Record Meter OVERS indication

Having switched from WL6 directly to WL8.5.2, noticed that the record window metering no longer indicates the number of “overs”, as in previous versions. This was a measurement our engineers came to rely on.

I am not finding a preference to turn this on/off, so has this feature been discarded or…?

Thanks.

This has been replaced by the “True Peak” option. When you get a sequence of overs, this translate to overs if True Peak is selected.

Thank You, PG, for the quick reply!

I am afraid the the True Peak metering does not report the information we are interested in.

  1. Under certain conditions, True Peak could predict >0dBFS even if the actual digital peak is not clipped.

  2. True Peak metering only indicates the held, max “True Peak” value. WL6 and earlier indicated a running total number of overs.

True Peak is an interesting metric, but until prevailing commercial music levels drop 5dB (cough/wheeze) then True Peak is mostly irrelevant. The previous metering features were very helpful for our work.

Actually, the feature is still there: see picture:
2014-11-14_07-52-24.png
Or do you mean something else?
If the window is too narrow (vertically), the values are not displayed by lack of room. Then just enlarge it.

Yes, that is the indication I am looking for. I am glad that it is still included.

For some reason, I cannot get that indicator to become visible here. I expanded the Record window to fill a 1024x768 monitor, but still no appearance. I also checked the main level meter, too (set to monitor input). Nothing there either. Both digital and true peak modes…

…the mystery continues… :wink:

The values appears only if there are overs.
Do you see the dB values anyway?

Yes, the dB peak hold value is being displayed, but without “# of overs”

I am overloading the ADC by >10dB with a 1kHz tone, therefore massively clipping.

Digital Peak reads 0dB, as expected.
True Peak reads only +0.46dB.

I did the tests using “record what is played back” and put some big gain to the Master Section. Hence clipping really occurs.

If you look at your recorded material, do you see clipping? If you perform a global analysis on the files, do you see clippings?

Yes, clipping: 32 consecutive samples pegged at 0dBFS, top and bottom. Global analysis detects them correctly.

Ah, then we are comparing apples and oranges. Here is my signal path:

Analog Signal → ADC → RME AES PCIe → ASIO → WL8 Record WIndow.

In your scenario, you are adding gain in the floating-point domain, therefore your will measure digital signal above 0dBFS. This is “over” but technically not clipping until exported back to a fixed-point format (I believe).

In my scenario there are no digital signals above 0dBFS, but the clipping is still real.

EDIT: Added some pics which show the captured, clipped waveform. As you can see, the samples clipped tally does not appear. When I turned the master section level up, then the tally did appear. So it seems this may only work during playback?


But if you’re clipping in the analog domain, the ADC will still deliver a 0 dB (hugely distorted) signal - or am I misunderstanding you?

Hi Arjan,

Right: 0dBFS, lots of consecutive 0’s in a row = clipping = flat top = distortion.

The old WL record window would keep a running tally of clips (like >2 consecutive samples @ 0dBFS). This was very helpful to know during and/or after a record pass.

I am unable to get this tally to appear in the WL8 record window, for an unknown reason.

@WYCA

With some more testing, I can reproduce what you describe. I remember now this is caused by a change that was introduced after WaveLab 6. In short, now WaveLab only detects real overs > 0 dB.
In WaveLab 6, WaveLab was detecting overs >= 0 dB.
Why this change? Because some peoples complained that 0 dB is not a real over (which is right).
But that also breaks your needs, which are legitimate.

The only way, will be to add an option in the level meter settings.

Thank you for acknowledging that.

If the previous metering function can be reinstated as an option in the next update or two, then you shall have some grateful engineers here. :slight_smile: