peak vu analogue RMS dbFS PPM ARRRRRRRGGGGGGGGGGHHHHHHHHHHHHHHH. . .

[Josh Bass]

I'm trying to expand my knowledge of audio, so let's see if I've got this right, and then ask a question or two:

We have two ways of metering/monitoring audio: VU and full scale. A VU meter generally goes from -whatever to +20, while an FS scale will go from -whatever to 0. In VU, +20 is the point at which you're hosed and your audio distorts, in FS, it's 0. Is this right?

Sometimes VU meters measure RMS/average values, sometimes they measure peak (I'm thinking of the SD302 mixer which can do both, but still uses the VU scale). IF they measure RMS, the "loudness" is averaged out around every third of a second, instead of "frame by frame" (if we're talking audio for video production here). So, you can have a situation where, if I snap my fingers, looking at a peak meter it would spike, while looking at RMS it might barely move, 'cause of the split-second interval of time that a finger snap occupies. Correct? Sort of?

Now, if we're looking at the digital meters on these camcorders (pro and prosumer), are these ALWAYS measuring peaks? Furthermore, if you see ANY dfFS meter (with the highest value being 0), is THAT always measuring peaks? I ask because the meter on my XL2 responds instantly and accurately to a quick sound like a finger snap, while the same sound will barely nudge my mixer's analogue VU meter (it's an FP33).

IF this digital cameras have peak meters only, then why even bother with RMS metering? Isn't all that's really important what trips the max recording level and distorts? Do I not understand the whole idea of peak metering?

Help the Bass expand his brain.

[Nick Flowers]

In the UK and Europe we tend to use PPMs, either BBC type and its derivatives (such as the old Nagra 'Modulometer') or Nordic. These have ballistics that respond with a quick rise time and a slow release time; those of a scientific leaning can supply the actual figures no doubt.

We're used to them and they work well with not only music and spoken word but with all other fx - although gunshots might present a problem. On the BBC type the numbers are from 1 to 7, with 4 being zero level, 6 being 100% and 7 representing the hope that your limiters are working. It's what you are used to: I couldn't use a VU or SMPTE as it would require thought. My eye and ear are accustomed to the PPM and my fingers act on the knob or fader automatically. I imagine that someone used to VUs would find PPMs irritating.

My old guv'nor used to tell me of a mixer they used in MGM's studio in Boreham Wood just after the war. No metering at all, except for a light that went on when the limiter was working. Maybe an RCA mixer? It kept you on your toes and you really, really used your ears.

[Josh Bass]

My bad. Sometimes I forget we're an international board. I meant in America, generally, we see those two types of metering.

[Daniel Epstein]

Truth is there is no standard for what the meters on camcorders mean unless they have indications on the meters. Every piece of gear seems to have its own way of displaying the signals it is recording. Some are easy to understand and some not. It seems to be up to the user to figure out what the individual piece of gear is doing. If you are using a new piece of gear to record on you have to test it out and see how it works with the other gear you are hooking up to it. Professional gear usually has better metering than consumer gear but not always. Sometimes reading the manual helps but not always. Preparation for a shoot may require some extra time figuring out the best way to use the gear.

[Ty Ford]

Josh Bass: I'm trying to expand my knowledge of audio, so let's see if I've got this right, and then ask a question or two:

We have two ways of metering/monitoring audio: VU and full scale. A VU meter generally goes from -whatever to +20, while an FS scale will go from -whatever to 0. In VU, +20 is the point at which you're hosed and your audio distorts, in FS, it's 0. Is this right?

>>>>>Shure location mixers with analog metering (needles that physically move) I've seen usually don't go above +4 or +6. Now that's an analog VU meter that is too slow to track transient peaks.

Sometimes VU meters measure RMS/average values, sometimes they measure peak (I'm thinking of the SD302 mixer which can do both, but still uses the VU scale). IF they measure RMS, the "loudness" is averaged out around every third of a second, instead of "frame by frame" (if we're talking audio for video production here). So, you can have a situation where, if I snap my fingers, looking at a peak meter it would spike, while looking at RMS it might barely move, 'cause of the split-second interval of time that a finger snap occupies. Correct? Sort of?

>>>>>>> I don't know about every third of a second, but the 302 measures rms and peak. I set my 442 to rms and peak, but only concern myself with peak.

Now, if we're looking at the digital meters on these camcorders (pro and prosumer), are these ALWAYS measuring peaks? Furthermore, if you see ANY dfFS meter (with the highest value being 0), is THAT always measuring peaks? I ask because the meter on my XL2 responds instantly and accurately to a quick sound like a finger snap, while the same sound will barely nudge my mixer's analogue VU meter (it's an FP33).

>>>>>> In my experience, the newer camcorders' LCD or LED meters do measure peak. I have an XL2. It measures peak. Yes, well, as you almost got to before, VU meters do not read peak. The transient passes so quickly that the needle mechanism can't respond fast enough. I think Shure added a peak LED at some point. In addition, analog was somewhat more forgiving about "overs" and could take a hit more gracefully.

>>>>>>>I have had one or two occasions in which the metering on the camera I was feeding was reading overs when I was reading +20 (after setting +20 = 0dbfs). I wrote it off to minor differences in calibration and meter sensitivity.

>>>>>>>Sidebar: Consider, when you're calibrating, tone from mixer to camera, you throw tone from the mixer and adjust the camera input controls. As the camera meters increase in value, you want to stop JUST AS the meters hit -20 (for example). If you increase further, but before the next meter increment shows (maybe at -18, -16 or higher), then you're not really calibrating at -20. That squink of difference in adjustment can make all the difference when you get too close to 0dbfs.

>>>>>>Anyway, I pulled my mixer levels back a bit just to be safe. I seldom allow audio that high, anyway. I usually aim for peaks around +14 so I have room for the excited moment.

IF this digital cameras have peak meters only, then why even bother with RMS metering? Isn't all that's really important what trips the max recording level and distorts? Do I not understand the whole idea of peak metering?

>>>>>Good point. It think we see both because VU is historically so well entrenched. There is valuable info in knowing the difference between VU (or RMS) and peak. That tells you how much compression and limiting is being used. The Dorroughs meters are a good example.
..Dorrough Electronics..

Regards,

Ty Ford

Help the Bass expand his brain.[/QUOTE]

>>>>>PS: You may notice that some transients e.g. handclaps, can sneak through the limiter on a mixer and spike your audio. That's because the transient is too quick for the limiter's attack setting. Choosing an attack setting must be done carefully. Too fast and some audio may not sound as good.

>>>>>As a rule, I set my limiter threshold to +16 against a +20 = 0dbfs. That gives me 4 dB of safety for these ultra fast transients.

[Josh Bass]

I've left mine on the factory setting. It's more of a pain to adjust an FP33 than a SD302 for that particular setting. I think the default on mine is somewhere around where you leave yours. (If the day comes I can justify a $1300 mixer to solve some of these issues, I'll be as happy as you are).


At any rate--

Thanks.

So let me get this straight. . .you're saying that being able to monitor VU and Peak at once allows you to see how much compression/limiting is going on (by looking at the difference between them, except during transient peaks)? Or did I misunderstand that?

As far as camcorder meters all being different--- A pox on all their houses. I guess this is where having a mixer that outputs full scale tone is especially helpful.

[Paul R Johnson]

I suspect that what we all want is a way to assess the average level, plus have some confirmation that we are not going over FS and introducing digital distortion - which is horrible sounding. So I can live with pretty well any kind of metering as long as I can see where the real maximum is. In truth, I work with plenty of expensive digital audio mixers, and have to put up with some very different display characteristics. In digital displays, the key feature is lots of segments - five or six simply isn't enough. The US origin VU meters were great when they had real needles - same as proper BBC type PPMs, but on-screen bar graphs and led/lcd segment meters don't really show what is happening. My preference is for fast acting led/lcd meters with a peak hold delay - as long as they show the low level sound near the bottom. I've discovered that I can use a 24T multi-track recorder as a display for my theatre mixer and its really sensitive at the bottom end of the scale - where my main mixer doesn't even light up the first segment. I use this now because I can see a mic is working without having to put headphones on and listening. It's annoying the Yamaha desk meters don't even register this. I suspect manufacturers are simply making sure that 0dB = 0dB although I suspect many over read for safety.

[Josh Bass]

You're saying there're meters (for this discussion, let's keep it constrained to video production oriented things, so mixers for video production) that go over 0, but 0 is still the point of distortion?

[Ty Ford]

Originally Posted by Josh Bass

So let me get this straight. . .you're saying that being able to monitor VU and Peak at once allows you to see how much compression/limiting is going on (by looking at the difference between them, except during transient peaks)? Or did I misunderstand that?.

You got it. The more compression and limiting, the smaller the difference between RAM and peak.

Regards,

Ty Ford

[Josh Bass]

The idea being that if you have a lot of limiting, your levels are too high to begin with?

[Gary Nattrass]

I have always used BBC PPM meters for over 28 years and the latest version I use is a plug in for Pro Tools and other software apps: Raw Material Software - PPMulator

Odb PPM = -18dbfs and for Tv work max peak is +8db which is 6 on the PPM or -10dbfs

[Jimmy Tuffrey]

Originally Posted by Gary Nattrass

Odb PPM = -18dbfs and for Tv work max peak is +8db which is 7 on the PPM or -10dbfs

That must be a typing mistake. You obviously meant 6 on the PPM.

[Gary Nattrass]

Originally Posted by Jimmy Tuffrey

That must be a typing mistake. You obviously meant 6 on the PPM.

OOps never could count to more than ten, as you say 6 is the max sorry, I tend to use the bargraphs more than the needle meters so forget about the numbers.

[Steve Oakley]

most cameras change their display style at -18 . white to yellow, solid to hollow. some even give you a tick mark for -18, most can also set for -12 if you want. after that its easy enough to figure out. as long as levels are making it most of the time to -18 you've got basic levels, but I prefer to run a little hotter then that. if you setup the FP33 so that its 0db = the camera's -18 you're pretty much good. with experience though, once I set tone up that way, I usually give the master gain another 6+ db with a HD100 camera. other cameras make like you to be dead on, others can take a hotter level. it seems that some camera meters have different response at different freq's :(. mackie mixers are the worst. run tone thru them and think all is golden, run program and it doesn't come close to matching. FP33 is better.

so anyways, with 0db analog = -18db digital you should be good. the FP33 has 16-18db of headroom. the limiters if engaged all the way open are set to +15db, so 0+15 still leaves you with 5db of headroom in the camera. I've got my limiters set at around +12/+13, then run a little hotter. makes crappy camera audio inputs more happy as you are getting the signal hotter then the noise. with the limiter set this way, only a super hot transient that the limiter misses might clip, and chances are you still might have enough headroom.

[Josh Bass]

Actually, we found that running tone at 0 out of the FP33, setting it at any level of the camera, can cause distortion in certain circumstances. It's very odd. . .I found that as I was saying "test/check" into a mic, with the tone at 0 on the FP33, -20 or -12 on the cam, you get a bit of distortion on the "s" and "ch" syllables. For real! Tried it with several different mics, even took it to a local rental house and tried the mixer with a HDX900 for comparison. Not an issue on the HDX.

Conclusion? Mixer's output, at 0, is simply too hot for the XL2, period (by this I mean going through the mic inputs or the beachtek/line level, happens either way). We found that running tone out of the mixer at -10 instead of 0 alleviates the issue, along with opening the XL2's pots to at least around 1 or 2 o'clock. I asked a local sound guy who uses an FP33 as his rig, and he mentioned he's had the same issues, with the same ideas for resolution.

At any rate, so you're suggesting running tone to the camera at -18 instead of 20? I'll try that.

As far as upping the master output on the mixer, I thought the idea of unity gain (the place where the master output on the mixer wants to be set) was that the preamps were neither amplifying nor cutting the signal out of the mixer. If you up it +6 db, isn't that introducing extra noise? Isn't it better to raise the volume on the faders?