Jump to content

Volume Fluctuation- Pumping Sound


HammerMan

Recommended Posts

Hello All,

 

I apologize if this is a dumb question but I'm perplexed by it. Been reading the forums here and searching the web with no real solution so I thought I'd just go ahead and ask it.

 

What I'm trying to do with WavePad is two things:

 

1. To make my song mixes relatively the same volume level compared to one another.

2. To put each song at a dB level that is consistent with commercially recorded songs for playback on a burned CD-R.

 

But when I use Amplify effect to make the sound louder, it puts out this volume "pumping" sound that fluctuates between the rhythmic peaks of the drumming. It makes the mid-range dominant instruments sound like they are warbling!

 

I understand that the real solution is to compress the stereo tracks to squash their dynamics making soft parts louder so the average level brings it to where I want to hear it, but even the compression presets are adding in this pumping effect even on the most subtle settings. I used to use CoolEdit96 to make the volume level louder and it never had this volume fluctuation problem. It would just crudely squash the mix so I could get a general idea of what the mixes sound like upon playback on a portable CD player, albeit with some array of distortion but it was definitely listenable. Maybe it had some soft clipping algorithm on the output to prevent full on digital nasty sounding artifacts?

 

So... How do I get rid of this "pumping" sound on the volume? Does it have to do with Automatic Gain Control or the AutoTrim Threshold settings? Tried messing with those but nothing is alleviated yet. Please understand that I'm trying to figure out a way to do this with little tweaking. I want to have a fast and crude way to just squash the mix and bring it up to portable CD player compatible output level and have it sound loud although inevitably a bit distorted.

 

Thanks

Link to comment
Share on other sites

Hi,

 

You can use the Normalize option, from the top menu click on Effects > Normalize.

 

To 'normalize' is to adjust the volume so that the loudest peak is equal to (or a percentage of) the maximum signal that can be used in digital audio. Usually you normalize files to 100% as the last stage in production to make it the loudest possible without distortion. Another reason to normalize is to have multiple tracks sound equally loud, or to have equal average loudness.

 

The 'Peak' normalization method finds the sample of the greatest magnitude within the file. Normalization is then done with this value as the peak. With the Normalize Peak Level set to 100% (0dB), the whole file will be amplified so that the peak reaches 0dB.

The 'Average Loudness (RMS)' normalization method normalizes according to the file's average loudness, or volume. Multiple files normalized to the same peak level using this method will have equal average loudness. The 'Normalize Peak Level' for this method should be set much lower than for the Peak method, because the average loudness will always be lower than the peak sample.

 

The 'Peak Loudness (RMS)' normalization method attempts to normalize according to how loud the loudest part of the file will sound. This is the best method to use to make multiple tracks sound equally loud. As with Average Loudness, the 'Normalize Peak Level' for this method should be set lower than for the Peak method, because the peak loudness is lower than the peak sample. The actual algorithm used takes the RMS of each 50ms window in the file, ranks the windows from loudest to quietest, and then takes the 95th percentile of these as the 'peak'. Note that no adjustment is made for humans' differing perception of different frequencies.

 

You can also contact NCH support at > supportnch@nchsoftware.com

 

Best regards,

Link to comment
Share on other sites

Thank you Telephony for the quick response!

 

Read through the manual and have experimented with the Normalizing options you mentioned. It is definitely the right way for matching the volume of tracks together, I agree. So that does answer my question in part.

 

However, I am very interested in the pumping effect that occurs from too much volume increase. It happens whether using Amplify, Normalize, Dynamic Range Compressor, or any effect that has the possibility of producing too much gain. The way I understand it is if the audio gets loud enough to exceed the upper digital ceiling (the max decibel full-scale quantisable values in digital) it must either TRULY hard clip, creating nasty, flat line digital distortion (not pleasing in the least since 10 to 30 samples might get distorted in a row for example!) or go through some form of "fudge-factor" of internal processing to save the listener from such harshness. That "fudge factor" could be

implemented in various ways.

 

Is it a "brick-wall" limiter? Brick-wall limiters have a release time (although the release time is set very near zero but never actually zero) in order for them to know when to release their limiting effect so they are never truly a complete "brick-wall." If it was a true "brick-wall", it would just clip instead of limit like a typical limiter circuit would.

 

Is it a look-ahead limiter? Look-ahead limiters delay the signal to see what is coming up so that transients can be caught by the limiter but in order to prevent clipping they "contour" the approach curve.

 

These both would explain the pumping effect if the volume is over the digital ceiling since both do what they can to prevent actual clipping. I've tested this out by amplifying the audio to insane saturated levels. I still never experience any obvious digital clipping but a pumping effect.

 

It would be spectacular to be able to change the characteristics of what happens when audio hits the digital ceiling so that the user can decide whether to introduce clipping or limiting effects into the signal or not. The characteristics could have different algorithms to choose from for desirable effect. In fact, I've read that some professional mastering engineers will clip digitally slightly in order for recordings to compete in the loudness war. Even though there is digital

distortion it is more pleasing than to soften the excitement that a music mix might have with limiters.

 

So what is occurring behind the curtain exactly? I just would like to know more about what WavePad actually does by default when processing audio that is too loud.

 

Thanks in advance.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...