Shooting next to a fountain

B

berniebeaudry

Well-known member
Hi All,
Just did a very interesting shoot near a large noisy park fountain. A couple approach the fountain toss in coins, make a wish, then turn toward each other and talk about what each other wished for. A third person comes in and gives some advice and then leaves. I put a hidden lav on the two women in the scene and boomed all with a Schoeps CMIT5U.
Boom on channel one, lavs mixed to channel two.
I boomed from underneath for the mediums to closeups and from above for the extreme wide shot. I did a test record on a digital recorder with some stand ins, and also rolled on some takes at the beginning of the day. Played those back for the director. He was also listening on cans later as we were shooting.
I was planning on getting the actors away from the fountain to do the lines wild but we all spaced it out and it didn't happen. I was listening to some of my tracks today from the recorder and I'm wishing we had done the wild lines. While you can hear the dialogue pretty well I'm not loving the amount of water sound that is coming through. I played with noise reduction on some from free audio software. Got a noise print from the fountain and it worked to take out all of it except a warble or two of the fountain. Didn't seem to affect the speech too much at all. Of course one would have to add the water sound back in to make this work at all, it sounds pretty weird. Tried some eq, tried muting the lav track, using a mix of the two tracks etc. Really just playing with minimal tools to determine how improvements could be made. This is a quick turnaround project that will be played at huge meeting so wanted some ideas in case the director has any issues after the fact. Yes I know ADR would be an option too, but am hoping it isn't necessary.
Thanks,
Bernie
 
I'm assuming that you can't go back and shoot any of this again?

I'd have left the fountain on for the wideshots and had no dialogue, and shut it off for any medium shots that involved dialgue and for all the close-ups that involved dialogue. Just shoot to where you can't see the falling water and then lay down a water effects track in post to cover all your cuts.

As far as removing the unwanted sound that you already have, you might download a free 10 day trial of Ozone from Izotope and in the multiband dynamics portion use a noise gate to try to cut the sound of the water out. Probably wouldn't work unless the sound of the fountain was really low in comparison to the level of your dialogue.

If you download it, they have an awesome PDF tutorial.
 
If using noise reduction truly did not negatively affect the dialogue, I think when you replace the fountain sound in post, it should hide whats left of the fountain sound from the original track.

Try using an expander (a gate that doesn't close all the way) to reduce it even more in between dialogue. Using a gate might call attention to the sound coming and going.

I don't know what program you're using, but Synchro Arts has a fully functional demo of VocAlign for download. It’s a great plug-in for ADR alignment. Something to keep in mind.

 
bmowell said:
I'm assuming that you can't go back and shoot any of this again?

I'd have left the fountain on for the wideshots and had no dialogue, and shut it off for any medium shots that involved dialgue and for all the close-ups that involved dialogue. Just shoot to where you can't see the falling water and then lay down a water effects track in post to cover all your cuts.

As far as removing the unwanted sound that you already have, you might download a free 10 day trial of Ozone from Izotope and in the multiband dynamics portion use a noise gate to try to cut the sound of the water out. Probably wouldn't work unless the sound of the fountain was really low in comparison to the level of your dialogue.

If you download it, they have an awesome PDF tutorial.
Click to expand...

Would have been nice to be able to shut it off, but its a very large decorative fountain in the middle of a park. It was a good twelve feet high. We would have done that if possible. No we can't reshoot, its a quick turnaround pseudo commercial for a business meeting. I got in good and tight for the close ups and that helps. I'm just concerned that when its played out on large speakers the ratio of bg will be too high. I'm still kicking myself for not getting the lines wild away from the fountain. I'm not involved with the post I'm just trying to get answers just in case I get the dreaded call from the director about the bgs being too high. He was listening too, but you never know. The tools I was playing with were on Audacity and Wavepad. They are pretty limited compared to PT or any other for real audio program. I'll check out Ozone and see what it can do on this.
Thanks,
Bernie
 
Hxd Ped said:
If using noise reduction truly did not negatively affect the dialogue, I think when you replace the fountain sound in post, it should hide whats left of the fountain sound from the original track.

Try using an expander (a gate that doesn't close all the way) to reduce it even more in between dialogue. Using a gate might call attention to the sound coming and going.

I don't know what program you're using, but Synchro Arts has a fully functional demo of VocAlign for download. It’s a great plug-in for ADR alignment. Something to keep in mind.

Click to expand...

I thought it was working but when I went back to it I couldn't duplicate the noise reduction I got. It is affecting the voice unfortunately. I was just playing with Audacity and Wavepad, two fairly limited free programs. Just trying to be proactive incase the director calls and isn't happy.
Thanks,
Bernie
 
As mentioned, you can use an expander in the FX bin and set a level that you think is comfortable during non-speaking lines. If you do have wild takes of the fountain, good. The fountain BG that you're trying to get rid of, however, will sound unnatural because there's a certain amount of water present during the speaking lines. How bad is it?

If you're planning to use lines from a different take and cut around to make it sound less noticeable try sidechaining the wild fountain track to the expander so whenever the expander is engaged, the wild track will presently fill in the gaps.
 
Hxd Ped said:
If using noise reduction truly did not negatively affect the dialogue, I think when you replace the fountain sound in post, it should hide whats left of the fountain sound from the original track.

Try using an expander (a gate that doesn't close all the way) to reduce it even more in between dialogue. Using a gate might call attention to the sound coming and going.

I don't know what program you're using, but Synchro Arts has a fully functional demo of VocAlign for download. It’s a great plug-in for ADR alignment. Something to keep in mind.
Click to expand...

Just to be pedantic, an expander is not a gate that does not close. An expander is basically like a compressor that is working in reverse. You select a ratio (like 1:2) and for every dB the input signal falls, the output will fall 2 dB.

Also, most gates have some sort of floor or low threshold that allow the gate to remain open regardless of input level to prevent chattering.
 
Alan H. Chang said:
As mentioned, you can use an expander in the FX bin and set a level that you think is comfortable during non-speaking lines. If you do have wild takes of the fountain, good. The fountain BG that you're trying to get rid of, however, will sound unnatural because there's a certain amount of water present during the speaking lines. How bad is it?

If you're planning to use lines from a different take and cut around to make it sound less noticeable try sidechaining the wild fountain track to the expander so whenever the expander is engaged, the wild track will presently fill in the gaps.
Click to expand...

You can hear the lines pretty well. The water isn't overpowering the dialogue but there is more bg sound than I personally would like to hear. The concern I have is that when this gets played on a pa system at the meeting its for, the bg sound might be annoying. I'm not involved in post on this other than following up on my tracks. I just wanted something to tell them if it even comes up. I downloaded a free version of RX just to see what that could accomplish and it was encouraging. The boom track worked better than the lav track. There is more high end energy on the lav track. I'm just assuming (hoping) they will cut the closeups audio for the main track and sync it up with the wide stuff as necessary. I tried RX without using any eq first so I know that will help some too. Anything that you can think of that the lav track could help with in this scenario? Its on the two women and its brighter sounding than the boom. The male voice is also being picked up by the lavs as well. I'll play with an expander if I have one here. Just got some decent monitor speakers just tonight so will give a better listen on those when I get them put in. Thanks again for any and all ideas.
Regards,
Bernie
 
Hi Bernie,

Most often your lavaliers will have less background water than the boom because of it's close proximity with the speaking heads. Yes, you can still use the lavalier tracks rather than the boom. How well does it sound? I'm not sure. When you mean bright, do you mean that it has a lot of high-frequency, harsh, or very crisp and clear? RX is really a handy tool, it can save you from situations like this. If you think the boom track works better than the lav, then by all means go with the boom.

The lavalier tracks could be used in conjunction with the boom track. Mix and level them till they blend well together. The lavalier track is going to sound more up front and closer and your boom track is going to have more background. Depending on the outcome after processing through RX, should the boom track become better sounding than go with that. If not, use your lavaliers as your primary and add the boom track in for background noise matchups as necessary. I say go with this route before using any additional FX. Less is more.
 
tigeba said:
Just to be pedantic, an expander is not a gate that does not close. An expander is basically like a compressor that is working in reverse. You select a ratio (like 1:2) and for every dB the input signal falls, the output will fall 2 dB.

Also, most gates have some sort of floor or low threshold that allow the gate to remain open regardless of input level to prevent chattering.
Click to expand...


That pretty much describes a gate that doesn't close to me. My description was probably over-simplistic.

And just so I can be un- (anti-?) pedantic, gates aren't just all on or all off. They have dynamic controlls before they close fully. (Many of them do, anyway.)

But like I said, my description was probably over-simplistic.:) No worries.
 
If you want to feel better about yourself, you can watch Primer. There is a fountain scene that sounds like Niagara Falls.

If you post a few short .wavs, maybe someone nice (like me, for example), could run it through some processes to try their hand at it.
 
ProfessorU said:
If you want to feel better about yourself, you can watch Primer. There is a fountain scene that sounds like Niagara Falls.

If you post a few short .wavs, maybe someone nice (like me, for example), could run it through some processes to try their hand at it.
Click to expand...

Hi Professor,
I just tried to upload a wav but its too big for the member upload center. I don't think it supports mp3 either. Suggestions for doing this? The upload center says it supports WMA files but I don't see a way to convert a WAV to that format. I shortened up the file and it was still 5.5 megs. I don't have any way to host as far as I know.
Bernie
 
Professor must be busy on a project or on vacation. Does anyone else want to play with my tracks and see what they can do? So far haven't had any negative feedback from the director so all may be well. Just curious what you would do with these tracks. If you need more clips I can load more up.
Thanks,
Bernie
 
http://panasonic.sixbullets.net/public/fountaintest1b.zip
The base noise level isn't terrible. I think if you mixed with that level, it would be acceptable to most directors, assuming you didn't add any extra noise.

I don't have a real monitor system with me, so I mixed the clip by eye while monitoring on earbuds, forgive any audio anomalies. I ran the stereo channels through independent noise reduction. Afterward, it felt like both the voices were a little to midrangey. I brought down 1k on the woman's voice about 3dB, and 500 on the man's voice about 3dB.
 
Here's a trick to try:

Duplicate your dialog track, and do extreme noise-reduction and/or EQ on one track to minimize the fountain sound, even if there are some artifacts/unnaturalness.

Then try blending the "effected" track with the original in various ratios. If this still sounds unnatural, try some different (less extreme) things on the effected track.

But, especially since your fountain level is not terrible, an approach like this might be able to enhance the voices without any unnatural sound... the original audio mixed in will cover a lot of artifacts from the effected track...
 
Thanks for checking this out. Its good to know that the backgrounds aren't too objectionable and that they're workable. I've been doing location sound for a long time and every once in awhile these moments of doubt creep in to wake me up. I can't listen to your new clip till I get home but I'm sure you've cleaned it up nicely. Haven't heard anything negative from the producer so I assume all is well. Or should I say wishing well?
Best Regards,
Bernie
 
Thanks Ulanta!
I'm gooing to play with your suggestions. Good to hear the backgrounds aren't terrible in your opinion. No news is good news and I haven't heard anything from the producer negative or otherwise so it seems all is fine. You wouldn't believe how loud that fouintain was. I think the CMIT did a decent job of it given the circumstances. A Sanken CS3E might have been a better choice but there aren't aroiund here for rental that I know of. Nest time I'll ge the lines wild too so I don't have to add more gray hairs to my head.
Best ,
Bernie
 
You must log in or register to reply here.
Back
Top