2,9VU / 5,10PPM/BBC (Solo/M view)
VUMT2 +0,6VU (0,6L/-1,3R) / 6,07PPM/BBC (6,07L/5,27R) (Duo/LR view)

Virtual Orchestra Mix (Summed to mono)
mvMeter +3,1VU / 6,5PPM/BBC (Stereo Mode)
VUMT2 +3,1VU / 6,61PPM/BBC

And just for sh*ts and giggles (not to mention "I don't have anything better to do" /sarcasm), I ran these test files again (*sigh - and again...*) and compared VUMT2 with mvMeter in "Mid" mode and "Stereo Mode"

• The Virtual Orchestra in Stereo (not summed), if we go by plain math, has a MID VU value of -1,85dB
• The measurement in Mid was -2,9 on both VUMT2 and mvMeter, and -1,3 on mvMeter in Stereo Mode
• If I "actually sum" the signal in Wavelab, I suddenly jump to +3,1VU. The signal did make a +3,5dB "jump" from -4,1dBFS to -0,6dBFS while summing (according to Wavelab's analysis), however only 1,5dB loudness wise (average signal strength)

• The Acoustic Guitar in stereo, would go by (again, plain math) by MID VU value of -1,3
• The measurement in Mid on the stereo file was +0,1 on both VUMT2 and mvMeter, and +0,2 on mvMeter in stereo Mode. So here we talk an offset of 1,2dB
• If I actually "sum" the signal in Wavelab, I have a sudden jump up +6,1VU, the signal did make a +5,8dB "jump" from -8,7dBFS to -2,89, however loudness wise (average signal strength) only 3dB.

You see where this is going?


"Stereo Signals" on a single-needle meter can (and do!) have the most different readouts. It's not just math, it's needle behavior, how the "internal summing / signal correction" is handled, how wide/narrow was the stereo field, etc.

Summed up however, it can be said that Stereo Signals on a Forced Single Needle Meter are not recommended/not considered optimal.

This is why there was so much attention put into the "summing" of MLk (Monetary Loudness)/SLk (Short Term Loudness) in ITU-R BS.1770-x loudness meters - to work against the flaws I just pointed out.


In fact, depending on how "Linear" a metering tool is while signal analysis, a "bass intensive" signal can have different results on different average signal strength meters.

This whole topic is more than just "a" (singular) can of worms - and we need to(!) understand what meter XYZ is actually capable off, what it was/is made for. We can't just make up certain features and hope for the best. This is why threads like this one exist.





Sorry, I have to hit the breaks now... been too long at this yet again...
*a planned 2 hour test, turned into another 10 hour testing and documentation day...* *sigh*

"Then why are you still in this thread?!"
Because the topic metering interests me, and I do to try to educate current and interested/possible future users with such posts...

I for one am glad you are still in this Compyfox, even if in sidelines now. Good information, kudos for extensive testing too.

Your kind words are appreciated, thanks.

Though maybe I should drop a PayPal Donation button or something to sponsor me a hot cup of cocoa ever so often... to revitalize my sanity.

[crimsonmerry]

@compyfox

At a first sight i'm not really thrilled with the strong way you write your thoughts about the GUI design, but since i'm fairly new to this forum, it took me some time to hold myself and see only the good things behind your posts.

To respect all of your extensive tests for this plugin, and for your generosity to educate us all, let me write a few lines, responding to your critics regarding the GUI.

CS-3301

The design were inspired from some gears i'm using in the studio. Portico II, Vintech pre, and some old Amek pre/eq and stereo comp/lim.

And for the layout, i did my observation from lots of channel strip plugin images available, and from there i took and avoid what i think is good or bad for the button and knob placement, size, spacing, shapes, drawing style, etc.

That's pretty much it.

My respond for mvMeter is not much different, i use all good things i can find from my references. But i don't think using the current color scheme is "unique" at all. White skin for those who prefers bright, black skin for those who prefers dark, and the classic skin is there to imitate those vintage VU meters for those who prefers oldskool.

According to my image references, VUMT2 is not the only (or perhaps the first) VU meter in a software who has value readouts on the bottom left nor right. But yes i like it more when they have a "frame" around them, it matches the top left and right buttons. So why not.

Coincidence? Inspired? Copy? Steal? Please pick whichever suits your mind best. I don't want to make any of the features less comfortable to use just because there's one, a few, or hundreds of plugins that already uses the idea.

[cprompt]

For anyone who doubts that compyfox knows what he's talking about:

A huge thanks to Roland Löhlbach (Compyfox), who has provided some additional metering presets with in-depth documentation for VUMT & VUMT deluxe.

Congrats!

[planetearth]

Your kind words are appreciated, thanks.

Though maybe I should drop a PayPal Donation button or something to sponsor me a hot cup of cocoa ever so often... to revitalize my sanity.

Didn't I suggest something similar about a week ago?

I'm also glad you're still in this, and I hope everything can stay civil for a while longer. I think we're all learning something here, and we'll all be better at judging how "good" a metering plug-in is, as well as determining which features we really need.

I believe this is making the TBProAudio mvMeter a better plug-in, too. I've found TBProAudio one of the better, more responsive developers out there (at least in my interaction with them), and I hope this turns out to be a win/win for everyone (including Compyfox and his Starbucks gift card.)

Steve

[Compyfox]

@compyfox

At a first sight i'm not really thrilled with the strong way you write your thoughts about the GUI design, but since i'm fairly new to this forum, it took me some time to hold myself and see only the good things behind your posts.

So there are more people behind the "we" and the "Legal Notice"... Welcome to the conversation indeed. I really wonder why there is a problem to make that public if you folks are proud of your work? Yet only one person is listed, and the official representative of the company is... well... non-communicative or "understanding".

I just don't get it...

Coincidence? Inspired? Copy? Steal? Please pick whichever suits your mind best. I don't want to make any of the features less comfortable to use just because there's one, a few, or hundreds of plugins that already uses the idea.

The problem is, that it's just too close to the competition (IMO and all that). Sometimes even blatant 1:1. This is what was really annoying me. I'm not saying the design is "bad" (in fact: mvMeter 100% = VUMT2 200% - kind of enjoying that), and yes - there are many, many similar tools out there. But they all look different, unique.

But in this case (which was my main UI criticism), the UI looks nearly the same, the manual was copied (especially after my criticism - although really, really simplified), the feature set is nearly identical... so what should I think about that?


Which is the main reason why I asked (and I'm now doing that again, 4th of 5th time at this point?):
"How long was the development time?"

I wanted to rule out that this was not(!) a copy-cat, but mere coincidence. But what happened after my first post, clearly steered towards the direction "copy, we don't give 2-flarts, and we won't admit it either". That's all I've been addressing.


That the plugin was (and still somewhat is) faulty on top of that, and that "some" of the usability could be improved (see "Stereo Mode") is a completely different issue.





@Eleventh:
I've had enough of your insults and non-productive/non-contributing commentary - I reported your posts



@cprompt:
I think we had enough "cross promotion" in the thread at this point. Yes, I did create these presets. Yes, I did invest a lot of time into them (and the bundled documentation). Thanks for the positive words - but everything about that belongs in the VUMT thread.

Didn't I suggest something similar about a week ago?

Payment of any form? Yes... sometimes I really wish I would be paid.

But no, I don't have a Starbucks gift card (does that exist for Germany?), I don't even have one "nearby" my area (they're all 5+km away, in midtown Berlin). Though I am considering something like Patreon as of late for all my side projects... different topic, different time.

I'm also glad you're still in this, and I hope everything can stay civil for a while longer. I think we're all learning something here, and we'll all be better at judging how "good" a metering plug-in is, as well as determining which features we really need.

Glad that (some of) my posts help on the educational side of things.

I believe this is making the TBProAudio mvMeter a better plug-in, too. I've found TBProAudio one of the better, more responsive developers out there (at least in my interaction with them), and I hope this turns out to be a win/win for everyone

I guess, this is why I bothered to begin with... to see if this plugin is any good, if it's accurate, if it might be adding to learning factors, etc. But somehow I did get the impression, that the developer just didn't want that if I look at the communication alone. Then again, sometimes you have to beat some eggs to make an omelette - I guess.

This whole "thing" sparked a discussion... whether or not some might like the direction, whether it's good or bad. So the plugin is being talked about - can't get any better promo than this I'd say.

So the best thing the developer can do at this point, is grab all of that, turn it on it's head and make it "usable" for him. The ball is in the game...

[Hink]

report duly noted and I echo planetearth's feelings...I hope this stays civil but not just for a while longer

[TB-ProAudio]

Update to V1.0.3:
- 2 new themes: retro and deluxe
- RTAS/AAX/VST3: GUI fully restored after project load
- WaveLab VST3: resize crash fixed

[murnau]

You know, you could have just reported the issues like a normal grown-up person, instead of ranting like these generous people who made the plugin committed some kind of crime against you.

But you didn't, you acted like the pompous little asshole you are.

No, it's okay the way it is because this is a public forum and 2nd some can learn something while reading the detailed Compyfox posts (The dev should send him some money for his awesome work on this).

Furthermore: You seem to think that grown-up persons can insult others. Makes sense..

[planetearth]

Update to V1.0.3:
- 2 new themes: retro and deluxe
- RTAS/AAX/VST3: GUI fully restored after project load
- WaveLab VST3: resize crash fixed

Thank you! Updating it now!

Steve

[planetearth]

Update to V1.0.3:
- 2 new themes: retro and deluxe
- RTAS/AAX/VST3: GUI fully restored after project load
- WaveLab VST3: resize crash fixed

You say this is an "update to V1.0.3", but the file on your website is "v1.2.6".

Just wanted to point that out, in case others are confused.

Steve