Hot audio without being in the red?

NorBro

NorBro

Major Contributor
Is it possible for audio to be hot/distorted, but software shows levels to be only barely in the yellow here or there. There are no red peaks. We recorded some stuff and the levels in the camera showed the audio was around -12dB, but it appears to have been recorded lower and then some parts (the pops) are hot/distorted for a millisecond here or there, but the levels don't show red.

This is the first time this has happened to me. Any thoughts? Thank you.
 
I'm not an expert on audio so I'm not about to try and answer but I'm sure the experts on here will want a lot more info. What equipment are you using and what was your setup. Give them something to work with.
Bill Wojo
 
Assuming youre plugging mic direct into cam, mic itself could be getting overloaded--had this happen once with an me66 in a small room where actors were shouting. Levels were fine, audio was NOT.

otherwise i would say something earlier in your chain is getting overloaded (mixer, transmitter, receiver, if any of those are applicable). If all of that isnt set up right then the cam/recorder levele are irrelevant.
 
Thank you...I'm using an HPX250 with a Sound Devices MM-1 pre-amp. My microphone is the Sennheiser MKH-416. Everything is straightforward as far as the connections. I have my internal microphone off in the camera. I'm going into the camera via LINE (the camera is set to LINE). INPUT 2 - and Channels 1 and 2 are active from there. I'm using phantom power from the preamp. The levels are bouncing where they should be in the camera, but in FCP X, everything is much lower and some parts are hot when they are barely in the yellow (if they even reach the yellow). And I hope hot is the right word here, but it basically sounds like static distortion for a second. It's too loud.
 
What was the phantom power setting on? Were the batteries were fresh in the MM1? 416 mics often suffer from low volts distortion.
On Outside Broadcasts with long cable runs this is often a problem with 416 / 816 mics. (they need to be remote powered)
 
Batteries were brand new and phantom was 48V.

OzGizmo said:
On Outside Broadcasts with long cable runs this is often a problem with 416 / 816 mics. (they need to be remote powered)
Click to expand...

What do you mean by this? Because it was an outside shoot (in the cold) and a long XLR cable was used...
 
Sometimes on LARGE sporting OB's (golf, motor racing, etc) often the mic is a long way from the mixer powering the mic (100m+) and with the voltage drop on the cable is enough to go below the mics voltage requirements. MKH 416 is 48v +/- 4v, MKH 60 48v +/- 4v, Rode NTG3 44-52v... some mics will operate from 12-52v phantom.
With long cable runs mics are often powered at the mic end and then sent to the mixer 'line' level or even 'high mic' level.

If your mic cable as less than 100m (330ft) in good condition (indoors or outdoors) this would NOT be the problem in your situation.
Old or corroded connectors will cause the same problem on shorter length cables.
 
Couple of things.

You could be distorting the mic, or the preamp, before hitting the camera, so all of that is possible. Low phantom power or low batteries generally will lower the point the mic and or preamp distort at so that is why that could be an issue. A faulty cable could be a problem. If one conductor has a bad connection it could be unbalancing the phantom power, thus giving you a lower level. Oxide dirt etc on connections can give some resistance. With a hot enough signal it can "arc through" the oxide/ dirt and give you a sudden increase in volume. The last is also usually accompanied by problems at all levels so I doubt it's your problem but? Swapping a cable is always a good first move since they are the most likely things to fail.

Then alass we come to FCP. Some companies don't believe in letting you have control or know what the heck is going on so FCP uses floating point math on the sound files meaning it will magically readjust so it can never clip. The problem with the behind the scenes magic is that you are not seeing what the sound file looks like but a reimagined version of that file. So you can't trust the meters in FCP to tell you what is really happening. Older versions set the zero low so that there would be extra headroom. Used to drive people nuts because levels would change on export.

Another problem has to do with metering in general and the difference between how digital meters mostly read and the way digital is converted to analog. The jist is that anything over -3dBfs on a conventional digital meter could clip the D? converter.

And the second metering issue is the US uses averaging meters. So a really fast transient will read lower than it actually is.

Oh and there is an oldie but goody. You could have a DC offset in the file. I haven't seen this one creap up in awhile but it used to happen fairly often in early digital days. Classically you would see it as the waveform centering above or below the center line.

I'm not sure if this will be visible in FCP but pull one of the sound clips into a sound editing app and see if it's offset. Or possibly if FCP has a DC offset filter give it a try.
 
Just a thought, but you say camera levels are right, but what about the preamp? It's very easy to have the gain up too high there and then all the camera is showing is level, but the distortion in the signal wouldn't be noticed.
 
Theoretically, optimum (digital) performance is -0.00. Lots of other variables in play for sure including the accuracy of meters. (and the D/A converters after the fact)
 
paulears said:
Just a thought, but you say camera levels are right, but what about the preamp? It's very easy to have the gain up too high there and then all the camera is showing is level, but the distortion in the signal wouldn't be noticed.
Click to expand...

I'm thinking this was the exact problem...but then how can you control this? And what would be the most you can use to be safe? It goes up to 66dB.

EDIT: I'm also thinking it was the cold temperature that drained the pre-amp's battery and caused the malfunction.
 
Last edited:
Rick R said:
Theoretically, optimum (digital) performance is -0.00. Lots of other variables in play for sure including the accuracy of meters. (and the D/A converters after the fact)
Click to expand...

Well he is starting analog and hearing the sound (analog). any discussion of digital audio needs to include at least the D/A conversion since you can't hear digital audio. Optimum "bit performance" would be -0.0 but since we have the necessity to output analog that has to be taken into account.

IF your signal peaks at -0.0 and that peak is more than one sample wide you will have distortion, because the analog wave crest is beyond zero (and since there is nothing beyond zero is clipped). You have to peak at -3 or lower before you will be safe from conversion clipping.

And since conversion is not optional the "optimal" reproduction (bits and sonic integrity) is -3dBfs.

But that is output. On input if you don't have a peak reading meter you basically have to play it by ear since an integrating meter becomes less and less accurate as transients get faster. For inputs SD's excellent limiter function is your best friend if you want to record hot.
 
This is a side question, but if I'm in an uncontrollable environment of sound, would it be wise to switch my camera's levels to auto?

My understanding is that many don't use auto levels because the camera increases the noise floor when there is no audio (almost like hunting for sound), but since I'll be using a preamp the noise floor will be low from the beginning. Is this correct thinking?
 
So are you using 'auto' on the preamp or 'auto' on the camera?
As most preamps don't have auto it will be the camera preamps / gain that will be doing the 'auto-ing'
 
Yea, the camera's auto levels. The preamp doesn't have auto. Normally, I use 28dB or 36dB of gain on the preamp and manually adjust the camera's levels for some fine tuning, but I'm thinking of just letting them run automatically in certain situations. No matter what, the camera's levels have to be adjusted somehow...I can't turn them off and just use the preamp.
 
you WANT to use auto IN THE FUTURE, but DID NOT used auto now ?

Using auto is always a bad idea.....always..... really always ? yes always....
 
NorBro said:
Is it possible for audio to be hot/distorted, but software shows levels to be only barely in the yellow here or there. There are no red peaks. We recorded some stuff and the levels in the camera showed the audio was around -12dB, but it appears to have been recorded lower and then some parts (the pops) are hot/distorted for a millisecond here or there, but the levels don't show red.

This is the first time this has happened to me. Any thoughts? Thank you.
Click to expand...

OK - quite a few cameras have the level control *afrter* the first amplification stage. This means that it is possible to overload the mic. pre-amp. *before* the sound gets to the level control. In this instance you will be recording distorted sound at a level that shows OK.

If you were recording loud sounds, this is the most likely.

Also - what were the meters? VU meters (termed "Virtually Useless" in the UK) were designed for old analogue recording systems and show an average level. You really need to monitor on peak reading meters.

Ideally you should record using a peak limiter that will prevent overloads (do *not* use ALC, it's horrible).

I hope this helps.
 
Were your limiters in the HPX250 on? That will prevent the audio from peaking in the camera, but it will also distort it should there be a loud burst. I'm not familiar with the MM-1, does it have tone for you to match the output to the 250s input?
 
Not sure but I don't think the MM1 has tone. Truthfully I don't use tone a lot with the MixPre. The SD has really good limiters and the MM1 I believe are the same as the MixPre. I set the recorder so I can't clip. I will scream into it or whatever you can do to get the limiters to engage. You then set the recorder so that it's not clipping when the limiters are engaged and you are essentially clip proof. I do use the tone to check the recorder level because I have noted whee the tone sits when I'm at the clipproof position. I do a lot of FX so I will often adjust the recorder for some things where I don't need to worry about clipping but it's great to be recording an unknown and know that the worst your going to get is some limiting as oposed to digital distortion.
 
You must log in or register to reply here.
Back
Top