So in the last film I shot, I learned the AGC lesson. It makes recording decent sound to the camera, next to impossible.

Not one to leave things alone, I recalled the Beachtek coming solution, and one that Dan Chung talked about using a 1K tone fed into the camera. I tried that, and Dan was right, but the problem was when you were monitoring through a mixer, the tone masked the voice. So I realized I needed an inaudible signal.

Last night I decided to try my hand at it. I went into my copy of Audacity ( a free sound program on the net ) and found that it included a tone generator. Since Beachtek is apparently using a high frequency inaudible sound, I started making tones, until I found a point where it was inaudible. I settled on a 15,400 setting. And then I made it 15 minutes long. Next I rendered it to an mp3 file. I then loaded the file to my IRiver mp3 player, and hooked up the earphones out to the auxillary imput on my old Beachtek DXA-4., and started the player playing the tone. I ran my Sennheiser mic into the other channel, an set the imput at about two clicks below top.

I recorded voice... Result, no noise during the silent periods. AGC was shut down.

I subsequently ran the same thing through my Eng 44, plugging the mp3 player into the auxillary port where the inaudible track ran in both channels, and then recorded to one of the other imputs using the Sennheiser. A winner again.

I also found by adjusting the volume on the mp3 output, you will affect the level of gains setting in the camera. So the volume control in effect can act to raise and lower your levels.

Give it a try, you will be pleasantly surprised.

Great info thanks for posting this!


So in the last film I shot, I learned the AGC lesson. It makes recording decent sound to the camera, next to impossible.

Not one to leave things alone, I recalled the Beachtek coming solution, and one that Dan Chung talked about using a 1K tone fed into the camera. I tried that, and Dan was right, but the problem was when you were monitoring through a mixer, the tone masked the voice. So I realized I needed an inaudible signal.

Last night I decided to try my hand at it. I went into my copy of Audacity ( a free sound program on the net ) and found that it included a tone generator. Since Beachtek is apparently using a high frequency inaudible sound, I started making tones, until I found a point where it was inaudible. I settled on a 15,400 setting. And then I made it 15 minutes long. Next I rendered it to an mp3 file. I then loaded the file to my IRiver mp3 player, and hooked up the earphones out to the auxillary imput on my old Beachtek DXA-4., and started the player playing the tone. I ran my Sennheiser mic into the other channel, an set the imput at about two clicks below top.

I recorded voice... Result, no noise during the silent periods. AGC was shut down.

I subsequently ran the same thing through my Eng 44, plugging the mp3 player into the auxillary port where the inaudible track ran in both channels, and then recorded to one of the other imputs using the Sennheiser. A winner again.

I also found by adjusting the volume on the mp3 output, you will affect the level of gains setting in the camera. So the volume control in effect can act to raise and lower your levels.

Give it a try, you will be pleasantly surprised.

You already said "no noise"...but I want to grill you for more information regarding the quality. Beachtek keeps saying their new DXA-5D will "greatly reduce the hiss" from the camera, which makes me nervous because Im planning on buying one as soon as it comes out, but I really need something that eliminates the hiss, not "greatly reduces it"...so my question for you: "is there ANY audible hiss? Did the tone eliminate the hiss completely?"

Thanks for doing this by the way, REALLY appreciate the effort and ingenuity on your part.

Nandan

Hi guys:

Well I just did a little testing, enough to satisfy me that with a little bit of work, excellent sound will be produced. Here is a wma file with those first two tests.

The first part is throught the XlR adapter only, without mixer. At first you will hear the camera straining to raise audio gain, to pick up the TV in the background, then I plug the mp3 player in, and the AGC shuts down.

In the second part, I am actually running through the Eng44 mixer, where it seems to work better, as I can control more things. I know a little work will get exact right settings.

Take a listen.

Chris... good test and great results... thanks for the post..

so would it be possible to create a very small, wafer battery powered, in-line tone generator? for one channel of course.

Hold the phone.... maybe the new ipod shuffle would do the trick! Small, cheap, sexy.... kind of like Madonna but younger!

...but I really need something that eliminates the hiss, not "greatly reduces it"...so my question for you: "is there ANY audible hiss? Did the tone eliminate the hiss completely?"

The key to eliminating hiss is to use a great preamp/mixer. That's where the high gain is. After that, feeding a relatively high-gain signal to the camera (as well as the pilot tone to overcome AGC) should minimize hiss.

I haven't tested it yet, so I can't speak to how low the hiss can be.

The preamp/mixer will also give you a headphone output, which is important for controlling levels. (It's not as good as monitoring from the recording device, but better than nothing.)

Even if hiss were completely eliminated (it won't be, but should be low), you will often have other unwanted sounds, such as heating, ventilation, and air conditioning (HVAC). So regardless, you will want to have noise reduction in your toolkit.

On the PC, I've found that GoldWave's NR works well. There's a free demo with something like 50 uses for free, and last I checked, the cost was only $50. You just copy a portion of the clip with noise-only, and apply it to the rest of your clip.

If the hiss is low and you want it to be zeroed out, NR will usually do the trick.

so would it be possible to create a very small, wafer battery powered, in-line tone generator? for one channel of course.

Hold the phone.... maybe the new ipod shuffle would do the trick! Small, cheap, sexy.... kind of like Madonna but younger!


Yes, good idea.... of course the Iphone would work just as well...

Yes, good idea.... of course the Iphone would work just as well...

but much bigger.

For what it's worth.... I've been shooting a huge testimonial project and using 2 5d2's with rode videomics on them.... so far it's worked out very well....

Last night I tried a y adapter at the mic input to see what would happen trying to add the signal input along side the a mic input. I had a Sony stereo mic in one of the stereo inputs, and the Iriver in the other. It did not work. Its possible my connector was not proper for the camera, but it sounded like one signal was interfering with the other, as you would expect. I have no doubt that if I actually split the channels, mic into one channel, and IRiver into other, that the result would be different. I will stop by radio shack today to see if I can locate that adapter.

I am coming to the conclusion that for editing reasons, it is better to keep the tone track separate. In post, in Vegas, I can eliminate one track, and double the other to both sides, and have a wave form that actually shows the peaks and valleys of the voice track. When the two are combined via mixer and/or XLR adapter, you have a solid block along most of the track, with occasional peak over from the voice track.

Here is a wma file with those first two tests.

Hi Chris,

Just to see what the 5D MK II audio would be like once processed, I took your audio file into Sony Sound Forge 9 and did the following:

1- Split the audio tracks and deleted the tone track.

2- Manually adjusted the volume of the voice track.

3- EQ'd the voice track to eliminate some of the high end ( -2dB at 9Khz )

4- EQ'd the voice track to boost some of the lower range ( +4.5dB at 200Hz )

5- EQ'd the voice track to eminate any low end rumble. ( - Infinite slope starting at 80 Hz )

6- Used the Sony noise reduction plug-in to bring down the background noise by -6dB.

7- Used the Sony WaveHammer plug-in to compress the dynamic range by a factor of 2:1 at -16dB.

8- Exported the voice track as a WMA file.

...This is the quick and dirty method of processing audio that I use for work. To me, the processed audio track from the 5D MKII sounds not bad at all.

- Guy

Guy:

Thanks for the work on that. I will review the process you employed tonight on some other material I have.

Guy, had an interesting thing occur tonight, and with my minimal sound experience, I can't figure it out.

So I have the mono setting engaged on my XLR adapter, so every is going to both tracks.

Not a big deal, since I can't hear the tone anyway. So I convert my files to Cineform, open Vegas, and drop them on the line. Play on the line, sound is still clear of any audible tone. I edit, and render to the Cineform 4.xxx codec, and play the resulting file, there is a continuous high tone being heard.

So then I render the sound to a wave file, and it is free of tone, great, I drop it on the track, turn off the original sound track, and render again to same Cineform file. More tone in the final render.

Of course, I will render to separate track to eleviate any issue of the tone in the voice track, but it makes an interesting problem.

Got any ideas ?

Found a resolution to issue in last post. I had the settings at 48000 hz in audio in the HD template I was using. When I dropped it to 44,000 herz, no more tone.. My sound guy thought it could be a harmonics issue... So maybe push the tone track up a little more to 17,000 area..

Hi Chris,

Personally, I would stick with sending the tone track to one channel, and then removing the tone track when editing. Unless you can be asolutely sure of removing the tone frequency completely from the finished audio ( say with a narrow band filter ), I would be concerned about it effecting things somewhere down the road.

Another option that you might consider is to add a high quality digital audio recorder as your main audio source, and use the 5DMKII audio simply as a sync track. I use a Sony PCM-D50 recorder for much of the video I shoot. It's an astounding recorder, even when just using the built-in mics, and it will run for 10+ hours on one set of 4 AA batteries.

Sony PCM-D50 ( about $450 from B&H Photo )
Sony | PCM-D50 - Professional Portable Stereo Digital | PCM-D50 (http://www.bhphotovideo.com/c/product/524130-REG/Sony_PCM_D50_PCM_D50_Professional_Portable.html)

Hi Chris,

Personally, I would stick with sending the tone track to one channel, and then removing the tone track when editing. Unless you can be asolutely sure of removing the tone frequency completely from the finished audio ( say with a narrow band filter ), I would be concerned about it effecting things somewhere down the road.



Yes, that is my conclusion too.. we have an ENG 44 mixer that will mix multiple mics.

As far as separate sound we will go that route, but I would like to feed that camera at same time, and this method will work for that, especially recording to one track, with tone segregated out. Thanks for the input.

And I have now gotten it to the point that in a single shooter no crew hand held situation, I mount a Sennheiser mic on my camera rig, and feed it through my Beachtek, and can get some decent sound.

As far as separate sound we will go that route, but I would like to feed that camera at same time, and this method will work for that, especially recording to one track, with tone segregated out. Thanks for the input.

And I have now gotten it to the point that in a single shooter no crew hand held situation, I mount a Sennheiser mic on my camera rig, and feed it through my Beachtek, and can get some decent sound.

I have not tried this, but I would imagine that you could use a Zoom H4N to record stereo ambient with the built in mics + boom & Lapel on its XLR inputs.

Then possibly attaching the headphone out to the 5D input.

If you want to use the pilot tone in addition to an external recording you might be able to use the H4N in multitrack mode to play the pilot tone while recording.

~Jay

I'd be interested if the Zoom can downmix an XLR input and a built in mic on one channel, while leaving the other channel as just XLR. Then we could play the high frequency tone into that XLR leaving one good channel for boom or radio mic and still getting some nice ambient. Is this possible?

Also you would need to pad the zoom's line or headphone output as the 5dmkII only takes mic level.

Dan

Ipod touch and iphone have a free app "Dog Whistle" and this has a variable tone output, maybe this could be incorporated as the tone generator

I have not tried this, but I would imagine that you could use a Zoom H4N to record stereo ambient with the built in mics + boom & Lapel on its XLR inputs.

...A word of caution. I haven't worked with the new H4N yet, but I owned the Zoom H4 for about a year. It was a pretty good recorder, but the XLR pre-amps were garbage. I ended up buying a Sound Devices MixPre to feed a clean signal to the Zoom H4.

I would hope that Zoom has learned from their past mistakes, though I've read one review of the H4N that stated that the mic pre-amps are still pretty noisy.

This is the reason why I got rid of the Zoom H4 and switched to the Sony PCM-D50, which has far better audio capabilities. ( i.e. Much better built-in mics, much much quieter recordings, and will last 10+ hours on one set of batteries, where the Zoom H4 could barely last 2 hours )

I decided to do a quick test to see just how noisy the H4N's pres are.

In my quick test they are very noisy. In a real world scenario I don't know how bad the signal to noise ratio would be.

My test was quick and simple:
I used a Beta 57 mic because that's what I had handy, and also because it isn't especially sensitive. I held the mic a foot away from my mouth and spoke in normal to soft voice. I had the rec level set to 80 out 100. The hiss was high enough to be distracting.

For comparison I plugged the same mic into a Mackie mixer which has VLZpro preamps, and set the gain to where I thought it should be, and sent the direct out into 1/4" input on the Zoom. With the record level at 11 out of 100 there was no preamp noise that I could hear, and my voice was significantly louder.

I can post the two clips somewhere if you would like to hear them.

I may try the same test using a Sony Lav that I have to see if preamp noise is an issue with that mic.

I guess at $350 for a 4 track recorder, you get what you pay for.

~Jay

Hmmm... Interesting review with sound samples of the Zoom H4n digital recorder

Brad Linder's blog: Zoom H4n audio take 2 - second time not so charming (http://www.bradlinder.net/2009/03/zoom-h4n-audio-take-2-second-time-not.html)

I was recently asked to test the Juiced Link CX231 on my camera, and give some feed back. That was probably based on my posting here, but I also posted in another thread about my first experience, and this repeats some of that.

First, let me say that the CX231 does not promise to resolve the AGC levels issue on the 5D. What I can see happening is Hudson or others cracking the AGC control, giving us gain adjustment as needed on the camera. In the meantime, if I want to get best sound into camera-- as many will want to do, the CX231 is a valuable tool.

Shooting with my Sennheiser ME-66, directly to the Juiced link seemed quieter in terms of background noise than my Beachtek DX4-A. This makes perfect sense because the Juiced Link has powered preamps on board. In fact, I believe the sound is similar to running my mic through my ENG44 and into the 5D.

I ran my high frequency silent test tone trick (using an mp3 file generated tone from Audacity at 15400 hz) into one side of the CX231 through my IRiver 895 player, and of course the ME66 in the other side, as discussed in my original post on this thread. With the gain frozen, the mic input has is very clean and back ground noise is very low. Again, at first blush, it seem as clean as when I did a similar set up with the ENG44. This set up should work fine for shooting one man single mic operation, until we are given gain control on the 5D MKII. I wondered if mic levels couldn't get high enough, with gain disabled on the 5D, but I was actually able to distort with too high of input on the mic side in one set up, so I don't think that is an issue, and a range of adjustments can be chosen from.

As I had tried with the Beachtec, I did try setting both the IRiver input and the Mic input on center with the Pan switches, but that created a warbling effect and ticking noise that made that unusable, I did not have that happen on my Beachtek, so I suspect it has to do with preamps and issues with mixing two signals.

Monitoring is still an issue with this set up. What I want to be able to do at minimum, is split the output from the CX231, and sending signal unchanged to the Camera, while using the other side of the split to act as monitor. At the suggestiong of Robert Rozak from Juiced Link, I am going to try this product:

http://www.boostaroo.com/

I hope to be able to split of of the output from CX231 while avoiding interference with the camera input.

Hopefully, the HUDSON project, which has already succeeded in adding a meter, and disabling the AG, will also provide monitoring through the AV port, in the same way that Canon does with the HV20, but we will have to make do until then.

I will report back with more testing of the CX231 and this camera combo.

I started out the night with the intention of doing a thorough on screen demonstration of the the tone generator/Juiced Link CX231 i had messed with last night. I pluged my little LCD4Video monitor so I could center myself in the picture. I started getting that whine I had experience in shooting the DVChallenge film earlier last month...

After some testing, it is clear that there is signal generating from the output cord that interferes with mic imput, and that is what I had experience the entire shoot, when we were hooked up to a monitor most of the time. And there doesn't seem to be a resolution.

Has anyone experienced similar issue? How about with HDMI out. Is it possible that complaint s about sound in this camera are related to that issue ?

Hi Chris, I just received my juicedLink today and tried it with the iPod. Even without an SD monitor, I heard some rogue noises. On the other hand, the juicedLink into the MicroTrackII looks to be as clean as - if not cleaner than - my Mackie VLZ-Pro mixer into an Audiophile 192 capture card. I need to calibrate things and do a proper A/B, but the initial results of the CX-231 into the MicroTrackII are very encouraging. Frankly, I got way too much crosstalk and harmonics from the iPod trick to be useful recording with the 5D at this point. Maybe when the gain can be properly controlled, the results will be better.

For now, I recommend double system all the way - as well as a clean preamp ahead of any low-cost recorders.

Jon, I am not sure what your set up is, so its hard to comment. What tone level are you using, and how are you feeding through the system. What are you setting the gain switches to on the Juiced Link, for instance.

I will note that when I run the tone into the juiced link at too high a level, I get a repeating, noise on the tone track. I suspect it is the AGC being hit with such a high volume trying to adjust. I am going to try to demonstrate the issues and solutions in a video I will work on this weekend or next.

While I understand your preference for double system, there are many situations where there is just a shooter, and that is what I am concerned with. From my rudimentary testing I think the combo of the Juiced Link and the tone generator can get us a quality input. I see it as a matter of calibration to get it to acceptable levels.

Hi Chris,

I'm running the Shuttle headphone output straight into one of the 5D channels (not having a 1/8" to XLR combination of cables/adapters at hand)

I'd rather be able to record into the 5D directly as well for some situations, but my test results were not promising. We really need Tramm's firmware to turn down the 5D gain. I really had to smother the 5D input to get the gain to reduce acceptably.

Functionally, it's the same number of boxes right now. When I remove the MicroTrackII, I end up adding the iPod Shuttle to the system.

Also, I have no visual meters or headphone monitor with the 5D. The sound quality (at least with my rudimentary setup) was poor compared to with the MicroTrackII. Sure, we would avoid syncing in post, but the quality hit and lack of monitoring are too high a price right now. Even if I improve my pilot signal results, the monitoring problem is still an issue.

With Tramm's firmware, we have a shot at recording into the camera using the preamp, but the jury is still out regarding sound quality and being able to monitor with headphones and the LCD at the same time.

So, for now, I recommend a double system with the 5D for good quality audio and monitoring. My fingers are crossed that we will have a good in-camera solution available from Magic Lantern firmware.

Okay I did some more testing tonight with the JuicedLink CX231, and I think I arrived at a combination of settings that works pretty well, at least for my purposes. Here is how I had the Juiced Link set up:

Mp3 Player into channel 1, set an line level. Gain set at High, trim at 11 oclock position. Volume setting on Mp3 was best at around 30, but good at 25 too.

Sennheisrer Me66 with K6 power source into channel 2, Gain at Medium, trim set to 1 oclock position.

Pan is set Right on Channel 1, Left on Channel 2.

I ran this test with a roomful of cranky and snippy girls. They were drafted and didn't like it. I wanted a variety of sound. Grandpa was a bit snippy as a result.

In Vegas, I selected left channel only, and then rendered an .wma for upload.

It is attached for your review.

For my purposes, I think I can control this set up pretty well, and repeat these results regularly, now that I have run enough test.

The Juiced Link should be even better once Hudson gets the level control and meters on board !

Check out the file attached

I ran this test with a roomful of cranky and snippy girls. They were drafted and didn't like it. I wanted a variety of sound. Grandpa was a bit snippy as a result.Ahh. Family dynamics with snippy kids. I've been there!

Looking at the levels, your noise floor is at about -40dB or so, which is fairly high. You had good, hot levels with headroom though. And I didn't hear anything funny - just white noise.

It would be interesting to do a quick recording with the same level settings in a silent (yeah, right) environment. Maybe go into a closet or a car, put the mic on a stand (no handling noise), and point it into a box lined with foam or a towel or something. That would give us an idea of the noise floor.

Anyway, -40dB noise floor isn't all that great. I got closer to -70dB with the juicedLink, MicroTrackII and Rode NT1-A, and that was with a computer running in the background. My effective gain was lower though. I had my voice at about -18dB when a couple of feet from the mic. That said, The NT1-A isn't all that sensitive. I was running High gain at about twelve noon when I got -70dB. Your preamp settings were much lower (medium), so that tells me that the noise floor was set by the camera.

BTW, my camera tests had a similar noise floor to your file.

Hopefully, the Magic Lantern will be able to reduce the camera gain (and noise) further.

The 5D2 has a 16 bit A/D. That gives a dynamic range of 96dB (6dB per bit.) At 40dB, the camera is only delivering 7-bit audio(!) The -70dB noise floor is closer to 12-bits of dynamic range - and I recorded at 24 bits! It looks like 16-bit recording is all I need for this combination of mic, pre and recorder. No reason to use any more memory space than needed...

The 5D2 has a 16 bit A/D. That gives a dynamic range of 96dB (6dB per bit.)

Yes, theoretically 16 bit will give you 96dB, but in real life this will most likely be in the low 80s.

-- peer

The 5D2 has a 16 bit A/D. That gives a dynamic range of 96dB (6dB per bit.) At 40dB, the camera is only delivering 7-bit audio(!) The -70dB noise floor is closer to 12-bits of dynamic range - and I recorded at 24 bits! It looks like 16-bit recording is all I need for this combination of mic, pre and recorder. No reason to use any more memory space than needed...

So Jon, I had this happen with my Eng Mixer, where when I ran audible tone, and used the panning switches not disimilar to the ones on the CX231, and I had bleed from the tone channel into the mic channel. I am wondering if there is possiblity of some bleed here raising the floor level with bleeding over of the tone. I am not so sure these pan switches are designed to completely isolate channels. Or should I be complaining to Sign Manufacturing ?

Chris,

FYI in my testing of the 1khz XLR tone generator I've found that I get tone bleed depending on which audio adapter I'm using. With a Beachtek DXA2s it is pretty bad, but with a DXA6 it is non-existant, both fed from my Sound Devices mixer. Its interesting that you are getting issues with the Juicedlink too. I pinning all my hopes on the DXA-5D now.

Dan

Dan:

I am not disatisfied with the sound. I was commenting on Jon's numbers, and wondering if the noise level of the floor could be attributed to that. I don't have evidence of it one way or the other. And my question was whether these pan switches allow some bleed.

Yes, theoretically 16 bit will give you 96dB, but in real life this will most likely be in the low 80s. Agreed. Not only is there analog noise, but clock jitter, and errors in the quantization steps. Getting 70dB with the NT1-A, juicedLink and MicrotrackII right out of the box (and with a PC running in the background) is pretty darned clean.

I am not disatisfied with the sound. I was commenting on Jon's numbers, and wondering if the noise level of the floor could be attributed to that. I don't have evidence of it one way or the other. And my question was whether these pan switches allow some bleed.Chris, I agree that the sound quality is fine, but the noise floor is high. What frequency signal did you use? I could try some filtering. Unfortunately, the track has very little silence, so it's hard to get a good noiseprint for noise reduction.

One problem with the whole approach is that the connection to the camera is unbalanced stereo in a single cable. No matter what we do, there will be some bleed. In fact, I got lots of bleed even when sending the pilot tone directly into the camera with a short dual-mono to stereo combiner plug. The problem is that we need to crank the signal so high in order to defeat the AGC. I should try a low tone like 200Hz. It will have less crosstalk and might be more effective at reducing gain at lower levels.

Before long the Magic Lantern solution will make this moot. We'll be able to get stereo signals into the camera, will be able to manage the gain of the preamps, and won't have to worry about artifacts from the pilot tone.

I see Juiced Link is approaching this for DSLR's, with a product release on the new year.

juicedLink (http://www.juicedlink.com/blog/index.html)