PPM scale

This page http://www.soundonsound.com/sos/jun00/articles/metring.htm has a nice comparison of different PPM scales:


Everything seems to make sense, except the DIN PPM scale. It is referenced at + 9dBu, but differs more than 9dB from the dBu-scale at -20 and below. Is this a special characteristic of the DIN scale or a mistake in the drawing?

;Matthias
PS: the VU scale looks strange as well at -20

Look at the bottom of the VU scale.

All other scales are log, the VU is sort-of linear. The DIN-PPM is equal-spacing-per-deciBel at the top half, and closed up at the bottom. most other PPM scales are equal-spacing-per-deciBel all the way across the scale (i.e. log). VUs are largely linear, but the diode at the bottom makes things nonlinear.

Keith

the DIN PPM scale is wrong.
all the DIN PPM meters are +6 dBu = 0 dB reading.

from a NTP PPM 277-400B callibration manual:
DIN: +6 dBu for 0 dB reading
NORDIC: +6 dBu for 6 dB reading
BBC: +8 dBu for "6" reading

That article on metering in Sound on Sound was certainly not one of Hugh Robjohn's best efforts; not only are there mistakes in the drawing, but also there's some very woolly writing where he talks about 'perceived volume levels'.
The VU meter is good for what it was designed for.... to give an indication of perceived volume.
The PPM is adequate for its intended purpose.... to watch modulation levels in a broadcast chain.
Digital meters look pretty.

Cover them all up and use your ears! :razz:

Cover them all up and use your ears! :grin:

Click to expand...

hahahaha ILOVE IT!

later
ts

You are quite correct. The DIN PPM scale on that diagram is wrong. When the graph was redrawn from my original sketch it was done incorectly and the error not noticed until it was too late. A corrected graph was included in a subsequent issue of SOS.

But this was a useful reminder for me to try and get the webpage upadted with the correct graph for future readers. Thanks.

Not sure quite what Ted is referring to when he complains of 'wooly writing' though. Care to elucidate Ted?

Hugh

[quote author="hugerr"]

Hugh[/quote]
Whatever the reason for joining here, always nice to see people coming aboard that are known from elsewhere ! Hope you stay around.

Bye,

Peter

Welcome, Hugh.

> what Ted is referring to when he complains of 'wooly writing'

He's probably just being Ted.

Reams have been written about 'perceived volume levels'. We really don't know much about it.

In context of a few-page article for working mixers, the article's coverage of 'perceived volume levels' is necessarily sketchy. As I think it says: meters can't tell you how loud it sounds.

I would pick a few nits:

"The VU (Volume Unit) meter is amongst the simplest of meter designs, and it has been used since the very beginning of the audio broadcasting and recording industry."

The VU Meter spec was approved in 1938/1939. We sure did a LOT of broadcast and recording before the VU Meter came along.

I'd say "mechanical meters, culminating in the iconic VU Meter and imitators".

I have a pre-VU-spec meter (new in box). Compared to a VU, it is a slug. Yet this is the type of metering that was commonly used. But peak-indicating meter were also used: RCA had a neon-lamp peak meter to go with their clanky film-sound modulators, and limiters were used and their meters do in effect indicate if peaks are below or above maximum level.

Neither the slug meters nor the peak meters give good approximation of "loudness". While Ted say to use ears, long days in the radio station or film sound studio don't give good loudness judgement. Apparently Weston, after solving the light-meter problem for photography, worked on the loudness meter problem. Their meter has a much stronger magnet than a generic DC meter, and they got a reasonable approximation of the integration times of the ear. It does not even try to correct Fletcher-Munson, nor read more than 20dB (which is odd, because their light meters used tapered magnets to cover 80dB ranges); still it is better than a slug-meter or just avoiding peak-clipping. And it did turn out to be very suitable for tape which does not have hard clipping.

"It was designed to display an approximation to the RMS voltage level of electrical signals"

It measues Average integrated over ear-like time constants. It would actually work better if it responded to RMS, but that's not easy in a simple meter. For sine the difference is just calibration (indeed the line-up level is specified in RMS terms); for other simple tones a correction can be applied; for speech/music it properly indicates Average and the peaks are just unknown (pick a number from 10 to 18).

> VU meters are inherently cheap

A true VU meter was never cheap. Part of the attraction of PPMs is that once we had transistors, you could PPM a cheap meter movement instead of the rather expensive movement needed for True-VU.

Of course the world is full of cheap "VU Scale" meters. Plain 50uA movements, though the better Japanese movements were a lot less sluggy than the old-old US movements. These do dance well, and can be as valid as a VU meter. Put a good Teac meter next to an Ampex on the same signal, and you will see the difference. Two True-VU meters will dance identically, meaning you can take your project to another room and still have a loudness reference.

LEDs soon killed mechanical meters except among traditionalists. A 10-LED array is about the price of a very cheap movement, 1/50th the price of a True VU, much more rugged, and easier to "style". And the meter-lamps never burn out.

Ah, but that's 500+ words about a mechanism that is totally obsolete. I don't think you can even buy a True VU mechanical meter today. If you do have one in old gear, you are probably running it into digital audio, and the digi-meter on that is your guide to overload. VUs are now decorative.

Like speaker design and building
Levels and Metering have been a special interest subject for me since the begging of my time in audio.

I really do know a great deal about the subject and yet it still doesn't help me explain to people how and why they have trouble.

At work we currently have an issue with one of our live to air programs and all involved have their techinical data to back up their stand.
trouble is the a ' perceived volume '.
As with all things, people have a perception and at the end of the day that perception does matter regardless of data or facts.

Right now at RO another ProTools and levels thread has sprung up and over at ProToolsForum.com we had a very long thread on the go before server conflict force a move.
I hope to resurrect that stuff and write an article for our new openings.

... a mechanism that is totally obsolete
VUs are now decorative.

Click to expand...

so how and why is that I seem to solve so many problems with just one simple passive VU meter ?

perhaps it is just that I am lucky enough to have a stock pile of good VU meters.

If I though about it I could probably achieve the same result with a true RMS meter and a Peak voltage meter.
:shock:
trouble is ... that these units are not passive and require active electronics.
SO can you ever be sure they tell the truth ?
I'm an old fart that likes to rely on the trusty old moving coil meter.

The passive VU meter is a must for the audio tool box ... along with elec tape and gaffa and all the other little helper in there.

I doubt this will ever go away.
and nor does trouble with interfacing gear and what is a balanced line.
:roll:
we love audio

Thanks for all the replies, i was actually only confused by the unequal spacing.

To the people who use PPMs etc. in their work: How important is the exactness of the bottom of the scale (like -30 below the red area) ? Do you need to know if the signal is at -32 or -36 or is it enough to know that something happens a bit below -30?


;Matthias

> why is that I seem to solve so many problems with just one simple passive VU meter?

Because you, Kev, are a blast from the past.

> I could probably achieve the same result with a true RMS meter

Nothing wrong with an RMS meter IF the time constants are similar to the ear. The integration time is more important than the distinction between Average and RMS: RMS is probably more-correct, but for the range of waveforms found in speech and music, the difference is lost in averaging and scaling.

> and a Peak voltage meter.

> these units are not passive and require active electronics. SO can you ever be sure they tell the truth?

If they are properly made, they tell the truth they measure. I have no problem believing my peak meters read peaks within 10%.

But I can easily devise, and pretty easily find, two waveforms with identical peaks and VERY different "loudness".

In straight mastering, just capturing the performance, on modern digital media, the peak meter is your only friend. As long as you don't try to put more than 32K levels into your 16-bit (32K levels) A/D, you got it.

(Though, if running through a soft analog stage, like a tube preamp, the average meter may be handy for telling when you get most of the music up into the "soft zone" so most of it comes out "flavored". The peak meter will tell when peaks are hot, but material of high or low peak/average ratio will be bent differently even at the same peak level.)

Some mix people like to have a "volume" indicator. Maybe not for one track, where the loudest peak should be just inside the summer's range and everything else is balanced by ear. But when compiling tracks into an album, I often run a "loudness" measurement on each track and plan the overall relation between tracks so they play nice together.

This even when all tracks were recorded live in one hour: the soft tracks may average 15dB below the loud tracks, which worked in context of a large quiet chapel with an attentive audience, but does not work on the average playback system unless the listener keeps jumping to the Volume knob. I set the last hand-clap on "Go Where.." (a rockin' spiritual) at +2dBfs (yes, clipped), find the average for that track is say -17dB, then for "Vesper.." (a daytime lullaby) I'll master it at -25dBfs average, 8dB lower (it may have been 20dB lower in performance).

> How important is the exactness of the bottom of the scale

I say "not very".

I like to see far down the scale. I sometimes track or master in un-quiet rooms. Of course when I first set-up I'll crank-up a quiet passage and see how bad the non-performance noise is: rumble, buzz, etc. and deal with that if needed. But when balancing at reference monitor level (my monitor gain knob is marked for just about zero VU = 85dB SPL) I may not be able to hear noise that will be audible in a quieter listening room. Seeing how far down the meters duck in silences assures me it won't be a problem. That means being able to see well below -50dBfs (I often work on a -72dBfs scale) but accuracy is not an issue: I play reference tracks known to play well in quiet rooms and see where the meters flutter.

I've just got to jump in again and take serious issue withh PRR.... :grin:
To say that a 'VU' movement is expensive, and to infer that any fool can chuck a few transistors at a cheap Chinese movement and make a PPM is an apalling misrepresentation!
Tell that to anyone who has worked on design for the BBC and you would risk physical violence!
A VU meter is a conventional moving-coil meter with a full scale deflection somewhat less than 500 microamps and with a particular design of lightweight alloy indicator needle, a specific scale and that lovely buff background colour.
The PPM (according to BBC Specs that I worked with) is certainly a simple moving-coil meter (1milliamp FSD), but it has a highly specialised movement so that it wil not only move extremely quickly, but the mass and mechanical damping are such that the overshoot is very carefully specified and controlled.
Meters that met the spec. were only manufactured by 'Turner' and 'Sifam' in the UK (now there is only 'Sifam' left), and they were extremely expensive; VU meters were always the cheap alternative.

Welcome Hugh!
I did think you were waffling just a little in the sidebar of the article; I couldn't work out if you agreed with the idea of measuring levels with a VU or not!
But the piece was intended as instructional and apart from the oddity in the diagram, it was very informative.

BTW, on my current gear I use 'nominally VU' meters using buffer amplifiers and meter movements made in Taiwan (as most are nowadays). The reason I like to use a VU type is neither nostalgia nor prettiness; it is because it's a useful continuous indication of overal volume. Adjustment to the last half-a-dB can be done against digital meters..... but I still wouldn't trust'em! :shock: