I own three microphones and two sound recording devices.
With this equipment I’ve captured Lego building noises for my YouTube videos in the last 5 years. I rarely use music in my videos, instead I try to capture authentic soundscapes with little background noise or reverberations. Besides Lego sounds, I’ve recorded outdoors ambient nature sounds, underwater sounds, and for example sounds on-board a Lego centrifuge.
Audio-Technica AT2020
The first microphone I bought 15 years ago was Audio-Technica AT2020. It cost around 90 EUR.
This is a condenser microphone (as are all my mics). The audio quality is quite good in my opinion. It has a little bit more background noise compared to the Rode NT1, but generally very low amount of noise. The polar pattern is cardioid, so it will pick sound from front and sides. Despite the old age, it still works just fine, so it seems to be very durable. I used this mic in my videos for a long time until a year ago Rode NT1 replaced it.
Sennheiser MKE 600
The second mic I bought 5 years ago was Sennheiser MKE 600. This is a shotgun mic and it cost 300 EUR.
The long and narrow form is a big benefit, as this mic fits on a camera. I often attach it on my Sony AX43 when I shoot outdoors or other places outside my normal work table. The polar pattern is narrow, so it will block noises outside the direction of the subject, which can help in limiting room reverberations. There is a low-cut switch button on top of the mic to filter out low frequencies. I have used that button many times, for example in windy conditions to prevent distortion. The audio quality is good – noise level is about the same as AT2020.
Rode NT1
A year ago I bought Rode NT1. It cost 270 EUR. The purpose was to update my old AT2020 with a similar mic that has less noise.
This is very much like the AT2020. Both are condenser microphones, physical size is similar, and the polar pattern is cardioid. The biggest difference is with noise level. I have measured NT1 to have about 10 dB less background noise compared to AT2020 and MKE600. That is a noticeable improvement. No need to add DeNoise filter in Premiere for this mic. The sound is also brighter compared to AT2020 and especially MKE600, so in post processing I usually need to filter out high frequencies a little bit more than normally.
Zoom H5
My first audio recording device was Zoom H5. It cost 315 EUR, so a fairly expensive device (perhaps too expensive for the purpose). After 5 years, I still use it today.
This device supports 48V Phantom Power, which all my microphones require. It has two XLR input sockets, so you can record stereo audio using two external mics. Volume for each microphone can be adjusted conveniently with the knobs on the front. The device also has a stereo microphone on top, but I have never really used it, only done a quick test to notice it has more noise than my other mics. The display shows volume level indicators that are handy in checking that everything is working before you start recording. The settings are easy to use, no problems there. Files are stored in an SD-card that is entered to the side. You can run the device with batteries, but I keep it always on DC power. In terms of audio quality, I’m pretty sure it is good enough for any mic, meaning the quality depends only on the mic you use, not the recording device.
Zoom H1n
The second audio recording device I bought was Zoom H1n. This little device cost only 80 EUR. The aim is to use it as a small travel recorder when shooting outdoors or other places where I don’t want the big setup of H5, two mics, and cables.
The device is conveniently small and you can attach it on top of my Sony AX43 camera. It records stereo audio with the integrated X/Y microphone on top. Audio quality is a little bit disappointing, as it produces more noise than any of my other mics, and you need to filter the noise in post-processing. But, it is a cheap device, so you can’t expect high quality. The record button isn’t the best, as I have at least two times noticed later that it failed to start recording. Now I always look at the red indicator after pressing the button to make sure it is recording.
Windscreens
I have three foam covers and one dead cat.
The foams protect from little gusts of air, but they are not enough in high wind situations. When I shot my 88 kg rock lift video outdoors, it happened to be a very windy day. I put foam covers on my AT2020 and MKE600, and even knit winter caps on top of the foams, but that was not enough. The audio was partly blown out, as you can hear from the video.
The dead cat windscreen is Zoom WSU-1 and it fits on my Zoom H1n. It does a little bit better job than the foams, but not much. Because protecting against wind is so difficult, I usually try to avoid filming in such weather.
Shock mounts
I have four shock mounts to protect against vibrations and handling impacts.
The big one on the left is Rode SM6 and I use it to mount Rode NT1. The one on lower right is Audio-Technica AT8458, and it fits on AT2020. The one on top is Movo SMM1 and I use it often to mount MKE 600 shotgun mic on my Sony AX43 camera. The little one on right is Sennheiser MZS 600, which also can be used to mount the MKE 600, but it does a little bit worse job than the Movo.
Stands
I have one big microphone stand: K&M 27105. It cost 37 EUR. Although this is a mic stand, I have mostly used it as a lighting stand, as it reaches to places my other straight lighting stands won’t. I have even put a camera on it a few times to film Lego builds straight from above.
I have a bunch of mini stands for microphones, LED lights and cameras. I have 5 gorillapods/monkeypods that are always good versatile stands. Manfrotto PIXI is very sturdy little tripod with a terrific ball socket on top. Benro BK15 is another one I use often, as it can be adjusted heightwise between 50…100 cm, although it is quite flimsy and falls easily. I have 3 pieces of K&M 232BK stands that are heavy and sturdy. Black Eye Filming Handle Tripod is not very good IMO, as you need to turn the rod to adjust height, and your camera will spin also.
Cables
I have four XLR cables to connect the mics to the recording devices. Two 3 meter cables and two 1 meter cables. Below is an image of the shorter ones.
Noise tests
Just out of curiosity, I did a quick measurement on the noise level between my microphones. The measurements were done in my quiet storage room. The volume level was normalized between mics by making noises of similar intensity and looking from the Zoom H5 intensity bars that the volume level is the same. The results are only indicative. I recorded silence with the mics and analyzed the result with Cool Edit Pro 2.1 Frequency Analysis tool. Frequency is on the x-axis between 20-20000 Hz and sound intensity is on the y-axis between -50…-120 dB.
On the graph you see Audio-Technica AT2020 and Sennheiser MKE 600 have virtually the same amount of noise. Zoom H1n is the worst, having about 10 dB more noise. Rode NT1 is the best, having about 10 dB less noise than AT2020 in middle and high frequencies.
The environment
One big aspect of recording audio is the environment where you operate. I record my videos mostly in a small storage room that is very quiet. The room has clothes piled on the shelves and jackets hanging from a rail, which dampen the echoes effectively. There is very little reverberations, only a little bit of low booming reverb because of the plasterboard shelves and walls. It is a good environment to record quiet Lego building noises. For example this 2.5 hour Lego building video is recorded in that space.
I have used this same storage room from the beginning. In fact, I selected this environment precisely for it’s acoustics. Otherwise it isn’t very good, as there is too little space for setting up lighting and camera angles, and I have to walk to a different room to edit videos. But in terms of audio quality, it is good.
Other rooms in my apartment have a lot of echos. I’ve shot a few videos (for example Googol gear train and 20 Mechanical Principles) entirely in my living room. They all have this spacious echoey acoustics, which isn’t particularly good. In those videos I used my shotgun mic close (30 cm) to the subject to limit the amount of echoes, but it improved the result only a little.
When I shoot outdoors, the nature soundscape creates a totally different vibe (example video). Lego building sounds disappear under the nature sounds, but that is fine, since I don’t make long building sessions there. Underwater soundscape is another completely different world (example video_1, example video 2). As is onboard camera sounds on a Lego car (example video) or a spinning centrifuge (example video). In those situations the audio becomes a big part of the video and creates an authentic, visceral experience.
Microphone setups
Here are a few examples how I’ve used my microphones. This first image is from my early videos. In green circles you see AT2020 on the left side and MKE600 on the right side. Both mics are about 40 cm distance to the center of the building table. Because they are positioned to the extreme opposite sides, forming a stereo sound from them results in a weird 3D sound. Kind of annoying if you listen with headphones. You can listen an example in this wheel spinning video.
This image is from my current setup. In green circles you see MKE600 on the left side and Rode NT1 on the right side. Distances from the table are 90 cm and 60 cm. Most often I use the Rode NT1 audio to create a mono sound. An example of this is from the Submarine 4.0 video building scenes. But sometimes I have felt the MKE600 is better with softer sounds, so in my Vacuum Pump video I used only that. You should hear some background noise as I didn’t use DeNoise filter.
This image is from my living room shooting the longest gear train video. MKE600 shotgun mic is located 30 cm away from the build. That is the only mic used in the video. You can hear some reverb for every click and snap from the spacious living room.
Example from Submarine 4.0 test scene. Audio is recorded using Zoom H1n with a foam windscreen attached. It is placed as close to the water container as it can be without being in the shot.
Recording settings
I use MP3 320kbps as the file format for Zoom H5. In the beginning I always selected the maximum quality, which is WAV 92kHz/24bit. But a year ago I questioned, what am I doing here? No-one can hear the difference. With MP3 320kbps I get 14x smaller file sizes, or about 150 MB for every hour or stereo audio, and the quality is virtually the same. Since then I have used mp3.
For the volume level I usually clap my hands and see how high the intensity bars go, and try to adjust volume knobs until the peak is at -12 dB. That leaves 12 dB of overhead, which is in theory 4 times louder, but is perceived by the ear a little bit less than that. For the two microphones, I usually set higher volume for the close-up mic and lower volume for the farther away mic. That way, if there are loud scenes that will distort the sound for the close-up mic, I can use the other one. Finding proper volume level requires a little bit trial and error. Too high volume leads to distortion and too low volume leads to noise. Both can be somewhat fixed in post-processing, as I have used Audacity’s Clip Fix and Adobe Premiere’s DeNoise with some success.
Post-processing
After I’ve done filming a scene, I import the video files and audio files to my editor program (Adobe Premiere Pro 2020). The first thing I do is synchronize the files. That is because I have two files: a video file (that includes audio from the camera mic) and an audio file (quality audio from a separate recording device). Usually I have pressed record button at the same time, so the files start about the same time, but not precisely at the same time. I often clap my hands together after I’ve pressed record to help synchronizing them, especially if it is a quiet scene. So, in the editor I will look at the first peak sound (hand clap) that is visible in both files audio waveform graph, and I will move the clips along the timeline until they are synchronized. The clips move at 1 frame increments, which means the result may be 0.5 frames (20 ms) out of sync, but I’ve noticed that is good enough and you cannot perceive the error.
The next thing is to organize audio tracks. I usually use 3 to 8 audio tracks for different purposes. Below is an example from my latest video. Audio track 1 is reserved for video audio that is muted. Track 2 is for scenes where I showcase different mechanical principles made from Lego. Track 3 is for scenes where I do Lego building. Track 4 is for the final machine testing. Track 5 is speed up footage for the final machine. Track 6 is for background music.
After organizing, I’ll add effects. This is done in Adobe Premiere’s Audio Track Mixer.
The first effect I always add to the end of the Master track is a Hard Limiter. I’ll use otherwise default settings but set the maximum amplitude to -1 dB. The purpose is to cut peaks, such as Lego clicks and snaps that can be noisy and make the audio distort. Usually the audio doesn’t reach that level, so the limiter rarely does anything, but it is a habit to put it there. Sometimes when I notice an annoyingly loud click that I want to quiet down, I’ll add a Hard Limiter effect directly on that particular audio clip and set for example -12 dB setting to it, which resolves the problem nicely.
The next thing is panning. Usually I use Fill Right with Left or Fill Left with Right effects that will just take one of the stereo tracks and use that to create a mono sound. Basically this selects which microphone I want to use, since I often record with two mics. In my old videos I didn’t use panning at all, which resulted in an ugly stereo effect, especially when listened with headphones, because the mics were positioned to the extreme left and right side of the subject. Later I started to use a Stereo Expander effect with 50% narrow setting to combine the stereo tracks into a slightly centered mono sound. But now I’ve gone to pure mono sound. If I record outside with Zoom H1n, then I use stereo sound, but otherwise it is mono.
If the audio has too much background noise, I might add a DeNoise effect. I put the Amount setting to a very small value like 2% or 5%, which is enough to cut about half of the noise. I don’t want to overdo it, since it will affect real sounds. I rarely use the DeNoise effect, because my mics are good, but especially with the H1n recorder when used in quiet indoor situations, you can hear the hissing noise.
I always add a Parametric Equalizer filter. I usually want to cut down high frequencies (above 2000 Hz) because the microphone is positioned very close (40 cm) to the subject and that makes the Lego clicking sounds too bright and in-your-face. Also, if I record in my storage room, I usually cut low frequencies (below 150 Hz) because it has wooden shelves and a wooden table that cause booming sounds here and there. Sometimes rarely I put a narrow notch filter on certain frequency to cut a noise that comes from my adjustable power supply for example. I have also used DeHummer for that same purpose.
Another effect I always add is a Dynamic Processing effect. That is used to compress the audio to even out quiet and loud parts. In my old videos I used heavy compressing, like 3:1 ratio, but later I’ve come down to 1.5:1. I think it sounds more natural that way. In Settings tab I always increase Look-Ahead Time to 10 ms and bring down Gain Processor Attack Time to 5 ms. That is because otherwise the compressor won’t be fast enough to react to sudden clicks, so they will sound too loud.
The last thing is to adjust volume levels. I usually aim for -12 dB in continuous loud parts. In the beginning I used much louder overall volume level, but I’ve come down as I think my videos should have a low-key atmosphere. Adjusting the volume needs to be done for audio tracks and sometimes for individual audio clips. Similar scenes should have a similar volume level. This is a phase that always requires some work. I may listen the complete video once with headphones and once with a cell phone or tablet to hear how it sounds. Besides the volume, I may adjust the effects. When listening with headphones, all details like the background noise may come a problem, so I may need to add e.g. a DeNoise filter. With a cell phone, the sound becomes brighter, and low frequencies are lost, so I may want to adjust the Parametric Equalizer.
That is all the processing I do in Adobe Premiere. I also have an old audio editor called Cool Edit Pro 2.1, which I still use sometimes for frequency analysis and other thing. And some times I’ve used Audacity for it’s Clip Fix effect, for fixing overblown audio e.g. from the SQ12 camera footage.
Brick Experiment Channel 
Pingback: Thoughts on Video Editing | Brick Experiment Channel
Your blog is insanely useful and detailed. Searched for audio / lighting / video making advice, and this is about the most clear, concise, specific advice I’ve seen.
Well done.
LikeLike