Freesound Project

The Adventures of a Freesound Member

My work background

I work as a Sound Designer and Re-Recording Mixer for The Discovery Channel where I work on promos and full length documentaries. We have access to most of the major sound effects libraries, but I like to use as many of my original recordings as possible-- partially because I get tired of hearing the same effects on shows and movies that I see, but also because a program often calls for something that doesn't exist in the commercial libraries. I now have a library of recordings that is over 10GB and has thousands of cues. I record a lot of ambiances, because they are the foundation of most of the work I do.

The beginning of an obsession

Richard Humphries

I started collecting sounds in the early 1990's. At first I used a mono Nagra recorder and an AKG 451 that belonged to my employer at the time. I wasn't very organized about it, but if I needed a sound, I would go out and get it. Later, after I had changed jobs, I started using my new employer's DAT machine and their Neumann RSM 191, which is a wonderful microphone-- I wish I still had access to it today.

In 2000 I started to think about putting together my own sound effects library. I did some research of the available technology at the time and came a cross the Zaxcom Deva II. The Deva II is a 4 channel portable field recorder that records on a hard disk, rather than tape. This is more common now, but at the time, the only other option was a DAT, or a Digital Nagra, using reel to reel tape. Today, there are many more options, like Zacom's Deva IV and V, Sound Devices 722 and 744, Fostex FR-2, Aaton Cantar, plus a myriad of MP3/Wave recorders. The Deva II was made for location film sound work and so has timecode in and out, along with 4 mic pre-amps, and 4 line in/outs. It records at 48kHz, 24 bits and it also has a pre-record buffer which will store up to 10 seconds of the input previous to when you hit record. I thought this would be a great feature for when I was out in the filed waiting for something to happen; I wouldn't have to always be rolling, I could wait for the sound and then hit record!

Since the Deva II could record 4 channels I decided that I would try to create a surround effects library. The idea was to record a front pair and a rear pair. I decided to make the front pair a Mid-Side pair and the rear a pair of spaced omnis. The mid-side pair I could easily spread into the three front speakers, and the spaced omni pair would give me a nice diffuse surround field.

For microphones I contemplated both the Schoeps colette series and the Sennheiser MKH series. I settled on a pair of Sennheiser MKH 20 omnis, and an MKH 30 and 40 mid side pair. The Sennheisers are reportedly better in humid conditions, and also reject RF noise more, but it was also a matter of price. I am slowly collecting a set of Schoeps mics too, since I really love their sound.

A recording expedition

My first real expedition with this new rig was a three week trip to West and East Africa. I started in Ouagadougou, Burkina Faso where I was attending the bi-annual film festival Fespaco. Pretty quickly I discovered how cumbersome traveling with that much gear can be. Added to that, my schedule and complete incompetence in French, Mòoré or any of the other 68 languages spoken in Burkina Faso, kept me in the center of the city the whole time, rather than out in the rural areas like I had hoped. None the less, I was able to do some recording of urban and suburban Ouagadougou. I've uploaded one of the recordings to Freesound here.

Aforementioned file was recorded here,
in Ouagadougou, Burkina Faso

Recording with 4 microphones is exponentially more complex than recording with two. It's nearly twice as much gear, and it's a whole lot of cables. It takes a while to put together, and usually by the time everything is set up, a boisterous crowd has gathered to see what it is that is going on. I also had a difficult time with what I thought was RF interference in at least one microphone at any given time. I later discovered that the phantom power of my recorder was probably not sufficient to power all 4 microphones.

The second half of my trip was in Ethiopia. I actually do speak enough Amharic to get around, so this trip was much more fruitful, though not entirely without difficulties. Again, I was plagued by the interference I was getting in one of my microphones. I also ran into some trouble by setting up outside of a government office. I was trying to record a group of minibus touts, but I also ended up recording an argument about whether I could record there and whether or not I was a spy. When I traveled outside of the capital to some more rural areas I didn't take my gear with me because it was cumbersome and because I was unsure of it's security. There were several occasions where I really wish that I had brought it along, particularly when I encountered a pack of hyenas and another time when I came a cross a traditional wedding party, with ululations and all.

In the end, I was able to get some useful recordings out of the trip, but I definitely had learned a couple lessons. The second lesson actually took me a few more expeditions to get through my head, but I'll share it first. It was that a multichannel recording is not very practical, and probably not even that useful. Initially it seemed useful, and marketable, but now that I have more experience mixing in surround, I would rather use several stereo ambiances that fit well together than a surround ambiance.

The first lesson I had learned was that I needed to have a more discrete, portable and potentially disposable recording setup. I ended up buying a minidisc recorder and a Audio Technica 822 stereo mic. I realized that this wouldn't give me the recording quality of the Deva, but a recording of slightly lesser quality is more important to me than not having the sound at all. I've recorded plenty of material using this setup and I take it with me nearly every where I go. If there is something specific that I need to record I will take my Deva setup (though now I only use two channels), but if I am just out and about and hear a sound I want to record, I just pull the minidisc out of my jacket pocket and start recording—usually without anyone even noticing.

Editing and cataloging sound

The biggest task in collecting sounds is editing and organizing them. This job starts while doing the recording itself. Over the years I have found that it is very important to take notes while recording and to put a voice slate on the recording. I can't even count the number of times that I was sure I would remember the specifics of what I was recording, only to have to put off the edit for several months-- then when I finally do get back to the recording, I can't remember what it was or where it was, or what I was intending to capture. Since I record a lot of ambiances, I usually mention the date, time of day, location, and what I am recording. It's not always possible to slate at the beginning of a recording, so now I try to at least give a description at the end of the recording.

Once I have made the recordings, I load them into my computer to edit them into usable pieces. I used to do this on my PC using SoundForge 5.0, but it renders really slowly and I don't like the quality of the processing. Now I usually edit on my Mac which is running Protools TDM 5.3. I don't usually do very much severe manipulation, unless I am trying to design a particular sound. If I had recorded in mid-side stereo, I will decode it to a standard stereo file and adjust the width to what sounds natural to me. I will usually adjust the level so that is sounds natural, or maybe a little louder than natural, with my speakers calibrated to 79dB spl, which is the level at which I mix. I realize that this often leaves a lot of wasted headroom, but it saves me the trouble of having to reduce the gain significantly when I am mixing. I also do some EQing to my sounds, particularly in the low end. The lower frequency components of sounds tend to build up upon one another, making a mix boomy or muddy and eating up its headroom in the process. I typically use a high pass filter to remove some rumble. I usually start at a very low frequency and then move the filter up until I feel like I'm loosing something important in the character of the sound, then I'll back it back down a bit and compare between the original and the filtered version until I think I've done just enough.

I will also edit the sound for content, finding the longest usable portions and sometimes editing several together. I remove mic noises, wind noises, perhaps an airplane or a car passing-- anything that I don't really want in the recording. For ambiances, I like to get 3 minutes of usable sound. Sometimes I will record for 10 minutes to get those 3 usable minutes, sometimes I might only get a minute or 30 seconds of usable sound. Sometime I don't get anything usable at all. Occasionally, I get something completely unexpected. On July fourth, I went out to record people lighting off fireworks. I was recording in one location that was giving me bad results for fireworks, but then a guy rode by on the squeakiest bicycle I had ever heard. I'm not sure what I'll use that sound for yet, but when I need it, I know I'll have it.

Once I've edited and processed the sound, I will usually create a Broadcast Wave file. If I had recorded on the Deva, I will keep this as a 24 bit 48 kHz wave file, If the source was a minidisc, I will create a 16bit 48kHz wave file. The file is named as descriptively as possible given the number of characters available. It then gets placed in a folder in my library of sounds. My library's hierarchy is probably not ideal, but it works for me. I have an outer layer of folder such as Outdoor Ambiances, Indoor Ambiances, Fire, Water, People, Mechanical, Electronic, Transportation, etc. Within each of those folders I have sub folders which divide the sounds into further categories. For example: Transportation has the subfolders Trains, Cars, Motorcycles, Bicycles, Boats and Planes. Planes has a further set of sub folders; Jets and Propellers. I also add all of my sounds to I-Tunes, so that I can quickly search and listen to the files. Unfortunately I Tunes doesn't recognize 24bit files, so I will probably purchase SoundMiner sometime soon, at which time I will begin the arduous task of describing all of my sounds into the metadata fields of the broadcast wave files.

The importance of Freesound

I discovered the Freesound Project through a Yahoo group. It seemed like a very interesting concept and one that I immediately saw the importance of for my line of work. I am constantly in search of recordings of remote places or unusual things-- how fantastic would it be to have people all around the globe sharing sounds from their localities?! While it appears that Freesound is frequented more by musicians than people in my particular line of work, I have seen the number of sounds that are useful to me grow significantly since I first became a member in April, 2005. Despite having made a significant investment of both time and money in my recordings (which I hope to recoup at some point by releasing a commercial effects library) I'm happy to share some of the sounds that I have recorded for others to use. Throughout my career I have benefited from people sharing their knowledge with me, and I've met some wonderful people who have been willing to share their recordings with me too, so I would like to continue that tradition by sharing what I can with those who share an interest in sound.

Comments...