Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums Audio Audio coming into FCP low

  • Craig Alan

    April 3, 2011 at 2:46 am

    Ty, please dummy this down a little. You’re saying that even after calibrating a camera to the mixer’s tone and not hearing any distortion when monitoring with headphones using the camera’s headphone out and the meters indicating acceptable levels, the sound that is actually recorded can be clipped and distorted? My experience is that if the sound is recorded a little low but clean I have some room in post to increase the level without adding unacceptable distortion.

    Can’t you test the camera by recording directly into the cam without a mixer and seeing in FCP if the levels match?

    I know when I used to record with sony pd170s I was careful not to let the sound get too much over -12 or it would distort. On other cameras I could be a little more generous. But I assumed this was due to PD170’s poor preamp which added noise and distortion.

    OSX 10.5.8; MacBookPro4,1 Intel Core 2 Duo 2.5 GHz
    ; Camcorders: Sony Z7U, Canon HV30/40, Sony vx2000/PD170; FCP certified; write professionally for a variety of media; teach video production in L.A.

  • Ty Ford

    April 3, 2011 at 5:12 am

    Craig,

    That’s exactly what I’m saying. I’ve heard it on more than half a dozen occasions. I’ve been recording audio since my mother bought me my first small reel to reel when I was nine. I went digital in my studio in October of 1990. You were quite safe as long as you didn’t go past 0 dB on the digital meters.

    We used to be afraid of ANY overs. Overs are anything over 0 dB. We subsequently learned that you need at least three consecutive overed samples before you could hear the clip. We also learned that clip LED didn’t always fire at 0dB. Some fire one, two or even three dB before 0 dB.

    There’s another thing that may come into play here. Some mixers can be altered to send, 0 dB, +4 dB or even +8 dB when they read 0 dB on their meters. The Sound Devices 442 is set stock at 0 dB, but it can be adjusted in the firmware. I leave it at 0 dB. Why is there +4 and +8 when the meters read 0dB? I don’t know the history.

    I send 0 dB tone from my 442 and set camera meters to -20 dB to calibrate the mixer to the camera so I don’t have to look at the camera meters. My 442 meters go up to +20. That should be right at 0dB on the camera meters and it usually is.

    Tone is not peak. It’s Average or RMS. Real program material has peaks. The 442 meters read peak. When I see peaks on the 442 meters, as I usually do, at +14 on my mixer, I know the camera meters are showing -6dB. I know this because I have watched both mixer and camera meters.

    With some cameras, even though my peaks are at -6 dB or below on the camera, I can plainly hear clipping. When this has happened one camera, when inspected, was found to be in need of repair. The others were not inspected, but they began clipping somewhere around +12 dB on the mixer, which was -8 dB on the camera. You don’t really have to be scared of this. If you listen to the camera headphone output, you’ll hear the click or sometimes a sort of metallic clink.

    When I run into this situation, I politely tell the camera op there’s a problem and he/she might want to get it checked out. In almost every case, this happens when feeding line level from the camera. And the camera is set to line level. When I switch from line to mic on both devices and recalibrate with the camera tone, the problem goes away.

    BTW, you MUST use mixer tone and not camera tone. I’ve had a couple of camera ops tell me they can use their tone. I politely explain that the reason I need to send mixer tone is so I can calibrate my mixer to the camera levels. If you use camera tone, there’s no way to know what level your mixer is sending to the camera.

    I don’t buy the thought that any properly operating cameras “get sort of distorted, but not clipped” when the levels run near but below 0 dB. Which cameras are these? Someone please toss in so we can see if we can make a list of cameras with crappy, uncalibrated audio inputs, mic or line.

    You may be right about the the pd150. Were you observing peak levels at the camera? Perhaps the problem was the POS pad that they put in front of the mic preamp in order to create a “line level” input. A proper line level input bypasses the preamp entirely, but it’s much cheaper to pas the mic preamp, so best practices goes out the window. I can see the manual, “In order to prevent mildly distorted audio, you really should even try to get close to 0 dB because we cheaped out. Thanks and have a nice day!”

    I can run up to 0 dB on my Canon XL2, but I do need to send it mic level if I use the XLR inputs. I wonder if the DVX100 had problems.

    You said, : My experience is that if the sound is recorded a little low but clean I have some room in post to increase the level without adding unacceptable distortion.”

    Why should you have to record “a little low?” Because the camera audio section is not designed properly? Well then I guess, as Peter Groom suggested earlier, you HAVE to. Maybe the camera can’t be calibrated properly. That would suck because I can buy a $199 audio recorder like the Sony M10 that can take line level to 0dB without even breathing hard. Maybe the camera was OK when it left the factory, but they used cheap capacitors that dried out, resulting in your not being able to set the level correctly without clipping and distortion. How likely are you to complain and send it back? And don’t think that wasn’t a calculation made somewhere within the company.

    When you bring up the level in post, you are also bringing up the noise floor, especially a problem if you’re recording only 16-bit audio. With a well-designed 24-bit circuit, you have a lot more to play with. Don’t get me started on HDV audio. it’s 384 kbps stereo mpeg; one fifth the data rate of 16-bit, 48 kHz,

    You asked: Can’t you test the camera by recording directly into the cam without a mixer and seeing in FCP if the levels match?

    You mean mics directly into the camera? Sure, but what are you trying to get at? Problems with the camera’s line level input or problems with the mixer? I wouldn’t want to work that way because the 442’s limiter is very good and allows me to keep my level higher above the noise with out going over.

    In summary, yes, I’m bent when I run into cameras that cheat the audio. On a practical basis, as Peter Groom suggests, we need to be aware of them. Doesn’t mean we have to like them. 🙂 However, when they are broke, they’re broke.

    Regards,

    Ty Ford

    Want better production audio?: Ty Ford’s Audio Bootcamp Field GuideWatch Ty play guitar

  • Craig Alan

    April 3, 2011 at 6:16 pm

    Wow!
    Thanks so much Ty.
    I always tell my students that if they were doing a low budget production and they had just enough coin for one professional, “hire an audio technician.”

    I no longer have the pd170s and XL2 to go back and answer your questions through testing. But I can tell you what I remember about capturing audio with them. On the xl2, like you mention, I used mic level in and calibrated my mixer as you mention and I got pretty good results. Still dependent on mic placement and talent and audio conditions. On the pd170, I think I used line level in and tried to keep the meters no more than a tad over -12 or the recorded sound would be distorted (in other words clipping).

    The advantage of the pd170 over the canon was its ability to record a really decent picture in all sorts of light. Great cam if you had no time for set up.

    Whenever I tried to ask folk about how to improve sound on the pd170, I was told to record double system or get a mod to bypass the audio circuits. I tried all sorts of different settings including line level and arrived at best I could do practices.

    There’s another thing that may come into play here. Some mixers can be altered to send, 0 dB, +4 dB or even +8 dB when they read 0 dB on their meters. The Sound Devices 442 is set stock at 0 dB, but it can be adjusted in the firmware. I leave it at 0 dB. Why is there +4 and +8 when the meters read 0dB? I don’t know the history.

    I wondered about this too. I had a shure mixer and called the tech department to ask about adjusting the output levels. The answer he gave was, “leave it alone. ” Why, I asked, would the manual go into details about changing it? His answer: If you need to ask, don’t worry about it.

    I know the camera meters are showing -6dB. I know this because I have watched both mixer and camera meters.

    So during sound checks: set the mixer levels using tone, then watch both mixer and camera meters to see if they remain in sync?

    With some cameras, even though my peaks are at -6 dB or below on the camera, I can plainly hear clipping.

    YES that was the case with brand new PD170s! I tried to keep the levels as close to -12 as possible.

    If you listen to the camera headphone output, you’ll hear the click or sometimes a sort of metallic clink.

    YES. I always got afraid when I heard that.

    When I run into this situation, I politely tell the camera op there’s a problem and he/she might want to get it checked out. In almost every case, this happens when feeding line level from the camera.

    FROM THE CAMERA??

    “In order to prevent mildly distorted audio, you really should even try to get close to 0 dB because we cheaped out. Thanks and have a nice day!”

    The unconfirmed opinion in its day was that the pd150/170/vx2000 had crappy audio in order to not have Sony take sales away from their own higher end cams.

    I can run up to 0 dB on my Canon XL2, but I do need to send it mic level.

    Yes my experience exactly.

    You said, my experience is that if the sound is recorded a little low but clean I have some room in post to increase the level without adding unacceptable distortion.”

    Why should you have to record “a little low?” Because the camera audio section is not designed properly? Well then I guess, as Peter Groom suggested earlier, you HAVE to. …

    When you bring up the level in post, you are also bringing up the noise floor, especially a problem if you’re recording only 16-bit audio.

    Yes it adds noise! Better than clipping.

    You asked: Can’t you test the camera by recording directly into the cam without a mixer and seeing in FCP if the levels match?

    You mean mics directly into the camera? Sure, but what are you trying to get at? Problems with the camera’s line level input or problems with the mixer?

    I was getting at your suggestion that the cam might not be calibrated correctly. Take the mixer out of the mix and see if the cam’s meters are true.

    In summary, yes, I’m bent when I run into cameras that cheat the audio. On a practical basis, as Peter Groom suggests, we need to be aware of them. Doesn’t mean we have to like them. 🙂 However, when they are broke, they’re broke.

    I think many of them are designed broken (cheaply).

    OSX 10.5.8; MacBookPro4,1 Intel Core 2 Duo 2.5 GHz
    ; Camcorders: Sony Z7U, Canon HV30/40, Sony vx2000/PD170; FCP certified; write professionally for a variety of media; teach video production in L.A.

  • Ty Ford

    April 3, 2011 at 6:31 pm

    [Craig Alan] “FROM THE CAMERA??”

    sorry, from the mixer.

    Ty

    Want better production audio?: Ty Ford’s Audio Bootcamp Field GuideWatch Ty play guitar

  • Ty Ford

    April 3, 2011 at 6:40 pm

    [Craig Alan] “I was getting at your suggestion that the cam might not be calibrated correctly. Take the mixer out of the mix and see if the cam’s meters are true.”

    Well yes, there is certainly the possibility the the camera metering can be off as well as or rather than the input itself, but I don’t think you need to take the miser out of the test.

    I sometimes wonder if somehow the camera people chose -10 rather than 0 for a reference. -10 is considered line level on consumer gear and that’s where a lot of this clipping seems to begin. The 442 mixer has switches for -10 reference output I have used them when feeding a consumer CD recorder with RCA jack inputs.

    I think you’d find out a lot if you put the camera on the bench and ran some real tests with good test gear.

    Regards,

    Ty Ford

    Want better production audio?: Ty Ford’s Audio Bootcamp Field GuideWatch Ty play guitar

  • Peter Sassi

    April 9, 2011 at 11:19 pm

    Steve, I just posted a similar issue above. I got an hour 2 cam corporate thing to edit with FCP. Audio was straight into the cameras. Tone on the HVX-200 is -12 which FCP loves.

    Tone on the Sony Z5 is -20 which FCP can’t hear very well. I had a terrible time bringing the -20 audio up to match the -12.

    So with so much product being edited on FCP now (used to be an Avid Guy) why is FCP set up to ‘hear’ -12 with usually no gain and -20 can be as much as 7 or 8?

    Copying and paste the normalization filter doesn’t work obviously because each piece of dialog is unique. What a pain.

    I’m open for suggestions.

    -peter

    Peter Sassi
    http://www.StoryTellersWeb.com

  • Ty Ford

    April 10, 2011 at 1:56 am

    Hi Peter and welcome to the cow audio forum.

    -20 is not a tone to listen to. It’s a tone generated by the mixer so that you can calibrate the meters on the mixer with those on the camera.

    0 dB tone leaving the mixer should be set at -20 dB on the camera of you are using any of the current location audio mixers like any of the Sound Devices mixers. You then (as the manual says) run audio levels on the mixer up to about + 14 on the mixer meters, which means they will hit about -6dB on the camera.

    Regards,

    Ty Ford

    Want better production audio?: Ty Ford’s Audio Bootcamp Field GuideWatch Ty play guitar

Page 2 of 2

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy