Topic: what causes this waveform? one side "higher" on EACH channel

[willndmb]

does anyone know why the "top" is higher then the bottom?
I have never noticed this before.
You can't hear anything different or weird
both channels are doing it so I figure its not the fact that one mic saw slightly farther from the sound source but I guess maybe it could be
just curious
tia

[page]

If you weren't recording something with a brass instrument in it, then my next guess would be for DC offset somewhere in the signal chain or something with the PA stacks.

[willndmb]

If you weren't recording something with a brass instrument in it, then my next guess would be for DC offset somewhere in the signal chain or something with the PA stacks.

bingo, brass
the recording was made on a on stage fill and was positioned about was far from the bass section of the speaker as could be
still has enough bass in the recording for my tastes but I see what happened now
thanks

Edit, bRass not bass

[jagraham]

What rig were you running?  Based on your sig, it shouldn't be the case, but I've had similar results running CA mics into handhelds with plug in power.  I just assumed it had to do with the insufficient power being provided.

[willndmb]

What rig were you running?  Based on your sig, it shouldn't be the case, but I've had similar results running CA mics into handhelds with plug in power.  I just assumed it had to do with the insufficient power being provided.

http://www.thetradersden.org/forums/showthread.php?t=101813
soundpro mini omni mics in hat > tb via pip/low setting > m10 line in

[George]

I had a similar wave for two bands I taped recently.  I was close up to the on stage PA and at loud volumes (during the loudest parts of a song) I can hear some brickwalling. 

[DigiGal]

Understanding DC offset <<-- Source Link
DC(Direct or continuous current) Offset is good for changing the baseline of an audio file (e.g. wav) to compensate for electrical mismatches between your audio card and the input device, thus obtaining a louder amplitude after normalization. If DC off set exists, getting more loudness in the mix maybe a problem. 

DC offset occurs when hardware, such as a sound card, adds DC current to a recorded audio signal. This current results in a recorded waveform that is not centered around the baseline (-infinity). In some cases, when DC-Offset is present, glitches and other unexpected results can occur when sound effects are applied to the audio file. You can compensate for this DC offset by adding a constant value to the samples in the sound file.


In the example above, the red line represents the baseline. The lower waveform exhibits DC offset; note that the waveform is centered approximately 2 dB above the baseline.

Technical Explanation
DC offset is an offsetting of a signal from zero. The term originated in electronics, where it refers to a direct current voltage, but the concept has been extended to any representation of a waveform. DC offset is the mean amplitude of the waveform; if the mean amplitude is zero, there is no DC offset.

DC offset is usually undesirable. For example, in audio processing, a sound that has DC offset will not be at its loudest possible volume when normalized (because the offset consumes headroom), and this problem can possibly extend to the mix as a whole, since a sound with DC offset and a sound without DC offset will have DC offset when mixed. It may also cause other artifacts depending on what is being done with the signal.

DC offset can be reduced in real-time by a one-pole one-zero high-pass filter. When one already has the entire waveform, subtracting the mean amplitude from each sample will remove the offset. Often, very low frequencies are called "slowly changing DC". While not technically accurate, a highpass filter can remove such a "changing offset" better because its cutoff does not extend to as low a bandwidth as the above method.

When describing a periodic function in the frequency domain, the term DC coefficient or DC component refers to the mean value of the waveform (possibly scaled according to the norm of the corresponding basis function of the frequency analysis filter bank). The name comes from the middle 20th century design of electrical line codes for use with transmission channels unable to transmit a DC voltage or current. In such usage, this coefficient represents the useless DC, whilst the coefficients representing various other frequencies are analogous to superimposed AC voltages or currents.

A waveform with a zero DC component is known as a DC-balanced waveform. DC-balanced waveforms are useful in communications systems, since they can be used on AC-coupled electrical connections to avoid voltage imbalance problems between connected systems or components. For this reason, most line codes are designed to produce DC-balanced waveforms. The most common classes of DC balanced line codes are constant-weight codes and paired-disparity codes.

[dallman]

^^Great explanation!! Very helpful as I have struggled with this occasionally when copying live music on old cassettes ^^

[DSatz]

DC offset is one thing; asymmetry in your waveforms is another. The example shown in the above article has some of both!

Real-world sound isn't always symmetrical, so it doesn't necessarily indicate a technical problem if you correct for DC offset but still find that the peaks in one direction (usually positive if you're recording in correct absolute polarity) are a little higher than the peaks in the opposite direction. In fact, with some sound sources it would indicate a problem if your waveforms were perfectly symmetrical!

--best regards

[rhinowing]

this looks very similar to the results I'd get when forced to run the old R-09 below unity gain due to overwhelming stage volume

[jj69]

I realize this topic is a few months old, but I just experienced the same result.  Weird looking waveform in Audition.  On my recording, the bottom has more energy than the top. 

It was a VERY loud hard rock show, and I had to turn the input on my PCM-M10 down to 3, which I believe is a bit below unity gain. 

Did the volume cause this?  Is there a way to "fix" it? 

[jj69]

Just to follow up on this, I just found a feature in iZotope RX2 that appears to correct this.  It's called Vari-phase rotation. 

Not sure if this was the correct tool to use, but it seems to have done the job. 

But I would still like to know what caused this in the first place. 

[bombdiggity]

^ As Dsatz notes above I'd not be sure that phenomenon needs "correcting".  Only if the recording did not sound like the show I'd think... 

At least one of my best sounding recordings looks like this (due to trumpet, which produces that sort of look, especially when soloing nearly unaccompanied for long stretches). 

[stevetoney]

I'd also add a clarification...Page was the first responder when the OP asked this a few months ago and he pointed out that this phenomenon is seen on waveforms with BRASS in the music, not bass...which it appears that the OP may have originally thought he meant.

[jj69]

Just to clarify, what I recorded is definitely not brass, and not limited to a specific portion of the perfromance.  This was a heavy rock band playing an hour set at excruciating volume.  The entire recording is asymmetrical in the negative direction.

I just found this article:

http://www.soundonsound.com/sos/may13/articles/qanda-0513-1.htm

The last paragraph seems to suggest I was on the right track with the "Vari-Phase Rotator" function in RX2:

"... natural waveform asymmetry cannot be ‘corrected’ with a high-pass filter, and a rather more complicated solution is required called a ‘phase rotator’. Generally, there is no need to ‘correct’ a naturally asymmetrical signal, but occasionally the asymmetry can restrict how much the signal can be amplified because the stronger half of the waveform will reach the clip level before the weaker side. By using a phase rotator process to alter the harmonic phase relationships, a more balanced symmetry can be established, allowing slightly more gain to be applied before both sides reach the clipping level at the same amplitude. Asymmetrical waveforms can also sometimes confuse the side-chain level-detection circuitry (or algorithms) of some compressors, resulting in less effective compression than might be expected." 

The article is referring to the"natural" asymmetry, as found in brass instruments, but I'm assuming the same "correction" procedure applies to a loud performance recorded at below unity gain. 

[jj69]

If you have clipped your microphones that will also cause asymmetrical waveforms.  That cannot be corrected by a phase rotator.

The recording does not seem to be clipped. 

[jj69]

Here is a two minute sample.  This is completely unedited, not even normalized.  Recorded at 48/24 with AT933 cardioids with a Church 9V battery box into a Sony PCM-M10. 

http://depositfiles.com/files/z4z2nxsxn

[jj69]

You mentioned something I should have pointed out.  The sound engineer was indeed a fill-in guy.  In the past, the sound has been excellent, and very, very loud, but comfortable if you're wearing earplugs. 

At this show, the sound was even louder than normal, and although I was wearing my foam earplugs, I kept subconsciously reaching up to be sure they were still in properly because it was so loud.  The sound in the room was not necessarly bad, but not as good as when the usual engineer is working the board. 

The vari-phase rotator in RX2 corrects the shape/appearance of the wave form, but I could not detect any audible change in the sound. 

Do you hear specifific frequencies that need attention? 

[jj69]

Thanks for the feedback and input. 

I'm not sure the recording is really worth investing a whole lot of time in.  Ironically, I was using no bass rolloff at all at this show. 

[Gutbucket]

I could not detect any audible change in the sound.

Phase rotators are designed that way; typically they are allpass filters which don't change frequency response.

They are also used in surround matrix encoding.  Surround channel info is phase rotated before being summed with the main L/R channels.  It's that phase difference which is used to 're-extract' the surround signals when decoded. You might notice some intersting effects if you play back your phase-rotated file through a home theater reciever and switch the Dolby/DTS/Logic7/Circle-Surround matrix decoder in/out (or you may not, since your rotated file isn't being summed with a non-rotated reference signal)

[jj69]

Unfortunatey, I just made another recording like this.  It's not as bad as the sample I posted, but the negative side of the wave is slightly larger than the positive side through the duration of the 75 minute recording. 

Now I'm starting to wonder if there's something wrong with my mics.  Can loud PA systems cause this?  These are all hard rock shows at high volume. 

Should I be using the "vari phase rotator" function to "correct" this, or is there a better tool to use on these recordings? 

This really has me concerned now.  I'm worried if a loud show damaged my AT933's (or the card capsules)? 

[willndmb]

I'd also add a clarification...Page was the first responder when the OP asked this a few months ago and he pointed out that this phenomenon is seen on waveforms with BRASS in the music, not bass...which it appears that the OP may have originally thought he meant.

bRass, yes
I just had a typo above

The waveform I posted in the op was a recording with a trumpet and trombone
It sounds great and after reading pages post I figured that was the answer FOR ME and THIS recording

[jj69]

Thanks for all the input on this.  I just stumbled onto this thread from a couple of weeks back, and it has all the answers:

http://taperssection.com/index.php?topic=162851.0

I was using the exact same mics (AT933) and my result is exactly the same.  The volume was too much for the mics, and it's soft clipping. 

I've gotta remember to switch to my AT943's with resistor mod for the louder shows. 

[Stagger]

While it is a different case from what turned out to have happened here, I can get that result sometimes with the c34. In this case it is because of the way that the S42 forms the patterns from the 4 diaphragms (2 for each channel). As you move between adding and subtracting different amounts of output from each cap to simulate directionality, you can wind up offsetting the track. It only happens in certain cases and it took me a long while to figure out how it was happening but it can.

[JimmieC]

The other night, I had this issue with the soundboard channels.  The microphones channels are fine.  Was using soundboard tape out (RCA) to DR-680 (XLR) with kind of an old RCA to XLR cable (over 10 years old), may be should get a new one made up.  Using Audacity to process the files and DC offset doesn't help.  Makes it hard to apply a little hard limiter and normalize.  No brass members in the band.