Eliminating mic bleed

Barry W. Hull wrote on 4/16/2013, 7:59 PM

Anyone have an opinion on the Audio-Technica BP4025 as an overhead mic to record either two or three people sitting around a fairly small table?

farss wrote on 4/16/2013, 9:07 PM

"Anyone have an opinion on the Audio-Technica BP4025 "

Never used one myself however that mic has been around for a long time and I've heard plenty of people speak well of it.
I have a similar mic, the Rode NT4 which is a little cheaper. Used it for many things from chamber orchestras to medium orchestras to two people speaking.

For people sitting around a table do not ignore using a PZM / boundary microphone. This type of microphone is one of the few microphones that really does seem to have a certain "magic" to it. There's some quite cheap omni boundary microphones from Shure, Samson and even Yum Cha ones from China on eBay.

Bob.

musicvid10 wrote on 4/16/2013, 9:23 PM

If you're going to get a PZM, and they are good in the round, get the Crown 6D. You can't go wrong. But despite their appearance, there is no rear rejection and no isolation. PZM usually has a dome pickup pattern and fairly modest front gain. Their one advantage is good phase correction of reflections.

For a 1/4-spherical (bandshell) pickup pattern, good rear rejection and front isolation, nothing beats my Crown PCC's. Much better for headon interviews and lecture / stage work IMO.

But boundary mics in Barry's situation tend to pick up table level noise, so someone tapping a pencil, drumming their fingers, or shuffling papers is going to be heard. That's why my first choice is still overheads, similar to what Bob or John Cline first suggested. I've done this with a pair of AT Pro45's with pretty good results, although the AT is a bit hot in the upper freqs.

I think it's important to note that in any given setting, each possible mic configuration comes with its own special set of advantages and disadvantages. In addition to environmental controls, a thorough understanding of pickup characteristics and

richard-amirault wrote on 4/16/2013, 10:13 PM

Just found this thread ... I was also going to suggest a boundry / PZM mic. I've used them successfully with a 4, 5 or 6 person panel discussion on a single side of a banquet table.

This is a demonstration of it's effectiveness with regard to on-camera mics, but you can see the quality of the sound with regard to each person on the panel when recorded with a PZM. With only two people a single PZM would work fine. I assume you don't want/need stereo.

NOTE: yes, there are disadvantages (as musicvid points out) especially with "table noise" .. but, you can "train" your talent not to do that.

John_Cline wrote on 4/16/2013, 10:26 PM

Boundary microphones are definitely worth considering. I've used them in various scenarios and they have worked exceptionally well, particularly when the speakers are seated at a table.

musicvid10 wrote on 4/16/2013, 10:37 PM

Nothing personal, brighterside, but that demonstration makes my point regarding pzm vs. pcc better that I ever could have through a description.

-- The audience and room noise and coughing are equally as loud as any of the participants, often overpowering them. You don't want that. 360 deg. coverage over the X-Z axes is the biggest disadvantage of PZM in a headon recording setting.
90% of that undesirable rear "program" would simply disappear if they had used a pair of PCC 160's instead.

-- The producer(s) of that video evidently forgot that those mics have low-cut filter switches on them. As a result of that, the rumble is excessive.

Barry W. Hull wrote on 4/17/2013, 7:46 AM

Lots of good advice, thank you.

We’re going to rent, and experiment, then decide. I never would have guessed that adding one stinkin’ mic to the mix would take a week’s worth of my brainpower trying to solve this dilemma.

One thing I've decided, I can tolerate 'decent' video, but I can't tolerate 'decent' audio. I want GREAT audio.

rs170a wrote on 4/17/2013, 8:39 AM

One thing I've decided, I can tolerate 'decent' video, but I can't tolerate 'decent' audio. I want GREAT audio.

If you want GREAT audio, hire a pro who does it for a living as he/she will have the gear and, more importantly, the skills to give you the GREAT audio that you want.

I say that somewhat tongue in cheek as I'm a one man band on most of my shoots and can't afford to hire a pro. It's never in the budget so I do the best that I can with my limited resources.

Mike

musicvid10 wrote on 4/17/2013, 10:53 AM

Location audio is tough, no matter how you slice it. A dedicated sound grip (not an overstressed camera operator) is indispensable.

richard-amirault wrote on 4/18/2013, 8:59 PM

Nothing personal, brighterside, but that demonstration makes my point regarding pzm vs. pcc better that I ever could have through a description.

No offense taken, musicvid. I bow to your expertise. You are a professional and I am an amateur in these things.

The video is mine. The purpose of it was to demonstrate the advantage of close mics RELATIVE TO on-camera mics. My ears are not as trained as yours (or maybe most folks) but the PZMs sound a lot better, to me, than the on-camera mic.

Now is there a better way to mic this type of panel? Yes, I'm sure there is. However, for these shoots .. I am running alone, often from room to room with only minutes to pack up, move to the next room, and set-up before hitting record. This is the best I can do in this situation.

musicvid10 wrote on 4/18/2013, 9:42 PM

Yes, having omnis between the panel and the audience is better than on-camera miking behind the audience. That point is well taken.
Directional is even better.

I have seen union sound designers in regional theaters fooled by the appearance of the PZMs. They look unidirectional; in use they are anything but.

Next time you do a panel discussion like this, beg, borrow, or steal a pair of PCCs and mount them the same way. You will be astounded at the front gain factor; you may even want to mic the audience to get a little more of them in the mix. Hint: Make sure the PCCs are facing the right way!

farss wrote on 4/19/2013, 2:59 AM

Another hint, a boundary microphone picks up the boundary layer travelling over a hard surface. Tablecloths etc probably don't help.

Bob.

Rory Cooper wrote on 4/19/2013, 3:57 AM

Bob what tools in Vegas you are using to get that “dead” no hiss on you audio?
Try and explain you mental process, where do you start?

At 1:14 the lady gives a ahah this is a throat sound and is the same level as the speech. how do you get it so that the levels are the same?

farss wrote on 4/19/2013, 5:34 AM

Bob what tools in Vegas you are using to get that “dead” no hiss on you audio?

Whilst some high end roll off can be used in Vegas to eliminate hiss I generally don't use that.
Instead I use reasonably good condensor microphones. They have a high output and therefore only need reasonable quality microphone preamps e.g. the ones in prosummer comeras Once you use dynamic micrphones you can need best possible microphone preamps e.g. kit from Sound Devices.

"At 1:14 the lady gives a ahah this is a throat sound and is the same level as the speech. how do you get it so that the levels are the same?"

That one is simple. Use a volume envelope to duck the offending cough etc.
The trick is to match the volume envelope to the offending sound. If it's got a sharp attack then I pull the envelope down very quickly and right in line with the start of the sound. If it has a gradual decay I match the trailing edge of the "duck". The aim is to make the offending sound less offending without leaving a noticeable hole in the sound of the room.

I also oftenly add a compressor to the buss. I set it up so the threshold is just making the compressor work, I set the ratio between 1.5 and 3, generally 2.0 is good. Now when I duck an offending sound on the T/L the signal going into the compressor is below the threshold so it doesn't get made to sound louder but the speech is.

Bob.

Rory Cooper wrote on 4/19/2013, 7:12 AM

thanks Bob, I am using a preamp but only when I know we have a power supply nearby, would be nice if they made a battery operated pre-amp with a VU meter, one that doesn’t cause a thumping sound in your head before you even use it = price can make your heart race. Keep your heart in good condition look through an audio visual catalog twice a day.

Thanks for the tip on the bus track. Will play around with that.

Barry W. Hull wrote on 5/4/2013, 7:13 PM

So, after much experimenting with various equipment, trying to walk that fine line between the variables of perfect audio and “good enough” and easy/simple setup/editing, we came to the following conclusions on a two/three person conversation, all sitting around a small table…

We’re wiring each person with the Audio-Technica AT898 Cardioid Lavalier microphone. (I almost went for a sweet Sennheiser shotgun, but chickened out, too expensive)

We had the talent move her head while talking, yeah, we could tell, buy hey, just don’t do that, we made some “outside” commotion to see how it bled through, had the other talent talk loud to see much mic bleed came through, and AT898 did a good job, really cut out off-axis sound.

We recorded the audio on the Tascam DR-680. It is easy, does a great job, very suitable for our needs.

We synched up three cameras and the Tascam with Plural Eyes, holy smokes is that thing FAST and accurate, a complete no brainer.

Instantly we could see how much better the cardioids did in this situation compared to the omni microphones, no comparison in quality of sound. When we muted the offending track, we could barely tell a difference, there was a teeny tiny bit of bleed, but not enough to give us that lousy “hollow blurry” sound, barely perceptible really.

We then tried the Vegas multi-camera editing tool, that worked like a champ too, did not take much practice to switch back and forth between cameras in a pleasing way, really only after 15 minutes of practice it looked good.

It didn’t matter that our three cameras had slightly different looks, not enough to be annoying and detract from the material. The Sony HDR-XR260 (tiny hand held) is a bit contrasty for my tastes, but the HVR-Z5U and the HXR-NX5U are incredible, really too much detail, but Beauty Box adds a nice touch of makeup.

So thanks for the advice in this thread. However, I still don’t care what anyone says, great audio is lots of trouble and a pain, and please don’t argue with me about that. I am retired from the US Navy, but I still have friends, and if anyone tells me great audio is easy, I will blow up your house.

Thank you, carry on.

http://www.audio-technica.com/cms/wired_mics/c56cac8605404472/index.html

http://tascam.com/product/dr-680/

musicvid10 wrote on 5/4/2013, 7:21 PM

Well Barry, after all of that I guess we'd like to see your product?

Barry W. Hull wrote on 5/4/2013, 7:28 PM

Thank you for asking Musicvid, soon will put something on Drop Box.

jerald wrote on 5/7/2013, 11:04 AM

If you have an already completed session & need to fix in post, something to try:

You mentioned that your issue is a slight "echoey", hollow sound. This indicates that one lav is picking up the other speaker with a short delay creating an echo and causing cancellation of certain frequencies.

Possibly you can set up a correction via inversion, delay, and summing, for each mic's media by using other mic's media to cancel the undesired waveform/signal/data component.

----------------------------------------
Setup:

Assuming two mono mics:
Place each mic on a mono track, so two tracks.
(for this explanation, I'll call one track 'Joe' and the other 'Jane.')
Pan each track to separate sides to facilitate tuning (explained below). Let's assume Joe is on left, Jane is on right.

Duplicate track 'Jane.'
Name this new duplicate track 'CancelJane'sEcho.'
Invert it's signal.
Pan it to coincide with track 'Joe's' location.
Lower the amplitude (loudness) of 'CancelJane'sEcho' to match the perceived level of the undesired echo.
Turn off 'Quantize to Frames' in Vegas.
Put headphones on.
Set up an appropriate loop to listen for tuning (explained below).
Zoom in on the timeline until you can move the CancelJane'sEcho track media in very small sub-frame amounts.

-----------------------------------------

Tuning:
While listening to the loop, slide the 'CancelJane'sEcho' media in tiny increments to the right first & left as needed until you hear that the undesired (Jane's voice) component in Joe's sound decreases to it's minimum.

Once you've been able to reduce the Jane'sEcho component to a minimum, lower and raise the level of the 'CancelJane'sEcho' track to improve tuning.

Repeat these two steps to find optimal improvement.

------------------------------------------------

Let me know if you have questions.
Hope this helps.

Barry W. Hull wrote on 5/8/2013, 6:04 AM

Jerald, thank you for the detailed advice and explanation. I will experiment with it in order to improve my audio capabilities.

However, on our current project the directional lav mics are reducing enough off-axis sound that the mic bleed is only slightly audible. When someone is speaking and we turn off their track, we can hear them coming through the other microphone, but it is muffled enough so that it doesn't lower the quality of the audio to an unacceptable level.

The amount of mic bleed we are experiencing would likely be unacceptable for high fidelity recordings, maybe theater or music, but for our needs it works well.

With the cardioid lav mics, our setup is now quick and easy and does not require any extra steps. We use Vasst Ultimate S Pro to add a slightly more pleasing sound to the voice, but other than that, nothing else is required.

farss wrote on 5/8/2013, 7:41 AM

You could probably address your bleed problem in post using gates, the default Noise Gate in the track header is a bit of a bust in my opinion. I've had much better luck with the Graphic Dynamics plugin that ships with SF. No doubt there's plenty of other gates available.

I think I've said this before but it bears repeating.
In general there's a 3:1 rule regarding mic proximity i.e. if two mics are 3 time further apart than they are from the sound sources bleed isn't going to be a problem.
Your problem is this rule doesn't work because any lapel mic is way off axis from the sound source and it's off axis, up close and from a quite directional sound source. Everything is working against a lapel mic working well. In fact as you've discovered it might do a pretty good job of picking up the other person. The few times I've shot wedding the one lapel mic on the groom does just as good a job of picking up the bride when they're facing one another, at least the mic is more on axis with the bride!

What all that says in summary is lapel mics are the mic of last resort. They should only be used when you have no other option.

The other piece of wisdom is using anything "wireless". Certainly the most expensive wireless gear is close to being as transparent as a balanced audio lead but the cheaper kit that uses FM isn't. Again this should be considered the choice of last resort.

Here is a short video from Rode University. OK, it obviously pushes Rode and ignores the high end shotguns such as the Sanken CS-3e however the same principles apply, you could use any half decent cardioid condenser mic from AT, AKG etc,

In summary your problems could be solved quite easily with a Rode NT5 or NT6 or even a NT3, the last option is under $300. I'd even say a Rode NT-1A could be used although being a LDC you need to be careful of any airflow in the studio.
The only other gear you need at most is a good boom stand. If you've got the money the offering from Manly is a staple in every studio but there's options at half the price that are pretty much as good unless you've got clients bent on trashing the studio :) I've used the Manfrotto Combi Stand to get up and over several times.

Bob.

Laurence wrote on 5/8/2013, 10:23 AM

Why do you want both mics in the final mix at the same time? I would just temporarily unlink the audio from the video, split the audio at points between the speakers, relink the audio and video, then select the mic for whoever is talking at a particular time.

musicvid10 wrote on 5/8/2013, 12:33 PM

In most interview situations some bleed is expected and shouldn't require much intervention. When it does, just open the mixer and ride gains in realtime playback, like a live engineer would. If you know the script, it's done in one pass.

Barry W. Hull wrote on 5/8/2013, 6:06 PM

Bob, thanks for that informative video, in our original shoot, we had all those problems, mic bleed, echo all over the place. We quieted the room with quilts and blankets and foam rubber, and traded our non-directional omni lavs for directional cardioid lavs. That took care of our issues. I read in another thread the pros and cons of hiding mics. In our case they don’t really bother us, not a big deal.

We tried out a directional shotgun mic and we found that it picked up some of the outside noises (loud neighbors) more than the cardioid lavs. We not only had the initial problem of sound bouncing around the room, but we had issues with noise coming into the room from outside.

Laurence, it’s not that we want, or don’t want, both mics on in the final mix, it’s just easier that way. I probably should have made clear that this is not a one-time project. These videos will be created at least once per week, so we wanted to avoid the need to manually switch tracks back and forth to whoever is talking at any particular time. These are more conversational, with oftentimes two or three people speaking at once.

Musicvid, you are spot on with your assessment of our level of bleed. With the omni mics it was a problem, but with the cardioids the bleed is so slight, it requires no intervention, can barely tell any difference when the off-speaker audio track is shut off. Fortunately, that allows us to avoid the need for manual manipulation after the fact.

After using the omni lavs we were skeptical that we would be able to do this without manual input in post because the omni lavs seemed to pick up EVERYTHING, but the directional cardioid lavs worked well, better than expected, really closing off any off-axis noise. I can now see how putting one omni-directional lav on the groom would pick up the bride and preacher just fine.

So far we’ve only tested in practice sessions, but when we produce a finished product I’ll post it to drop box, let you audiophiles rip it to shreds. I have very thick skin, but as of now, we're pleased.

I'll admit again, that Senni (cool audiophile lingo) shotgun mic we tested, not sure what model it was, was a very sweet solid nice microphone that sounded great. It just seemed to pick up slightly more off-axis than the Audio-Technica lavs.