Subject:VU Metering Tip in SF 7.0!!!

Posted by: ReliableSound

Date:9/26/2003 9:39:47 PM

Just a note about VU metering. Most professional VU/RMS metering is done with a sine wave @ 300ms integration. However SF 7.0 uses a square wave. The difference? A 3dB difference that's what. Generate a square wave test tone, and then one with a sine wave (assuming they are at the same digital level) and you'll see a 3dB difference between them on the VU meter. SF 7.0 lets you change the VU meter 0dB reference but not the fact it's calculated with a square wave. So the work around is this: in the preferences/other menu, type in -23 (in the dialog box that has the -20 SMPTE option) assuming you're using U.S. reference levels. Otherwise type in a value that is -3dB than the options you're using. Cool Edit Pro (now Adobe Audition) has the option to calculate RMS levels with square or sine wave, and WaveLab uses a square wave with no option. BTW, this still doesn't fix the way SF calculates RMS, run the Statistics option on your test tones, and you'll see the RMS levels are still different. E-mail Sony, let them know we need the option for Square AND Sine wave RMS calculation in order for mixes to transfer to broadcast equipment!

Subject:RE: VU Metering Tip in SF 7.0!!!

Reply by: philsayer

Date:9/27/2003 1:43:02 PM

I'm nowhere near your level of knowledge, but this is clearly an important issue... if I leave the VU meter settings alone, should I therefore assume that "true 0dB" is actually occurring at -3dB?

Or is it more complex than that?

Thanks!

Subject:RE: VU Metering Tip in SF 7.0!!!

Reply by: Sonic

Date:9/27/2003 5:50:07 PM

To my knowledge, there's no such thing as "square wave" or "sine wave" RMS. Either it is root-mean-squared, or it isn't. The RMS level of an infinite sine wave of 0 dB amplitude is -3 dB. The RMS level of an infinite square wave of 0 dB amplitude is 0 dB. That's not Sound Forge in particular, that's just the math.

The VU meter calibration value is somewhat arbitrary in the digital realm, but typical settings are shown in the Options box. Given 0 VU = -20 dB digital, 0 VU on the meter corresponds to -20 dB RMS, for example, a -17 dB (digital) amplitude sine wave.

To set it up in a listening enviroment, you'd have to adjust the sound card settings such that a 0 VU test tone (typically 1 kHz) displayed in Sound Forge is yielding +4 dBu or whatever physical calibration you choose at the analog output. Settings vary depending on the type of facility.

I looked in Cool Edit Pro 2 to attempt to deduce what you're getting at, but could find no such setting. Please clarify how to get to this and perhaps I'll see what you are talking about.

Then again, I suppose it's all a moot point anyway as long as you can set the 0 VU calibration to whatever you want for your particular environment.

J.

Subject:RE: VU Metering Tip in SF 7.0!!!

Reply by: ReliableSound

Date:10/1/2003 3:10:54 PM

I've e-mailed the SF tech support and they acknowledge this situation. In Cool Edit Pro/Adobe Audition, on the toolbar goto Analyze/Statistics. On the General tab, at the bottom you'll see two options under RMS Settings: 0dB = FS Sine Wave (or) 0dB = FS Square wave. You'll notice the stats change by 3dB when this is changed. I've worked at TV stations for ten years now, I know they use sine waves to represent 0dB. Yes, there is a difference on a VU meter, however not on a digital level meter. SF should add this option too. Cool Edit Pro is used in almost all radio stations, so Cool Edit Pro/Adobe Audition is experienced with this.

Subject:RE: VU Metering Tip in SF 7.0!!!

Reply by: Terpsichore

Date:10/13/2003 12:44:29 PM

I'd like to clarify this topic, which seems trivial but really isn't.

Yes, a full scale sine will have a mathematical value of -3dB rms, while a full scale square wave will calculate to 0dB rms. However, according to the AES-17 standard, a full scale sine should read 0dB rms, and should also read 0dB peak. AES-17 defines them to be equal FOR A SINE WAVE ONLY. A full scale square wave would then read +3dB rms. This is the standard. I believe Cool Edit follows this.

Sound Forge, on the other hand has not followed it. They've chosen the mathematical model in lieu of the standard. No harm, I suppose, but one has to remember the difference. I've always been frustrated, when selecting a passage and doing a statistics check on it, to find the reported rms level to be 3dB low with respect to the standard.

Now in SF7.0, which I don't have, I undertand the 0dB rms calibration point is selectable, at least for the new meters. This is positive step forward. I suspect the statistics is still "wrong" however. Perhaps someone can check this out for me.

Jim