How to "read" VU meters

[iqi616]

I posted this in the Recording Hot thread before a(nother) fight broke out so I don't think I'll get an answer there.

What is the proper way to "read" VU meters? i.e. what constitutes tracking at 0VU?

You put in a signal and the needle bumps up and down. Should I be setting the input signal gain (assuming the source is a mic here - not anything easily calibratable like a keyboard) so that the needle maxes at 0VU, or so that it tends to hover around 0 VU +/- 3, or so that it hovers up to 0VU (say between -6 and 0) with occasional visits into the red?

[chadsxe]

http://www.roadogz.com/stories/downunder/vumeters.htm

[bennychico11]

0VU = -18dBFS


VU meters were originally calibrated so that they looked "good" with spoken voice. My opinion: the medium of the sound should fall around 0VU, soft sections will of course be a little quieter, and louder sections can peak up above every once and awhile. Never slam the meters to the top...0VU is where your gear likes to function at.

[Tim Gillett]

VU's are quite slow reacting. They dont show transient signal peaks well. They can sort of approximate to our human feeling of "loudness" and that's where they can be useful. Compare the difference in displayed level between a constant tone from an organ, and a single, sharp snare drum beat. Compare that to the readings on a peak meter.

They also have a very limited range, showing only the top 20 db or so of the volume range. For some types of music recording with big contrasts in level throughout the piece, the VU needle may be hardly moving off its rest for much of the time and that's fine.

I think VU's should be used with an understanding of their strengths and weaknesses and especially used alongside a peak reading meter which has a wider range of resolution.
Many analog tape machines had both a VU meter and a peak indicator display. Together they gave you more information than one on its own could.

Cheers Tim.

[SouthSIDE Glen]

Originally Posted by bennychico11

My opinion: the medium of the sound should fall around 0VU, soft sections will of course be a little quieter, and louder sections can peak up above every once and awhile.

I'd pretty much concour with this description, with the caveat that it can change slightly from device to device. While in general, VU metering tends to be "slow" or "average" reading, the actual ballistics of the meters varies somewhat from manufacturer to manufacturer. (There are even such things as "peak VU" meters with fast ballistics, but those tend to be rarer on pro and prosumer grade equipment, and tend to be found more on higher-grade consumer or audiophile recorders.)

The idea here is mostly to "ballpark" it, I think, and then to learn the individual personalities of your individual pieces of gear. There are some who will say to have a slow-ballistic VU meter peak at no higher than 0VU. Others will say it's OK to shoot a dB or two past on the highest visible peaks. Still others will say to eyeball "average" it at 0VU. A lot depends upon the meters themselves as well as the gear itself. Some meters are a bit slower or faster than others, and some analog circuits "like" running slightly colder or hotter than others.

Ballpark it around 0VU so you are not peaking too low below it or too high above it and you should be OK. From there, play with your individual gear boxes and see if each one sounds better or at least different slightly higher or slightly lower, and make whatever adjustments per session that you see fit from there.

Your Mileage Will DEFINITELY Vary .

G.

[iqi616]

Thanks for the replies everyone.

So I guess the consensus is that it depends on the individual meter and that I just need to get used to it.

[lomky]

Would it advantageous to run a 1k tone (or equivalent) through the convertors-preamp-vu meter-convertor and make sure there is 0vu throughout? I seem to remember doing this in my audio visual school days.

[iqi616]

Thanks for the reply . My question is more about what to do if the input is from a mic with the assumption that the path from the preamp to converters has already been calibrated for 0 VU. The answer seems to be experience with the individual meters and double-checking on the PPM before going for a real take.

[Fletcher]

Originally Posted by bennychico11

0VU = -18dBFS

Not necessarily... it can also equal -12, -14, -16, -20 or -24... there is no standard.

0VU should equal +4dbu which can be set to any reference level for the digital domain you would like.

One thing to remember is that a "VU" meter is an average of the signal, their ballistics are not set up to measure peaks [that would be the job of a PPM (Peak Program) meter]... so with things like drums, a good rule of thumb to have any headroom left is kik to -10 VU, kik and bass to -5 VU or below.

Best of luck with all you do.

[bennychico11]

Originally Posted by Fletcher

Not necessarily... it can also equal -12, -14, -16, -20 or -24... there is no standard.

i meant generally speaking when setting levels. There is no set standard, but most converters want to work with line level at or around -18dBFS.

[Fletcher]

Originally Posted by bennychico11

i meant generally speaking when setting levels. There is no set standard, but most converters want to work with line level at or around -18dBFS.

Most converters could care less what level signal you feed them so long as it does not exceed 0dbfs... how much headroom you define under 0dbfs is your business [matter of preference, zero standard].

[Farview]

Originally Posted by bennychico11

i meant generally speaking when setting levels. There is no set standard, but most converters want to work with line level at or around -18dBFS.

The converters don't care, the preamps do.

Originally Posted by iqi616

So I guess the consensus is that it depends on the individual meter and that I just need to get used to it.

The idea iis for the needle to be bouncing around 0dbVU. Like Fletcher pointed out, drums and other instruments with huge transients are a little tricky. You need to compensate for the fact that the meter will not read the transient. (for example: if you record a snare drum that read 0dbvu, you will probably be clipping the crap out of the recorder)

[bennychico11]

Originally Posted by Farview

The converters don't care, the preamps do.

so then what's with the posts of everyone (yourself I believe too) saying that 0VU should come in roughly at -18dBFS. I know you can push it to whatever level you like (just like in analog) but isn't that where the converters are calibrated?

how much headroom you define under 0dbfs is your business [matter of preference, zero standard]

by "zero standard", what do you mean? Not 0dBFS standard...i'm guessing

[Tim Gillett]

Originally Posted by bennychico11

so then what's with the posts of everyone (yourself I believe too) saying that 0VU should come in roughly at -18dBFS. I know you can push it to whatever level you like (just like in analog) but isn't that where the converters are calibrated?



by "zero standard", what do you mean? Not 0dBFS standard...i'm guessing

I would see the 18db differential as a rough rule of thumb based on the fact the VU under reads steep front transients. So in relation to the digital peak, you make the VU over read, to compensate. You're giving the VU a head start because it's slow.

It's impossible to set an exact figure in stone because it depends on the program content. Even different human voices have different peak to average ratios.

With analog tape machines the record amp was designed with headroom a similar amount above 0VU because even though there was increasing tape distortion above 0VU, it was gradual, and not nearly as harsh as amp clipping. So they made sure the tape was well saturated (heavily distorting) before the record amp clipped. At levels where the record amp was clipping, you already had massive tape distortion and so the clipping distortion was irrelevent.

If I was recording a snare drum line the 18db difference would save me from overcooking it.
If it was a Gregorian chant, that 18db safety margin could be reduced a lot, I would think.

Tim

[Farview]

Originally Posted by bennychico11

so then what's with the posts of everyone (yourself I believe too) saying that 0VU should come in roughly at -18dBFS. I know you can push it to whatever level you like (just like in analog) but isn't that where the converters are calibrated?

0VU is line level. A lot of converters are calibrated so that line level equals -18dbfs. It doesn't matter to the converters what the level is (below full scale), the point of those discussions is that the preamps don't like being pushed 18db above line level.

Originally Posted by bennychico11

by "zero standard", what do you mean? Not 0dBFS standard...i'm guessing

I think he means that there is no standard, none, zip, nada, zero.

[Fletcher]

Originally Posted by Farview

0VU is line level. A lot of converters are calibrated so that line level equals -18dbfs. It doesn't matter to the converters what the level is (below full scale), the point of those discussions is that the preamps don't like being pushed 18db above line level.

There are many pre-amps that can be pushed far beyond +18, there are many that can't be pushed beyond +12 without serious distortion. It depends on the design of the pre-amp.

I think he means that there is no standard, none, zip, nada, zero.

You forgot "zilch"

FWIW... in the analog domain, again, depending upon the machine and the tape forumulation, you can push things farther than you think... for instance, RGI 911 has a maximum operating headroom of +15... so, when you align the tape machine you can align it for 185nWb/m which is a "0" level and have 15db of headroom [above "0VU"]... or you can align it for 250nWb/m [+3] and have 12 db of headroom or 355nWb/m [+6] and have 9db of headroom [etc., etc., etc... there are a whole lot of ways to align a tape machine... some more creative than others].

In the analog situation you're trying to balance "headroom" to noise followed by how far you can go before the distortion from the tape becomes an impediment to the sound you're trying to achieve [remember, back in the day people weren't using tape as an "effect", they were trying to store the audio in the cleanest, clearest possible manner... it wasn't until the mid to late 70's that people started to really get into using the anomalies of tape in a "creative" manner... which grew into tape abuse in the 80's and 90's mostly due to ignorance of how the shit actually worked].

FWIW, there have been several of us who have been lobbying both the AES "Standards Committee as well as the "P & E Wing" [Producers and Engineers Wing] of NARAS to adopt a standard... which for all intents and purposes could be an absolutely arbitrary number... like -18dbfs... but it absolutely would not suck to have at least a guideline standard. It took them on the order of 40 years for "Pin 2 Hot" to become the XLR standard... and if I'm not mistaken that was never an "official" standard but more of a gentleman's agreement throughout the industry to adopt "Pin 2 Hot" for XLR connectors.

Peace

[Tim Gillett]

Perhaps we need to distinguish between a VU meter and a voltmeter, and between db's and volts.

VU stands for "volume unit". A VU meter is strictly a relative measurer only. Sure, in a recorder it ends up being calibrated to one voltage or another but it still reads in db's (relative) not volts (absolute).

0VU is not any particular voltage. We should really be talking in terms of voltages or dbm, dbV etc which are engineering conventions which translate into voltages.


0dbm or 0dbV are real levels by which different pieces of equipment can be made to match because they are real voltages. 0VU has no meaning between pieces of equipment.

Tim

[Fletcher]

Originally Posted by Tim Gillett

Perhaps we need to distinguish between a VU meter and a voltmeter, and between db's and volts.

OK

VU stands for "volume unit".

Right.

A VU meter is strictly a relative measurer only. Sure, in a recorder it ends up being calibrated to one voltage or another but it still reads in db's (relative) not volts (absolute).

In a recorder, on a console, the output meter on a mic-pre, compressor, etc. is all calibrated to a standard... so yes, the meter is a "relative measurement" but the standard for the claibration of that meter is not "relative"

0VU is not any particular voltage. We should really be talking in terms of voltages or dbm, dbV etc which are engineering conventions which translate into voltages.

0dbm or 0dbV are real levels by which different pieces of equipment can be made to match because they are real voltages. 0VU has no meaning between pieces of equipment.

Uhhhh... no.

0VU is indeed a standard voltage.

0dbu=.775 volts, 0VU is +4dbu which is equal to 1.225 volts. That is the standard, convention, whatever the hell you want to call it... it was handed down on "Tablet 3" when Moses descended from Mt. Sinai... we could call it the 11th commandment if you'd like but I think it was actually somewhere around 23 [with a bullet] on the commandment top 40... if I remember correctly the actual text read something like "Thou Shalt Calibrate All Meters Used In Professional Applications To +4 dbu".

[ofajen]

Originally Posted by Fletcher

In a recorder, on a console, the output meter on a mic-pre, compressor, etc. is all calibrated to a standard... so yes, the meter is a "relative measurement" but the standard for the claibration of that meter is not "relative"



Uhhhh... no.

0VU is indeed a standard voltage.

0dbu=.775 volts, 0VU is +4dbu which is equal to 1.225 volts. That is the standard, convention, whatever the hell you want to call it... it was handed down on "Tablet 3" when Moses descended from Mt. Sinai... we could call it the 11th commandment if you'd like but I think it was actually somewhere around 23 [with a bullet] on the commandment top 40... if I remember correctly the actual text read something like "Thou Shalt Calibrate All Meters Used In Professional Applications To +4 dbu".

I'm afraid this is misleading for the readers of this forum.

Back in the day when most all recording was done in pro studios, with balanced equipment with high voltage rails and clipping levels in the vicinity of +28 dBu and recording and mixing was done on analog tape recorders, the +4 dBu standard made perfect sense. Actually, my old AG-440B has a jumper to switch between the broadcast standard level of +8 dBu and the studio level of +4 dBu.

But now we're talking about HR.com, where lots of folks use digital recorders and gear with lesser clip levels. That means the old 14 dB crest factor limit imposed by tape may not apply. And gear with clip levels of +22 dBU (say with 15 V rails) or even as low as +14 dBu (9V power supply) are not at all rare.

So, here on HR.com, we now need to have a different conversation about standardizing studio line levels. Setting 0VU at +4 dBu may be totally inappropriate in these applications.


1) Say you want to stay away from the clip level because the gear starts to sound ugly a bit below the clip level. Starting with a mixer or preamp or interface with a clip level of +22 dBu, and giving yourself, say, 8 dB of cushion, you're down to +14 dBu as your peak level.

2) Then you have input signals that may have a crest factor of 20 dB if they're not heavily compressed (mine often aren't, for example). So you really would want a reference line level of -6 dBu to give yourself 20 dB of headroom and 8 dB of cushion above that.

The bottom line is that folks here need to know and work carefully within the limits of their gear, and a +4 dBu studio operating level may be well beyond the capability of their gear.

Cheers,

Otto

[Farview]

Originally Posted by ofajen

I'm afraid this is misleading for the readers of this forum.

Back in the day when most all recording was done in pro studios, with balanced equipment with high voltage rails and clipping levels in the vicinity of +28 dBu and recording and mixing was done on analog tape recorders, the +4 dBu standard made perfect sense. Actually, my old AG-440B has a jumper to switch between the broadcast standard level of +8 dBu and the studio level of +4 dBu.

But now we're talking about HR.com, where lots of folks use digital recorders and gear with lesser clip levels. That means the old 14 dB crest factor limit imposed by tape may not apply. And gear with clip levels of +22 dBU (say with 15 V rails) or even as low as +14 dBu (9V power supply) are not at all rare.

So, here on HR.com, we now need to have a different conversation about standardizing studio line levels. Setting 0VU at +4 dBu may be totally inappropriate in these applications.


1) Say you want to stay away from the clip level because the gear starts to sound ugly a bit below the clip level. Starting with a mixer or preamp or interface with a clip level of +22 dBu, and giving yourself, say, 8 dB of cushion, you're down to +14 dBu as your peak level.

2) Then you have input signals that may have a crest factor of 20 dB if they're not heavily compressed (mine often aren't, for example). So you really would want a reference line level of -6 dBu to give yourself 20 dB of headroom and 8 dB of cushion above that.

The bottom line is that folks here need to know and work carefully within the limits of their gear, and a +4 dBu studio operating level may be well beyond the capability of their gear.

Cheers,

Otto

I think we are making this harder than it needs to be.
Most of the people here at HR are probably using interfaces that can not be calibrated by the user. They are stuck with the factory calibration. Everyone should find out what line level is calibrated to (in the digital world) on their interface and try to keep the level bouncing around that. It really shouldn't involve a multi-meter and an ociliscope, just a basic understanding of gain structure.

[Fletcher]

Well Otto... you do have a point on the integration of "-10" and "+4" stuff... though the operating headroom of an analog unit doesn't really come into play here as we're talking about "0"... but you're right, there is -10 "semi-pro" stuff all over the place and it can be a bitch to interface that with +4 "pro" stuff... but, even the lowly RNC runs at +4 [going to +28 before clipping even though it runs on a "wallwart"... which is again why the 'internal headroom discussion is off the table].

Jay also has a point that there is damn little "semi-pro" hardware I've seen where you could actually calibrate the meters so the point is essentially moot in that relm. My point in the tech "waka/waka" talk was that there is a standard for calibrating meters, that they aren't set arbitrarily... nothing more, nothing less.

[ofajen]

Originally Posted by Fletcher

My point in the tech "waka/waka" talk was that there is a standard for calibrating meters, that they aren't set arbitrarily... nothing more, nothing less.

A good point, that is the standard for meters. About the only gear I have that can easily change is my old Altec 1567a. It has VU range settings from 0 to +12. It's the exception that proves the rule. The fact that most meters can't be recalibrated just adds to the complexity if you're gear isn't quite up to the +4 dBu challenge.

I just wanted to make sure that folks here using a mixture of gear understand that +4 dBu might be too high of a studio operating level for some of their gear to provide and still have the headroom to perform up to spec. Kind of a digression from the calibration issue, but also important for folks to understand.

BTW, the RNC spec sheet online claims a clip level of +22.5 dBu, which is about what I always assumed it was. It does have an unbalanced signal path, after all.

Cheers,

Otto

[Fletcher]

OK... +22.5 is still pretty damn healthy when you're talking about most project studio hardware... the desk in my room only does +24 and it's class A discrete running on bi-polar 24 volt rails [with a 15 amp PSU]...

As for playing "mix and match" with +4 and -10 gear... you really shouldn't do it without "bump boxes" unless you're looking for VERY compromised results from all of the involved equipment... and yeah, I know not everybody knows about this kind of crap but it does make a difference... it should probably be part of the training for the Banjo Mart sales mooks that pimp the great majority of the mis-matched crap on the planet but alas it is not... on the bright side, that's why there are information sites on the subject such as this one... knowledge is power and all that.

Peace.