Audio from camcorder sucks!

[lalkana]

The audio captured by both my DV and digital 8 camcorders seem to be very bad. The sound on the DV is usually low, while the D8 seems to pick up a lot of background or machine noise. This means my DVD's have great sound on the menus and background music as well as the slideshows, while the sound on the video sucks. Will an external mic improve things and/or is there a way to clean this up before burning. I have Pinnacle Clean, but not sure the best things to tweak as I am not really a sound person.

[Suntan]

Don't own a Cam so I can't say from experience, but from what I have heard (no pun intended) a good quality directional mic will do wonders.

I would say research for a good mic. After that the next logical step would be to look at recording the audio seperately and mux together afterwords. A Sony MD recorder (about $100 to $150) will record at CD quality (depending on the ADCs used) which you can then mix with the video in post production.

-Suntan

[anubis13]

i've dealt with this in broadcasting school
that's a common problem. most built in microphones suck. an external mic will help, depending on the mic. i like a battery powered condenser with dual pickup for stereo right off.

low priced one
http://www.musiciansfriend.com/srs7/sid=040809135224024166062169390635/g=rec/s=mics/se...e_pid/271083X/

the one we used with canon XL1 at the school
http://www.musiciansfriend.com/srs7/sid=040809135224024166062169390635/g=rec/s=mics/se...se_pid/271591/

BTW what cameras are you using.

[Kurt S]

Another problem with using the audio from a camcorder is that fact that unless you have a pro model or a very high end recorder, most consumer models use compression or automatic volume gain. This can add tons of extra noise.

You might try running the audio through a descent wave editing program such as Adobe's Audition or Audacity and do some noise reduction and even some expantion on the sound track.

[troyvcd1]

I would suggest getting one that mounts directly on your camcorder and is made for a camcorder like the sony ms908c. I got this one and my audio is 10 times better then before.

[Dosdan]

Another problem with using the audio from a camcorder is that fact that unless you have a pro model or a very high end recorder, most consumer models use compression or automatic volume gain. This can add tons of extra noise.

You might try running the audio through a descent wave editing program such as Adobe's Audition or Audacity and do some noise reduction and even some expantion on the sound track.

This is my pet hate with my JVC GR-DV900 camcorder: high quality 16-bit sound combined with a viscious AVC. Useless for good-quality music and even speech recording. I have to use an external recording medium (Minidisc) and try to combine it with the video. There are two problems to consider:

1. Temporal Offset. The DV and the MD usually don't start at exactly the same time so you have to work out the offset. If the MD starts after the DV then I have to add silence to the beginning of the MD to resync; if the MD starts before the DV I have to delete the early part of the MD. I use multitrack mode in Audobe Audition for this. I look for an obviously similar point in the two wave forms. Not always easy: the two wave forms are not exactly the same because DV is LPCM and MD is lossy ATRAC3. Also since both waveforms must have the same sampling rate in a multtrack session (DV is 48K; MD is 44.1K) and since the ouput (DVD) is 48K, I upsample the MD to 48K.

2. Sample Rate difference. The crystal oscillators in the DV and MD I believe can have an accuracy of +/- .005% (50ppm). Taking a worst-case scenario, where one was fully + .005% and the other was -.005% then the relative difference would be 1 in 10,000. Over 1 hr (3,600 secs - my DV recording are typicall 50-90 mins) that means you could have a difference of +/- 360ms at the end from the beginning. This time difference on speech would be very annoying. I find a distinct part of the wavefom near the end that I can use to determine the time difference in the track times (after adjusting the temporal starting offset) and then use Audition's Time Stretching to stretch/shrink the MD to exactly match the DV duration.

What a lot of extra effort to work around a stupid design decision in consumer DV camcorders that needn't be there. It's not like adding a manual recording level control is high tech!

[Jester700]

I use the DV/MD combo as well, and over the time span I use (a 45 minute music set or similar length wedding ceremony) I don't experience offset once the tracks are synced.

Big agreement on the AGC - there should ALWAYS be the ability to switch it out, especially on cams with external mic jacks.

The difference in waveform is a lot more to do with AGC vs. lack thereof and EQ differences than it does with PCM vs. ATRAC. ATRAC is lossy, but not THAT lossy.

[Cornucopia]

My 2¢...

Cam Mounted Mic VS. External Mic

Cam Pro's:
Sound Arrives at Same Time and From Same Direction (no sync or perspective issues)
More Portable when kept all-as-one

Cam Con's:
CAMERA/RECORDER NOISE! (Hell, even $50k cams make noise!)
Cam/Eye Optimal position isn't always Mic/Ear Optimal position

External Pro's:
Wider Choice of Mic Types
Optimal Positioning if Boom/Zoom/Shotgun/WirelessLav

External Con's:
Extra Cabling/Adapters to have to deal with
You have to THINK where to put the mic and use something to sync with

You can probably guess where I lean.

You want better sound, try these out (in rough order of increasing cost and quality):

1. Get external Mic-specialized for your type of work
2. Get LowImpedance/Balanced Mic (with adapters if needed)
3. Get Mic field mixer (and listen through headphones)
4. Get Phantom-powered Condenser Mic
5. Get a Number of Mics for the various types of work
6. Get ProCamera (w/switchable Manual/AGC control) or Separate AudioRecorder (possibly MD--but I don't like ATRAC compression)
7. Get #6, but with Balanced XLR inputs so no adapters needed
8. Complete double-systems with posted ADR and Foley 8)

(These all assume you've tried all the usual things, like shooting in a quieter area, getting closer to the subject, blanketing the camera, mic/line level switches set optimally, etc)

To keep sync, get a Clapboard. and/or use 2-pop before and after.

Good test to know where to put the mic for best sound: Close your eyes and stop up one ear so you only hear in Mono.

HTH,
Scott

[Jester700]

I use a camera flash to sync 3 cams together. I never found I had any trouble syncing audio, but a clapper (or similar sharp percussive sound with visual action) would be great.

[lalkana]

Thanks all of you for the great suggestions!

I think I will try a decent quality cam mic. I do have an external mic, but it is not decent. The sound from it is worse than from the camcorder. My Panasonic has very little motor noise, so I think a mounted mic will be okay. I worked with a friend once trying to sync up audio recorded on another device. The sound is probably better, but I don't really want to go through all that. I may give it another try sometime when I'd in the mood for torture

[lalkana]

Oh, forgot to mention -- if it makes a difference my cams are a Panasonic PVDV-701 and a Sony TRV-350. As I mentioned the Panasonic sound is low, the Sony sound has a lot of background noise. I think the issue is how the gain is working (or not working). I will research the suggested mics and go from there.

[Dosdan]

Further to time difference due to slightly different clock rates when using MD with DV. I've just been working on preparing a 69 min. speaking track. Time offset at the end was 353ms which is 84 in 1,000,000. This required a time shrink factor of .999916 to ensure that the end was still in sync.

[Cornucopia]

Further to time difference due to slightly different clock rates when using MD with DV. I've just been working on preparing a 69 min. speaking track. Time offset at the end was 353ms which is 84 in 1,000,000. This required a time shrink factor of .999916 to ensure that the end was still in sync.

Dosdan,
Sounds like the crystal clocks that set the sample rates weren't all that tightly synced.
Did you have to do a 44.1 --> 48 kHz SRC and then do the Time Compression? Man, that's alot of extra work that could be avoided...
Try this:
Use an audio utility that can change the WAV headers.
Set the MD (44.1kHz) file to whatever 44100 * .999916 = (?44096?)
--This part won't take but just a minute to fix (not like ~69 min.)
Then, do 44096 --> 48kHz SRC.
Should be lined up just fine, and only 1 long conversion!

(Actually, this is another reason I don't like MD. Like most other LowestCommonDenominator Consumer stuff, their crystal clocks suck)


Scott

>>>>>>>>>>>
edit: IIRC, CoolEdit Pro can edit WAV headers

[Dosdan]

Further to time difference due to slightly different clock rates when using MD with DV. I've just been working on preparing a 69 min. speaking track. Time offset at the end was 353ms which is 84 in 1,000,000. This required a time shrink factor of .999916 to ensure that the end was still in sync.

Use an audio utility that can change the WAV headers.
Set the MD (44.1kHz) file to whatever 44100 * .999916 = (?44096?)
--This part won't take but just a minute to fix (not like ~69 min.)
Then, do 44096 --> 48kHz SRC.
Should be lined up just fine, and only 1 long conversion!

I do use CEP/Audition to convert the header of Type 2 AVIs which are actually 48K but are seen as 32K.

Your approach will work if the sample rate difference is similar for each session. More of these sessions will show me if that's so or whether it's going to be different on each occasion. Thanks for the tip.