The No.1 Website for Pro Audio
i need a correct RMS meter
Old 4th May 2009
  #1
Gear Nut
 

i need a correct RMS meter

I have been using Roger Nichols/Waves/and the the meters in wavelab - i have been told that they are all out by 3db is this information correct, so all my mixs are down 3db - and why would they be out by 3db?
thanks
jim
Old 4th May 2009
  #2
Mastering
 

Quote:
Originally Posted by asylum View Post
I have been using Roger Nichols/Waves/and the the meters in wavelab - i have been told that they are all out by 3db is this information correct, so all my mixs are down 3db - and why would they be out by 3db?
thanks
jim
There is an official AES and IEC-supported standard for RMS meter calibration. All RMS meters should support that standard or else they are wrong. I won't get into the technical and scientific explanations for why the IEC 61606:1997 standard is the correct one to use, as all you want to do is find a meter that's correct.

There are a number of terminologies that people try to use to describe a meter which is correctly-calibrated. Some people speak about "sine wave-based" and "square-wave based" calibration, but I find that confusing. Here's how to test if the meter is correct:

1) Send in a sine wave whose peak level is x. The RMS level in dBFS should read the SAME as the peak. If it reads 3 dB lower, it does not meet the standard.

2) Download the -20 dBFS RMS pink noise signal from our site. Digital Domain-Downloads

If this signal reads (on the average) -20 dBFS on your RMS meter, then you can feel reasonably comfortable the meter is accurate and conforms to the standard. This test is not 100% correct because it does not test the actual RMS algorithm, it only tests how the meter behaves with two different test signals, but at least this test should show that the meter reads correctly with these two types of signals.

If the meter proves to be off, you will be doing the folks a service by citing the IEC standard and telling them that it does not meet the standard. Tell the to get in touch with me if they want to argue the science behind it, and I'll explain to them the historical precedents and the logic behind it.

As for the Roger Nichols, I understand there was a controversy over who owns the company. RN Digital (the new company) produces a meter called "Inspector XL" and the K-System meters in there meet the standard. Wavelab since version 6 is correct. Previous versions were 3 dB off.

I hope this helps,


Bob
Old 4th May 2009
  #3
Gear Maniac
 
fader8's Avatar
 

Quote:
Originally Posted by bob katz View Post
1) Send in a sine wave whose peak level is x. The RMS level in dBFS should read the SAME as the peak. If it reads 3 dB lower, it does not meet the standard.
Hi Bob,
OK, I'm going to call you out on this one. Are you saying that this standard no longer considers rms voltage to be .707 peak for a pure sine wave?

Could you perhaps quote the relevant passage in that standard?
Old 4th May 2009
  #4
Mastering
 

Quote:
Originally Posted by fader8 View Post
Hi Bob,
OK, I'm going to call you out on this one. Are you saying that this standard no longer considers rms voltage to be .707 peak for a pure sine wave?

Could you perhaps quote the relevant passage in that standard?

Now that I've opened my big mouth you want me to explain, too :-).

OK, here goes. For over 70 years, ALL averaging (including RMS) meters and peak meters that read in decibels have been calibrated with a sine wave. You can find some classic analog style PEAK-responding dB meters that read 0 dBu with an 0.775 V RMS sine wave, whose peak level is 1.09 volts, also defined as 0 dBu peak. This was done in order to find a correlation point and with a test signal that everyone had available.

Now when digital technology was developed, many of the technical engineers who designed RMS meters ignored or were not aware of this precedent and decided that since the RMS value of a sine wave is 0.707 of its peak, (or 3 dB lower than its peak), then the decibel level should also be set 3 dB lower. But this is IRRELEVANT to how the decibel reference level can be set, since that's all it is: a REFERENCE point. And redefining the decibel reference this way would contradict 70 years of precedent.

A decibel is always to be considered a REFERENCE. And the reference for IEC 61606 defines 0 dB for ALL TYPES of detectors as the same level as the maximum peak level when presented with a sine wave. In addition, it states that the decibel level should read the same for a sine wave regardless of the type of detector, RMS, median averaging, peak, or some other unusual detector. This allows for consistency and when you work at it, ease of use. It defines an alignment point.

Dorrough meters have always been set up this way, and mastering engineers have aligned their VU meters as well, with a sine wave, and defined compared to the level seen by the peak meter. The Dorrough VU reads the same as the peak level when a sine wave is put in.

Or think about it this way. The right triangle law says that the RMS value of the sine wave is 0.707 of its peak. But nothing says that you cannot define both of those voltages as 0 dB for reference purposes.
But if you want to reject 70 years of precedent, you're welcome to petition the IEC and have them change the standard!

Hope this helps,


Bob
Old 4th May 2009
  #5
Lives for gear
 
Cellotron's Avatar
 

Verified Member
For VST (PC & Mac) or SAWStudio capable workstations the Sonoris Meter makes an excellent choice - it features "correct" RMS metering, plus LeqA metering (which can actually give a more accurate idea of perceived loudness than RMS) , along K-system metering and standard peak metering in dBFs.

Meter - Sonoris Audio Engineering

Best regards,
Steve Berson
Old 5th May 2009
  #6
Gear Maniac
 
fader8's Avatar
 

Quote:
Originally Posted by bob katz View Post
NThe right triangle law says that the RMS value of the sine wave is 0.707 of its peak. But nothing says that you cannot define both of those voltages as 0 dB for reference purposes.
Absolutely agreed! Believe me, having spent a few years in noise and vibration measurement, sliding decibel references are old hat for me. What triggered my response was that I read your statement with the implication in mind that you were suggesting the "same" meter with a fixed 0dB reference should read the same with RMS or peak.

Semantics. It's the mind killer.

Interestingly, the term RMS applied to a meter that is indexed in dB is a kind of misnomer in itself, given that dB in that context expresses a power ratio between two voltages. So it "implies" an RMS power measurement, which isn't a real term. Oh what a wicked web we weave . . . .
Old 5th May 2009
  #7
Lives for gear
 
Coyoteous's Avatar
 

Inspector XL reading high with BK pink noise?
Attached Thumbnails
i need a correct RMS meter-picture-1.jpg  
Old 5th May 2009
  #8
Lives for gear
 
Jesse Graffam's Avatar
 

No settings for RMS average on Inspector XL? I would think that's pretty important to know what those actually are set to, for a comparison.

What does the "Peak-RMS Parity" button do?

Old 5th May 2009
  #9
Gear Maniac
 
fader8's Avatar
 

Quote:
Originally Posted by uncajesse View Post
No settings for RMS average on Inspector XL? I would think that's pretty important to know what those actually are set to, for a comparison.

What does the "Peak-RMS Parity" button do?
It does what we've been discussing here, ie no difference in calibration with peak and RMS if you're looking at a sine wave.
Old 5th May 2009
  #10
Lives for gear
 
Coyoteous's Avatar
 

Quote:
Originally Posted by uncajesse View Post
No settings for RMS average on Inspector XL? I would think that's pretty important to know what those actually are set to, for a comparison.

What does the "Peak-RMS Parity" button do?

Okay, is that the difference here... we're seeing RMS max (+.59/+.64)?

Notice also that this is K-20 and only a few out of many settings can be adjusted.

I like XL. I use it a lot on my laptops, but it doesn't jibe with my Dorroughs (plural). I used to do a lot of "speed mastering" for a plant... sometimes 1 hr. programs in less than an hour. Don't worry, these were/are generally be very poor recordings that needed some help.

Anyway, I have 2 stereo analog Dorroughs that go to +14. I like to calibrate them with an attenuator for my target crest factor, so that my peak goes to say +11, and then I shoot for an RMS max of 0 or so on the meters. On mine there are red LED's for -1, 0 and +1.

This is just for roughing in levels against a limiter, which I'm watching for max GR while I jump around on the timeline... mainly visually dense areas. Yes, I can do this without even hearing the material, but that's another story. This would just get me some basic settings that I would, of course, go back and refine.

So, back to the meters: I occasionally do similar work on a laptop now, though with out the time constraints... other than the fact that it doesn't pay much and I want to hear it as little as possible. I use XL in K-14 and shoot for about +3, but when I check playback on my old systems it's about a dB high.

Okay, that was a lot of typing for not much. Don't try this at home, and believe me, most of these jobs don't warrant the effort I put into them, but sometimes a few "right" adjustments can make a big difference and a happy customer.
Old 5th May 2009
  #11
Mastering
 

Quote:
Originally Posted by Coyoteous View Post
Inspector XL reading high with BK pink noise?
That does look rather suspicious. As I understand it (and Roger should correct me if I'm wrong), Roger Nichols no longer has anything to do with these meters. The company now producing them is RN Digital. I have them installed (courtesty of the company and I'll check them with the pink noise and recheck the pink noise against a couple of RMS meters and let you know.

The averaging time is really only necessary to make it easier to read the meter, but also requires that you watch it for a longer time. I usually have no trouble determining the level within 0.1 dB within about 10 seconds with an averaging time of about 3 seconds. To generate the pink noise file on our site I used Spectrafoo's RMS meter (which I trust), set the averaging time to very long and waited and waited and tweaked and tweaked until I was 100% certain the level fluctuated equal amounts above and below -20 dBFS RMS. It can't be more than 0.1 dB off and it is probably less than 0.1 dB off. After checking Holman's narrow-band pink noise using the same method we agree within less than 0.1 dB and I don't even know what RMS meter Holman uses, so we're in good company.

BK
Old 5th May 2009
  #12
Lives for gear
 
Jesse Graffam's Avatar
 

Quote:
Originally Posted by fader8 View Post
It does what we've been discussing here, ie no difference in calibration with peak and RMS if you're looking at a sine wave.
That's kind of what I thought. So... why isn't it engaged Coyoteous?


Quote:
Originally Posted by bob katz View Post
I'll check them with the pink noise and recheck the pink noise against a couple of RMS meters and let you know.
Coolio. Btw, the LKFS K-System + True Peak meter is still on the way. We're just making sure it's purrrrfect before anyone sees it, namely yourself before anyone else. Having 20 things on the stove at once doesn't help.

We have the CPU use so crazy low now, despite the eye candy. I ran it for 9 hours a few days ago, and the CPU time was 1 second. !!

Last edited by Jesse Graffam; 5th May 2009 at 03:05 PM.. Reason: leapfrog reply to Bob
Old 5th May 2009
  #13
Mastering
 

Quote:
Originally Posted by Coyoteous View Post
Okay, is that the difference here... we're seeing RMS max (+.59/+.64)?

Notice also that this is K-20 and only a few out of many settings can be adjusted.

I like XL. I use it a lot on my laptops, but it doesn't jibe with my Dorroughs (plural).
On pink noise there is a good reason why it doesn't jibe with the Dorroughs. The Dorroughs use simple averaging and that's why they're as much as 1 dB off on the pink noise. HOWEVER, for program material, either RMS or standard averaging is more than adequate to read the VU.

When the IEC short term ballistics becomes a standard we'll see things become a little more standardized in terms of the loudness measurement.

In the old days, Hewlett Packard averaging meters used to say RMS but were really averaging as illustrated in the small print on the meter face, and in order to distinguish themselves, TRUE RMS meters read TRUE RMS on their face! Maybe we should start saying "True RMS" again because the term RMS has been bandied about for a long time and many people don't understand that RMS involves more calculation than simple averaging.
Old 5th May 2009
  #14
Lives for gear
 
Jesse Graffam's Avatar
 

Quote:
Originally Posted by bob katz View Post
Maybe we should start saying "True RMS" again because the term RMS has been bandied about for a long time and many people don't understand that RMS involves more calculation than simple averaging.
I agree, along with True Peak. That almost warrants a new level of K-System as far as metering goes. True-K ? heh
Old 5th May 2009
  #15
Mastering
 

Quote:
Originally Posted by uncajesse View Post
I agree, along with True Peak. That almost warrants a new level of K-System as far as metering goes. True-K ? heh
Yeah... I don't have control over any of the manufacturers, just an advisory role. This is a non-profit venture! I do have plans for future K-System meters, they will employ the new IEC weighting and the peak section will be almost non-existent, just a warning light. This is because we want to discourage engineers from reading the peak meters, they are deceiving and largely irrelevant to the loudness. The abiliyt to Peak-normalize is what resulted from the invention of the compact disc, and it's what got us into this mess.

BK
Old 5th May 2009
  #16
Lives for gear
 
Jesse Graffam's Avatar
 

Quote:
Originally Posted by bob katz View Post
Yeah... I don't have control over any of the manufacturers, just an advisory role. This is a non-profit venture! I do have plans for future K-System meters, they will employ the new IEC weighting and the peak section will be almost non-existent, just a warning light. This is because we want to discourage engineers from reading the peak meters, they are deceiving and largely irrelevant to the loudness. The abiliyt to Peak-normalize is what resulted from the invention of the compact disc, and it's what got us into this mess.

BK

For mastering I do see where you're coming from, but for mixing - i think it is important to see the peaks all the time, so one can tell how slammed something is (crest) regardless of the absolute peak. That's often where stuff gets ruined in a mix, before it ever hits -0dBfs within the track, or bus.

Last edited by Jesse Graffam; 5th May 2009 at 03:31 PM.. Reason: idea
Old 5th May 2009
  #17
Mastering
 

Quote:
Originally Posted by uncajesse View Post
For mastering I do see where you're coming from, but for mixing - i think it is important to see the peaks all the time, so one can tell how slammed something is (crest) regardless of the absolute peak. That's often where stuff gets ruined in a mix, before it ever hits -0dBfs within the track, or bus.
I see your point.

It is also true that if you work at a 0 dB monitor gain and take away the mentality that "I have to protect the meters", that also removes some of the mentality of putting in compressors just to protect the meters and hopefully puts you in the position of mixing just to the ear. But Tom (Minister) in another post betted that he could make a squashed K-20 recording with no trouble at all, and he's right. It's just that the mix engineers who have been working K-20 and sending me mixes of everything from heavy metal to soft ballads have not been squashing. There seems to be a change of mindset there.

But I hear you, having the peak level turned on during mixing can possibly be a tool to aid the ear. I say "possibly".... If you learn how to mix openly by ear first and have occasionally glanced at the peak meter and determined that for the last 10 sessions your crest factor was at least 14 dB and they sounded great, then I guess you are justified at using the peak meter as a guide to see if your 11th mix will sound as "open". But meter reading per se is a dangerous precedent, you have to work by ear.

BK
Old 5th May 2009
  #18
Gear Maniac
 
fader8's Avatar
 

Quote:
Originally Posted by bob katz View Post
But meter reading per se is a dangerous precedent, you have to work by ear.
It always seems to come back to that, doesn't it? Interesting how we continue with our repeated and vain attempts with technology to measure something that our ears can tell us so clearly.

(Soapbox Mode=On)

I'm of a mind to advocate the removal of the term RMS from our collective audio vocabulary. I believe its relevance is questionable. Bob, earlier you mentioned historical precedent and I think that it bears out my conclusion. As I'm sure you know, RMS was initially developed as a method to calculate the AC voltage required that would be the equivalent work performed by its DC counterpart. Specifically, regarding heat.

So, a DC voltage of .707 is required to heat a purely resistive load to the same temperature it would reach if we had a one volt sinusoidal signal. While "True RMS" solves the problem for heating things with complex periodic waveforms, I believe it still doesn't tell us anything particularly meaningful about audio, unless of course I wanted to know how loud I need my living room stereo to be before it starts melting the foam rubber in my sofa.

I think no matter what we do with RMS, even a true RMS measurement, we're fundamentally relating the measurement process to heat, which ignores most of the time vs. energy aspects of musical sound that we use as cues to perceive loudness. So in a sense, the only relevant meter is the peak meter, specifically the intersample peak meter, since it directly relates to a real world value where nonlinearity can occur digitally in a real sound system.

So if I'm going to lobby the IEC about anything, it would be to remove the concept of RMS from any standard dealing with the measurement of audio levels.

(Soapbox Mode=Off)

Cheers,
Randall
Old 5th May 2009
  #19
Gear Nut
 

Well, I guess I'll chime in

To me, the best RMS Meter is the traditional VU meter. To make this workable it needed to be driven through an amplifier and a calibrated attenuator, adhering to the 600 ohm source that it was designed to work with many decades ago.

The readings on the meter have relevance to perceived loudness, especially loudness change, but it does not take into account the ear's sensitivity to different frequencies at different loudness levels (Fletcher-Munson )

For what it's worth - this is my opinion.

Bob
Old 6th May 2009
  #20
Lives for gear
 
Coyoteous's Avatar
 

Quote:
Originally Posted by bob katz View Post
That does look rather suspicious. As I understand it (and Roger should correct me if I'm wrong), Roger Nichols no longer has anything to do with these meters. The company now producing them is RN Digital. I have them installed (courtesty of the company and I'll check them with the pink noise and recheck the pink noise against a couple of RMS meters and let you know.
From what I've been able to tell about all of this, Roger Nichols really never had much to do with the development of Inspector XL. I bought Inspector XL from its developer, Elemental Audio Systems for $89, when it first came out. I paid another $20 to cross-grade to the Roger Nichols Digital version, which then became the RN Digital version when Mr. Nichols and Reinhold Probst had their falling out.

It has not changed functionally AFAICT through any of these company changes. I believe Reinhold Probst is still the principal behind RN Digital, though their website looks like a Waves clone these days. I also believe that Reinhold Probst is the president of: Dilo | Home which seems to specialize in Sulfur hexafluoride (SF6) detection and recovery.

I assume that the anonymous party or parties behind Elemental Audio Systems is/are still the real developer for RN Digital... anybody know this this might be? I've wondered for years, and I personally think they do very good work. Many thought that RN had developed the original EAS plug-ins, though others thought that he was only acting as if he did... 'nuff said?

That said, there are some other odd and/or inaccurate behaviors in IXL, mostly when looking at very low levels, noise floors, etc. For the most part the components of the suite seem to work as advertised and are handy for all but scientific work, IMO (because of these problems)... whew, are y'all still with me?

Quote:
Originally Posted by uncajesse View Post
That's kind of what I thought. So... why isn't it engaged Coyoteous?
The IXL Level meter (as shown... there is also a horizontal version that shows less settings) has 6 different formats: Log, Linear, K-20/RMS, K-14/RMS, K-12/RMS and Digital PPM. Only Log and Linear have all the settings available. So... that is why parity is not shown as engaged, even though it is offset.

Quote:
Originally Posted by bob katz View Post
On pink noise there is a good reason why it doesn't jibe with the Dorroughs. The Dorroughs use simple averaging and that's why they're as much as 1 dB off on the pink noise. HOWEVER, for program material, either RMS or standard averaging is more than adequate to read the VU.
Thanks, I've wondered about this for a long time. Like Bob Dennis, I actually prefer good old VU meters. I've been digitizing audio since we used to do it with pencil and paper... well, pre-DAT anyway. I got used to a combination of digital peak meters and VU early on... Dorrough after that.
Old 6th May 2009
  #21
Mastering
 

Quote:
Originally Posted by Superdisc View Post
Well, I guess I'll chime in

To me, the best RMS Meter is the traditional VU meter. To make this workable it needed to be driven through an amplifier and a calibrated attenuator, adhering to the 600 ohm source that it was designed to work with many decades ago.

The readings on the meter have relevance to perceived loudness, especially loudness change, but it does not take into account the ear's sensitivity to different frequencies at different loudness levels (Fletcher-Munson )

For what it's worth - this is my opinion.

Bob
Well, as you know, a VU has never been "true RMS", for what that's worth. It's a simple average and if you go back far enough, its ballistics depend on a copper-oxide rectifier which is no longer made, so all current VU meters are approximations of the original! Regardless, RMS is most suitable for calibrating with pink noise. After that it's not an important averaging method for looking at program material.

I've been experimenting with some competing meters on various program material, with varying time constants and loudness weightings. Bob Orban has a beta meter which includes a traditional VU, the new IEC LU standard, CBS Loudness and one other. When you see this demo, you begin to see the VU as dancing around much too much and quite distracting. Syllabic response is initially attractive, especially for those who are mixing speech, but I believe that a more modern-day short-term loudness meter will have a slower time constant than the VU because it's easy to read. And when you watch it while listening, it more truly reflects the ear's reactions because the ear's loudness integration time is not as fast as VU's. That's one reason why the K-System meters have a 600 ms time constant as opposed to the VU meter's 300 ms.

Bob, if you are interested in trying out the Orban comparison to see if you agree with me, write to me privately, at bobkatz[atsign]digido.com.
Old 6th May 2009
  #22
Gear Nut
 

Quote:
Originally Posted by Coyoteous View Post

Thanks, I've wondered about this for a long time. Like Bob Dennis, I actually prefer good old VU meters. I've been digitizing audio since we used to do it with pencil and paper... well, pre-DAT anyway. I got used to a combination of digital peak meters and VU early on... Dorrough after that.
It's a bit hard for me to understand all the hubabaloo generated around "true" RMS. The bottom line is you want a meter that tells you something about how something will sound from a loudness perspective. This is used to "focus" your hearing because your ears only will hear what you're paying attention to (see how soft the wife sounds in the 3rd period??).

The marvous thing about the VU Meter is the balastics. You can "see" the various instruments causing the meter movement. When you see the vocal causing peaks that are rising to zero and then see a chorus where the vocals cause swells to +2, the vocal sounds louder by a certain amount - the same loudness difference as the vocal peaking to -4 and swelling to "-2"

You could easily see how "compressed" or "uncompressed" the audio was by looking at the meter. In many cases with "meters" in software a very compressed master doesn't look that much different than a lightly compressed master. I go "yuk," why am I even looking at this??

The balastics of the VU meter has it rise from a standstill in 300 ms. to a "0" display with a maximum overshoot of 1 dB. With a lot of digital meters you can get that 300 ms rise time, but I haven't seen any that do a good job of duplicating the "fall" time that happens with a real VU meter caused by a light meter mechasim working against a spring. That "overshoot" part of the spec, done mechanically, largely determines the "fall" time.

I've more or less stopped looking for a better meter. I do know that I hate the way modern software is constantly trying to display the whole dynamic range in one scale, or the way that they try to "linearize" the dB scale so the difference between 0 dB and -2 dB is the same distance on the scale as - 4 dB and -6 dB. Why do this? Does it help you judge loudness or even loudness difference between "sections" of the piece? The answer is "Not really"

With A VU meter driven by an amplifier and with a calibrated attenuator, you can "turn up" the lower sections to see where they read and how the instruments are influncing the meter's movement. The VU meter has a maximum "scale" of about 23-24 dB, and with only about 12 dB being that useful, but with the amp/attenuator this can expand another 20 dB.

So, that's my take on all of this (my 2c).

Bob
Old 6th May 2009
  #23
Gear Nut
 

Quote:
Originally Posted by bob katz View Post
I've been experimenting with some competing meters on various program material, with varying time constants and loudness weightings. Bob Orban has a beta meter which includes a traditional VU, the new IEC LU standard, CBS Loudness and one other. When you see this demo, you begin to see the VU as dancing around much too much and quite distracting. Syllabic response is initially attractive, especially for those who are mixing speech, but I believe that a more modern-day short-term loudness meter will have a slower time constant than the VU because it's easy to read. And when you watch it while listening, it more truly reflects the ear's reactions because the ear's loudness integration time is not as fast as VU's. That's one reason why the K-System meters have a 600 ms time constant as opposed to the VU meter's 300 ms.

Bob, if you are interested in trying out the Orban comparison to see if you agree with me, write to me privately, at bobkatz[atsign]digido.com.
Hmm...

For some reason your responce post came up before my original post. .

The "bouncing around" is prercisely what I like about the VU.

To me "Loudness" is best judged by ear - the VU meter gives me information about loudness change within the tune/cut - a key to applying intelligent volume envelopes, and it's just a guide. Be that as it may, I'm still open to something else being better, I just haven't found it.

Thanks for your views on this.
Old 6th May 2009
  #24
Mastering
 

Quote:
Originally Posted by Superdisc View Post
Hmm...

The "bouncing around" is prercisely what I like about the VU.
Yes, it's what makes it initially attractive because it "follows the words" , but it's not an accurate representation of what the ear calls "louder" or "softer". Our integration time is not 300 ms long and the syllabic response, however "intuitively" attractive, is actually wrong. You begin to think something is instantaneously louder when the meter jumps up, when it is not as much louder as it looks.

We trained engineers who were brought up on the VU have forgotten how much TRAINING was necessary to learn how to interpret the VU and why. If you sit someone in front of a VU for the first time, for many reasons that I shouldn't need to get into at this time, they will tend to overreact to the changes in the meter, they will tend to follow their eyes instead of their ears. Look at all the rationalizations that we have to teach him about what's "tolerable" for short periods of time, etc. Rules which should be broken at certain times and why....

I'm not saying that we didn't get excellent results when using a VU, arguably more consistent results than those using a PPM, but these were trained results from much experience. Today's engineers deserve a more accurate meter which needs even less training.

BK
Old 6th May 2009
  #25
Gear Nut
 

Quote:
Originally Posted by bob katz View Post
Yes, it's what makes it initially attractive because it "follows the words" ,
Well I'm going to risk being really irritating.

You've said several times that it "follows the voice", but I've always felt that it followed everything.

What I mean by that is that if a bass swell reads out -2 at one point in the cut and then it later reads "0" it's as much as a loudness difference to the ear as a voice peak reading 0 and then later reading +2.

What you can't do is try to relate how loud a bass swell is compared to a voice peak with a VU meter. You can't always tell how loud one mix is by ear relative to another easily (with lots of experience you may be able to start to do this).

At Motown, for disc mastering, we would distinguish between low-frequency modulation, midrange instrument modulation and higher-frequency modulation. We would also distinguish between "peaks", "swells" and "constant" levels (stuck needle type meter action) and write down notes on this in a 5 digit code with a possible added 5 digits when a volume envelope was considered needed.

With these notes, and not relistening to the mixes, we could pretty much set up the side to give us a consistent loudness level throughout the side and be at the level we wanted. Of course we did listen and rarely there would be adjustments that would be helpful (like EQ).

I certanly understand that you can't just look at a VU Meter without considering these factors and judge loudness differences from the VU Meter.

Quote:

We trained engineers who were brought up on the VU have forgotten how much TRAINING was necessary to learn how to interpret the VU and why. If you sit someone in front of a VU for the first time, for many reasons that I shouldn't need to get into at this time, they will tend to overreact to the changes in the meter, they will tend to follow their eyes instead of their ears. Look at all the rationalizations that we have to teach him about what's "tolerable" for short periods of time, etc. Rules which should be broken at certain times and why....
Well training is definately needed and it took a lot of experience to come up with Motown's leveling technique, but, once developed, it didn't take long for me to train a green "engineer" to read a VU Meter - maybe 2-3 hours. It took longer for me to train them to set up the lathe for the cut.

Quote:

I'm not saying that we didn't get excellent results when using a VU, arguably more consistent results than those using a PPM, but these were trained results from much experience. Today's engineers deserve a more accurate meter which needs even less training.

BK
Well I'm certainly for having a meter that is easier to use for engineers, as long as the meter is capable of giving the same kind of information that a VU does regarding loudness and loudness changes.

I DON"T mean any disrespect by all of this.

Bob
Old 6th May 2009
  #26
Gear Nut
 

all this i have been out

So if i,m correct most meters are 3db out approx. Waves/RN/PSP/Sonalksis/ but the new meter from tischmeyer is correct?

So on a final mix just to get an estimation say were aiming for -14 dbfs rms and approx.0.1 peak - this is just an average - my meters were really reading -17rms - I guess its not that big a deal if its goin to a mastering engineer- i wish had of known earlier, or manufacturerers should state if there is a standered they adhere too.----please dont say use your ears - please!
Old 6th May 2009
  #27
Mastering
 

Quote:
Originally Posted by Superdisc View Post
Well I'm going to risk being really irritating.

You've said several times that it "follows the voice", but I've always felt that it followed everything.
Right you are, I was just using shorthand language. Anyway, as you say, judging bass loudness is the weakest link of the VU meter. And in this world, how many new engineers are going to get the custom 3 hour mentoring of Bob Dennis to learn how to properly use a meter? I doubt Full Sail or any of the Universities give proper time to this important topic.

Quote:

Well I'm certainly for having a meter that is easier to use for engineers, as long as the meter is capable of giving the same kind of information that a VU does regarding loudness and loudness changes.
That's why I'm VERY interested in your reaction to the new ITU curve and Orban's take on the ballistics.

Quote:

I DON"T mean any disrespect by all of this.

Bob
You definitely have been totally respectful and clearly your mind is open.

BTW, it's impossible to post on the Internet without accidentily offending someone or getting a misunderstanding. The other day I posted something in Gearslutz about RN Digital's meters and I got an email from the designer who said that he heard that I had said the RN Digital's meters are 3 dB off, yet clearly I said nothing of the sort... Just remember, the real human being is usually a lot less frightening than his posts. Oh well, happy Internet y'all.

BK
Old 8th May 2009
  #28
Gear Nut
 

Quote:
Originally Posted by bob katz View Post
And in this world, how many new engineers are going to get the custom 3 hour mentoring of Bob Dennis to learn how to properly use a meter? I doubt Full Sail or any of the Universities give proper time to this important topic.
Students who intern with me get that mentoring, but using a meter to the degree we are talking about would be in a more advanced class. My class is an introductory mixing & mastering course of 50 clock hours (5 hous/week for 10 weeks). I do have articles on it though (and it will be in my text that's coming out this summer) - reading and practice goes a long way with the serious student.

I'd be supprised to hear that Full (...) Sail gives much, or any, time to teaching it -nor would any other of the major educational programs for that matter. Reading a meter is almost a lost art.

Quote:


That's why I'm VERY interested in your reaction to the new ITU curve and Orban's take on the ballistics. BK
Where do I find the ITU specifications?

We'll see about the Orban Meter. If you think it has potential, it probably does.

Bob
Old 9th May 2009
  #29
Old 9th May 2009
  #30
Lives for gear
 
dcollins's Avatar
 

Verified Member
[QUOTE=uncajesse;4168984]ITU BS.1770-1
stashbox.org : free anonymous file hosting
[/quote

Here is the Orban meter referenced earlier, which is as fascinating as any meter............

ORBAN Loudness Meter



DC
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump