HELP! SD302/442 sync level issue

[Hiro Nakamura]

This is a problem that has happened lots of times, I just wanna know WHY!
(It happened on both my mixers, 302 and 442)


The problem is :
After I sync my tone level with a camera (it happened to all the cameras, Panasonic HPX3000, SONY F900R, recently Alex Alexa)
I level the camera with the 1k tone to -20db. BUT what happened was....the mic getting the audio in to my mixer, which I keep the level around 0db on my mixer, BUT the camera receiving the audio 10db higher.
In the pass, I didn't bother, after I leveled the camera, if this problem happens, I just re level the tone to -30db, then I could still use my mixer to mixing the audio level around 0db, so... it will input to the camera around -20.
BUT I just wonder why it's like this!!!


Recently I was on an arri shoot, the level is always changing, first day I leveled the tone to the camera -30, like the problem, it gave 10db louder , when I switched back to my 416...but the first then the level got back to the normal level, which after like 2 hours mixing, I noticed everything that going into the camera is 10db higher ....which is around -10db, makes a bit easy to clipping..


That frustrated me .... anyone could help me to explain this ;(


Thanks a lot in advance guys!


Hiro

 

[Rippie]

Are you sure your mixer meters are in PEAK mode?
Sounds like your mixer meters might be in VU which would give you the results you mention.

The only metering mode I use is VU/Peak together. (the non "hold" version which holds the peak signal on the meters for a much longer period)

This way you can see your average signal but also the peaks at the same time.

 

[Steve House]

The problem is :
After I sync my tone level with a camera (it happened to all the cameras, Panasonic HPX3000, SONY F900R, recently Alex Alexa)
I level the camera with the 1k tone to -20db. BUT what happened was....the mic getting the audio in to my mixer, which I keep the level around 0db on my mixer, BUT the camera receiving the audio 10db higher.
In the pass, I didn't bother, after I leveled the camera, if this problem happens, I just re level the tone to -30db, then I could still use my mixer to mixing the audio level around 0db, so... it will input to the camera around -20.
BUT I just wonder why it's like this!!!


Recently I was on an arri shoot, the level is always changing, first day I leveled the tone to the camera -30, like the problem, it gave 10db louder , when I switched back to my 416...but the first then the level got back to the normal level, which after like 2 hours mixing, I noticed everything that going into the camera is 10db higher ....which is around -10db, makes a bit easy to clipping..

What you are seeing is the result of the different energy distribution in a sine-wave envelope versus a complex envelope such as human speech. With a sine-wave, the reading of an averaging meter and a peak reading meter will be the same. With speech, an envelope with an RMS -20dBFS average level has peaks about 8 to 10 dB higher. Thus what you are seeing is perfectly normal - after calibrating the camera levels to -20dBFS while sending it a 0dBVU sine-wave tone and switching to speech also reading 0dBVU on the mixer's meters, the camera is receiving an average signal level of about -20dBFS which indicates as about -12dBFS on the camera's peak-reading meters. Do NOT try to compensate by reducing the calibration point to -30dBFS, you're recording too low and raising levels in post will bring up the noise as well. With -20dBFS as the calbration point and keeping speech hoving around 0dBVU, you have about 12 dB of headroom before clipping is at risk. Make sure you engage the mixer's limiters, that's what they're for.

 

[Hiro Nakamura]

Do NOT try to compensate by reducing the calibration point to -30dBFS, you're recording too low and raising levels in post will bring up the noise as well. With -20dBFS as the calbration point and keeping speech hoving around 0dBVU, you have about 12 dB of headroom before clipping is at risk. Make sure you engage the mixer's limiters, that's what they're for.

The problem that I am having is that if I calibrate the camera to -20db with 302's 1k tone, after turn back the microphone mode from the 1k tone, the calibrated level on the camera automatically switch 10db highter, which means the level I just leveled (-20db) become -10db when I use the microphone. when I read from my 302 the peaking is around "0", but on the camera is on -10 ( sorry about my english, hope it makes sense) then I only have another 9 db to work with ....

I talked to Sound Devices, Stephen told me that there might be some padding cable in between ....

Any ideas?

 

[Rippie]

You say you're "peaking on the 302 at 0"
but are you 100% sure the meter is showing PEAK? (there are 4 USER choices you can go between)

It would totally explain whats happening to you perfectly and your audio mixer isnt magically switching after sending tone, and the camera isn't magically switching
either so this really by FAR what you should confirm or not.

If you can 100% be sure after flicking the peak/vu button on the 302 that its definitely not that we can start narrowing it down from there but
please confirm this.

 

[Steve House]

The problem that I am having is that if I calibrate the camera to -20db with 302's 1k tone, after turn back the microphone mode from the 1k tone, the calibrated level on the camera automatically switch 10db highter, which means the level I just leveled (-20db) become -10db when I use the microphone. when I read from my 302 the peaking is around "0", but on the camera is on -10 ( sorry about my english, hope it makes sense) then I only have another 9 db to work with ....

I talked to Sound Devices, Stephen told me that there might be some padding cable in between ....

Any ideas?

You are seeing an artifact of the camera's peak reading meters and it is perfectly normal. Imagine you have a meter that you can switch between average and peak reading. You set it to averaging and send it a sinewave tone, adjusting the level to read -20dBFS. If you then switch the meter to peak reading, it will still read -20dBFS, exactly the same. For a pure sinewave tone, the average meter and the peak meter read identical levels. But now try the same experiment with pink noise (which has much the same energy distribution as does speech). Set the meter to averaging, send it pink noise, and adjust the level so the meter reads -20dBFS. Then switch the meter to peak. You'll see it immediately jump up to about -10dBFS even though the actual level of the signal hasn't changed at all.

If you have an NLE that has a tone generator and switchable meter ballistics you can actually see this in action. Record a -20 dBFS sine wave. Play it back with the meter set to averaging and adjust the output to -20dBFS. Switch the meter to peak and play the tone without touching the levels. It will still read -20dBFS. Now try it with pink noise. Record with the level set to -20dBFS. Play it back with the mter set to averaging and adjust the levels so the meter reads -20. Without touching the levels, switch the meter to peak reading and play the noise file again. This time it will read about -10.

The upshot is if you send a 0VU tone from the mixer to the camera and adjust the camera recording level to -20 on the camera meters, switching to speech that also reads 0VU on the mixer meters will show a reading of -10dBFS (actually they're probably closer to -12) on the camera's peak meters. This is as it should be and is the right level to record at. Reducing the camera levels like you say you're doing means the voice PEAKS are at -20 when they OUGHT to be at -10, you're recording 10 dB TOO LOW.

The correct procedure is to turn on 0VU tone from the mixer and set the camera input levels so the camera meter reads -20. Lock down the camera input level controls so they won't move. On speech, adjust the mixer channel trim and fader controls so the speech averages between 0VU and +4dBVU on the mixer meter. The camera meters will be bouncing up around -12 to -10 dBFS amd that's exactly where they should be.

You can set the mixer meter to show BOTH average and peak level at the same time. If you feed it speech and set the levels so the continuous line of LEDS showing the average level reads about 0VU, the single LED indicating the peak will be about 8 to 10 dB higher.

 

[Hiro Nakamura]

You say you're "peaking on the 302 at 0"
but are you 100% sure the meter is showing PEAK? (there are 4 USER choices you can go between)

It would totally explain whats happening to you perfectly and your audio mixer isnt magically switching after sending tone, and the camera isn't magically switching
either so this really by FAR what you should confirm or not.

If you can 100% be sure after flicking the peak/vu button on the 302 that its definitely not that we can start narrowing it down from there but
please confirm this.

Thank Rippie for your input, I am pretty sure that the VU meter I see is alright, of course I didn't use the one you mention, I used the metering setting, I think it's the one, it 's like the old VU meter, it only shows once when the audio input in.... I kept the level around -6 to 0, so... even the VU meter that I looked at it is wrong setting, it should not be clipping, that's like 20db higher to get clipping isn't it?

hiro

 

[Hiro Nakamura]

The correct procedure is to turn on 0VU tone from the mixer and set the camera input levels so the camera meter reads -20. Lock down the camera input level controls so they won't move. On speech, adjust the mixer channel trim and fader controls so the speech averages between 0VU and +4dBVU on the mixer meter. The camera meters will be bouncing up around -12 to -10 dBFS amd that's exactly where they should be..

Thanks Steve for your help. I have different opinion of looking at the VU meter. I think the reason why I line up the level with the camera to -20, is to give 20db buffer before the audio hits the clipping level whichi is "0" on the camera. So...after I line my tone up with the camera, when I work on the mixing, I would raise the audio like primary audio (speech) to around "0" on my mixer, so technically, Camera will get around -20 on the camera. In this way, it's easy for me to adjust the levels. because if the audio level on my mixer is hitting 0, then I know how much audio level that the camera getting. IF I wanna the camera getting 10 higher, I could level up the 1k tone to -10 on the camera, then I know that my 0 is -10 on the camera.

Sorry again about my broken english, hopefully it makes sense.

Thanks a lot Steve for your help

Hiro

 

[Hiro Nakamura]

If you can 100% be sure after flicking the peak/vu button on the 302 that its definitely not that we can start narrowing it down from there but
please confirm this.

Hey Rippie,
I realized that sometimes I did mistakenly pressed the VU/Peak button. Then I will change back to my preferred option, which is the one like the old VU meter, it only shows the audio in and out. But just now I am checking the 4 options of VU/Peak, there seems to have two of the meters which are similar as the old VU meter style, but one of them is showing higher level, another is showing lower......I might turn the lower level option on. then I boost the audio higher, on my mixer it seemed to have 0, but actually it's 10 higher....

I will have a test now ... to find out

Thanks a lot for your help.... thanks all of you guys.

Appreciate!!!

Hiro

 

[Steve House]

The problem that I am having is that if I calibrate the camera to -20db with 302's 1k tone, after turn back the microphone mode from the 1k tone, the calibrated level on the camera automatically switch 10db highter, which means the level I just leveled (-20db) become -10db when I use the microphone. when I read from my 302 the peaking is around "0", but on the camera is on -10 ( sorry about my english, hope it makes sense) then I only have another 9 db to work with ....

I talked to Sound Devices, Stephen told me that there might be some padding cable in between ....

Any ideas?

Your average signal level in the camera did NOT change when you switched from tone to microphone. If the mixer's meter is showing 0VU on speech, the average signal level of the speech coming into your camera is still -20 dBFS in the camera if you have calibrated 0VU tone to read -20dBFS on the camera meter. What makes it SEEM like it's 10dB higher is you're reading peaks in the speech that don't exist in the tone. The peak level is -10 but the average level is still -20. You have 10 dB of headroon which is more than enough for normal speech.

Hey Rippie,
I realized that sometimes I did mistakenly pressed the VU/Peak button. Then I will change back to my preferred option, which is the one like the old VU meter, it only shows the audio in and out. But just now I am checking the 4 options of VU/Peak, there seems to have two of the meters which are similar as the old VU meter style, but one of them is showing higher level, another is showing lower......I might turn the lower level option on. then I boost the audio higher, on my mixer it seemed to have 0, but actually it's 10 higher....

I will have a test now ... to find out

Thanks a lot for your help.... thanks all of you guys.

Appreciate!!!

Hiro

You really need to study up on the differences between an averaging meter ballistic (the old VU meters) and a oeak reading meter (your camera meter, for one). On normal speech waveforms, peaks are about 8 to 10 dB above the RMS average level. On a speech signal, the peak meter will read about 10 dB higher than the average meter for the same exact actual signal level. The averaging meters simply don't respond fast enough to indicate the peaks that the peak meter shows.

Why do you want 20 dB of headroom? The mixer's limiters kick in well befre that.

 

[berniebeaudry]

Hello Hiro,
Steve and Rippie have done a great job explaining the difference between peak and vu metering. I have both the 302 and 442 and I always use the peak/vu, or peak setting on my meters. You can't change the metering on the camera so why not change it on the mixer so they come close to reacting the same way? That way you can be confident you're sending the camera a good level. Not too high, not too low. Sometimes levels that are too low can be more of a problem than you might think. The limiter on both mixers work very well. I've set mine at 17 which is 3 down fromd 20 which is the highest the meter can go. 20 on the mixer would correspond to 0 on the camera. Working this way its almost impossible to send too much level to the camera assuming your inputs are trimmed properly.
Bernie

 

[Rippie]

Remember what VU stands for!


Virtually useless.

There is absolutely zero reason why you shouldnt use the VU/Peak setting on the sound devices equipment.
If you traditionally mix using average (non peak) signal you can see it..
If you traditionally mix using peak meters you can see it.

No brainer.


Could be worse though.. I used to own a Shure FP42 back in the day.
Those VU meters were so ridiculously slow that unless the person was screaming for a few seconds the meters wouldn't move more than a few millimeters,
but at least the mixer also sounded like crap

 

[Hiro Nakamura]

Your average
Why do you want 20 dB of headroom? The mixer's limiters kick in well befre that.

I thought the TV station keep the audio level around -18 to -20 on the HD tapes... Is that right ? or I am completed wrong on this?
I think if I give a buffer like 10db, that makes me worried about clipping ...

Correct me if I am wrong please

Thanks Steve

Hiro

 

[Hiro Nakamura]

Remember what VU stands for!


Virtually useless.

There is absolutely zero reason why you shouldnt use the VU/Peak setting on the sound devices equipment.
If you traditionally mix using average (non peak) signal you can see it..
If you traditionally mix using peak meters you can see it.

No brainer.


Could be worse though.. I used to own a Shure FP42 back in the day.
Those VU meters were so ridiculously slow that unless the person was screaming for a few seconds the meters wouldn't move more than a few millimeters,
but at least the mixer also sounded like crap

Thanks so much Rippie for your help.
I am gonna change to VU/Peak mode in the future.

Thanks...Btw the wireless works great on my shoot;D

Hiro

 

[Steve House]

I thought the TV station keep the audio level around -18 to -20 on the HD tapes... Is that right ? or I am completed wrong on this?
I think if I give a buffer like 10db, that makes me worried about clipping ...

Correct me if I am wrong please

Thanks Steve

Hiro

Note that ALIGNMENT level and RECORDING level are two different things. Also the level you record at in the field is not necessarily the level it ends up being in the final mix.

Standard alignment levels are ...
North America: -20dBFS = 0VU = +4dBu
EBU: -18dBFS = 0VU = 0dBu

Most HD broadcast specs call for average dialog levels, averaged over the entire program ("dialnorm"), to be about -24dBFS, peaks on music and SFX can reach up to -3dBFS or so. But this is for the final mix ... for recording the original location master, to get the best signal/noise performance, you want to record as hot as possible without clipping.

 

[Harlockr]

Pretty informative Steve, I also was wondering if I was doing something wrong while calibrating my mixer with the cam : ) Now I get a more technical reason, and everything makes more sense.
Thanks guys, I just thought would be nice to let you know that the responses helps a lot more people

 

[Hiro Nakamura]

Note that ALIGNMENT level and RECORDING level are two different things. Also the level you record at in the field is not necessarily the level it ends up being in the final mix.

Standard alignment levels are ...
North America: -20dBFS = 0VU = +4dBu
EBU: -18dBFS = 0VU = 0dBu

Most HD broadcast specs call for average dialog levels, averaged over the entire program ("dialnorm"), to be about -24dBFS, peaks on music and SFX can reach up to -3dBFS or so. But this is for the final mix ... for recording the original location master, to get the best signal/noise performance, you want to record as hot as possible without clipping.

Thanks so much Steve for your help.
Appreciate!

Hiro