• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

microphone 2 basic audio recording questions

Ok let me start off by saying that audio is not my field of expertise, AT ALL. As you will notice by the questions I'm about to ask.

I'm trying to record a simple short voice-over for a video I'm editing. There will be a soundtrack playing underneath the voice-over.


Set-up:
Rode NTG3 connected to a Tascam DR-70D with an XLR cable going into the 1/L channel.
The microphone is phantom powered and the Tascam is recording in "stereo" mode, WAV 24bit, sample rate 48K, with gain set to "mid".


Question 1:
I keep reading online that sound should be recorded between -20dB and -12dB, but this results in volume that is way, way too low for my liking when imported into Premiere. And significantly lower than any video I've ever seen online as well.

Is this recording supposed to be amplified in post? Wouldn't that generate worse quality than just recording at a higher dB?
If it's supposed to be amplified afterwards, what's the best way to do this?


Question 2:
When I import my recorded audio I notice there's only a waveform on the left channel. Obviously this is because the microphone is only connected to the left channel, but how do I get sound on both channels?
Should I change something about the way I record, or is there a way to copy/paste it to the right channel?


If there are any other tips or stuff that I might be doing wrong, I'd love to hear it. As I said.. total noob.

Cheers
 
Ok let me start off by saying that audio is not my field of expertise, AT ALL.

Nothing unusual there.

I'm trying to record a simple short voice-over for a video I'm editing. There will be a soundtrack playing underneath the voice-over.

Set-up:
Rode NTG3 connected to a Tascam DR-70D with an XLR cable going into the 1/L channel.
The microphone is phantom powered and the Tascam is recording in "stereo" mode, WAV 24bit, sample rate 48K, with gain set to "mid".

Okay. Why are you recording the VO to the DR-70 and not directly into your computer?

1. You should be recording in MONO.
2. 24/48 is standard for film/video/game work.
3. "mid" may not be enough gain, depending upon how the VO is performed. Many VO artists speak very softly and "eat" the mic to give their voice more presence and to take advantage of the "proximity" effect to give the voice more depth/resonance.

Question 1:
I keep reading online that sound should be recorded between -20dB and -12dB, but this results in volume that is way, way too low for my liking when imported into Premiere. And significantly lower than any video I've ever seen online as well.

The -20dB/-12dB is standard for on-set production sound. The concept is that the PSM (Production Sound Mixer) has plenty of headroom as a safety in the case of loud transients. Example - The characters are speaking softly, suddenly start screaming. The PSM would be reacting to the screams by turning the input level down, but the initial scream is probably very disorted or highly flattened by the limiter (production sound folks always use a limiter, but hope that it never becomes engaged) or both - distorted AND flattened. Since you are working in a controlled environment you have the opportunity to "experiment" (looking for the appropriate gain/level settings) prior to recording - a luxury not afforded on-set.

It is not as important for PSMs to ride the faders as it was before digital audio recorders. Back when tape was used (can anyone say "NAGRA"?) the biggest enemy was tape hiss, so you always tried for the "hottest" signal you could get onto analog audio tape. Since digital recorders don't have tape hiss it is much easier to get a great signal-to-noise (S/N) ratio.

Is this recording supposed to be amplified in post? Wouldn't that generate worse quality than just recording at a higher dB? If it's supposed to be amplified afterwards, what's the best way to do this?

You can increase the signal level of the audio clips once you have them in your NLE/DAW. In Avid Pro Tools I usually use the "Gain" or "Normalize" audio suite plug-in; there should be something similar in Premier. After opening the Gain/Normalize plug-in you highlight the clip(s), hit the "analyze" button and it should show you the the dB in either RMS or Peak (I use Peak). Then the plug-in should allow you to increase the gain by an amount that you specify. Depending upon the sound I usually increase the gain to -5.0 so I have a little extra headroom when I start pumping the clip(s) through EQ and other processors. I'm sure that there is a YouTube clip out there to explain how to do all of this in Premier, or one of our resident editors can point you in the right direction.

Question 2:
When I import my recorded audio I notice there's only a waveform on the left channel. Obviously this is because the microphone is only connected to the left channel, but how do I get sound on both channels? Should I change something about the way I record, or is there a way to copy/paste it to the right channel?

As I state previously, you should be recording in MONO!!! The DR-70 allows you to do this. If the VO has already been recorded you should be able to split the stereo track into two mono tracks. Erase the blank track and pan the active track to the center. Then gain-stage the clips (Gain/Normalize).

Expect some frustration and confusion; you are entering a whole new and very different world that requires a fair amount of technical skill/knowledge in addition to your discipline and creativity.

I'll be here (locked down in my house) if you have any further questions.
 
Last edited:
Thanks so much! Your post gave me a lot of insight and some of the terminology helped me dive further into understanding more about how to properly go about this. While probably far from perfect, I've got a result right now that I'm very happy with!


Okay. Why are you recording the VO to the DR-70 and not directly into your computer?

This was because I don't have a way to directly connect the microphone into my laptop, and I wanted to take this opportunity to familiarize myself with these new pieces of equipment. Also, my laptop makes a noise significant enough for the mic to pick up, and I wanted to get a recording as clean as possible.
 
Depending upon how much money you want to spend there are numerous ways to connect your NTG-3 to your laptop. You can use an impedance matcher (between $15 and $50) and plug directly into the mic input of your laptop. You can get a passable USB audio interface starting at around $175, and there are plenty of used interfaces in decent condition out there. And, of course, you can spend lots more if you have it.

There are ways to keep the noise of the laptop contained as well. Even something as simple as the positions of the mic and the laptop. Since the NTG-3 is a very directional mic making sure that the laptop is behind wherever the mic is pointing can reduce quite a bit of noise. There are very simple sound containment things you can make or buy that can substantially reduce your noise issues.

I'm down here in the weeds with the low/no/mini/micro budget filmmakers, and can help you find inexpensive solutions to your audio problems. I would suggest that you read "The Location Sound Bible" by Ric Viers. The first third/half of the book does a good job of explaining audio basics amd most especially sound-for-picture. It definitely helps if you speak the language and understand the basics; at the very least you will be able to intelligently discuss audio issues with sound folks.

Peace,

Uncle Bob
 
I dont want to hijack the thread but whenever i try to record directly to my computer windows royally screws it up. there is a slight fade out/fade everytime the speaking starts. its like it stops recording when the input signal drops below a certain level and then there is a lag time starting it back up.

in my settings i have it so none of the ports lose power if they are inactive, but after about an hour trying to fix that i gave up.
found it was easier to just record the audio into my tascam and then port it from the sd card.. more work flow but it works every time
 
What are you using as an interface? What software?

there is a slight fade out/fade every time the speaking starts. its like it stops recording when the input signal drops below a certain level and then there is a lag time starting it back up.

It appears to be an issue with settings somewhere; maybe a "noise gate" was used in the signal path at some point. A noise gates are (almost) literally a gate that opens and closes; it can be set to remain closed until a loud enough signal opens the gate. There are settings on many noise gates that tell the gate to open or close or close slowly or rapidly. This was used to supress tape hiss back in the old analog days. You would set the gate to remain closed until a signal louder than tape hiss comes along and tells the gate to open. The gate was usually set to close slowly so there was not an abrupt change in the sound - the tape hiss faded out rather than stopping suddenly.

There are also compressor/limiters which work in a similar way but have a different effect upon the audio signal; limiters put a ceiling on (or limits) how loud the signal is before it is passed onto the next link in the audio chain. As with a gate the affect of the limiter can be set to react slowly or quickly. Limiters are quite often a part of an audio compressor. A compressor, in addition to limiting the top loudness of the signal, can also increase the loudness of quieter signals and, as with the gate and the limiter, there are settings that control how much the signal is compressed and how slowly or rapidly that will happen. Compressor/limiters are quite often used on voices, both spoken and singing, to smooth out volume inconsistencies in the performance. (Of course, these days compressors are overused to increase the "apparent" loudness of sounds to the point where there is very little dynamic range in modern music - but thats my gripe and different thread.)

The point of this technology lesson is that, from your description, it sounds like compressor/limiter/gate settings are affecting the signal you are attempting to record. Since you say that the signal fades out AFTER you start recording it sounds like a compressor/limiter/gate with a very slow opening response time and a rapid closing response time - the gate opens slowly and closes very quickly.

Just a guess on my part as to what you issues are. If you give me a very detailed description of your hardware, software and signal chain we can probably walk our way through the issue.
 
that would be great!

I have a sound devices mm-1 preamp and I am plugging the headphone monitor directly into my computer's microphone port on the front of the tower.

My computer specs

Windows 10 Home 10.0.18362 Build 18362

Code:
Tower:
https://www.amazon.com/gp/product/B00GMG5KD8

Motherboard:
ASUS ROG Maximus IX Hero LGA1151 DDR4 DP HDMI M.2 USB 3.1 ATX Motherboard

One Graphics Card:
EVGA GeForce GTX 1080 Ti SC2 GAMING, 11GB GDDR5X, iCX Technology - 9 Thermal Sensors & RGB LED G/P/M, Asynch Fan, Optimized Airflow Design Graphics Card 11G-P4-6593-KR

Processor:
Intel 7th Gen Intel Core Desktop Processor i7-7700K (BX80677I77700K)

Ram:
4x Crucial 16GB Single DDR4 2400 MT/s (PC4-19200) DR x8 Unbuffered DIMM 288-Pin Memory - CT16G4DFD824A

Hard Drive:
Samsung 850 EVO 1TB 2.5-Inch SATA III Internal SSD (MZ-75E1T0B/AM)

Power supply:
Corsair Power Supplies , RMX 850W CP-9020093-NA

And here is some audio for you to cringe at lol, just recorded it to demonstrate the issue

Recorded with "voice recorder" windows app

 
Last edited:
It sounds like Voice Recorder takes a moment or so to engage, or, the limiter function on the MM-1 is behaving strangely (unlikely, but a possibility). Or the issue is with the mic input on your computer (has it's own limiter/gate?). Do you have a cheap mic around to check that out?

Remember that computers are actually stupid; they only do what they are told to do.

Lord, I HATE troubleshooting! It always turns out to be:

A) Something really weird in the way the softwares are interacting.
B) Something incompatible between the pieces of hardware.
C) You're (I've been) doing something really stupid that you (i) haven't pick up on yet (wrong settings, incorrect connections, etc.).
D) Any combination of A, B & C.

I would suggest that you try using the XLR output of the MM-1 and use an impedance matcher into the mic input of your computer. I have no idea if it will work or not, but it's a place to start. Or perhaps a better voice recording app. Or even a cheap/free DAW like Pro Tools First, SoundBridge or Audacity. Here's a list: https://thehomerecordings.com/free-daw/

As always I suggest getting the proper tool for the job - a real audio interface. The PreSonus AudioBox is only $100. Of course, you can spend LOTS and LOTS more.
 
I have a mic on my oculus rift hmd that i can use to record something with audacity.

I was thinking more in terms of using the Oculus in place of your MM-1 connection just to test the MIC input of the computer with everything else as before. Does the same thing happen? What I'm getting at is the MIC input jack of your computer is meant for consumer devices - i.e. consumer input levels. The signal from the MM-1 may be too hot, so a limiter of some kind kicks in and takes a while to find the proper attenuation level. So one solution - if you do not want to go with a mini-pin adapter from the XLR of the MM-1 - is to lower the headphone output level. Maybe start low (2 or 3) and work you way up until the limiter kicks in. Oh, are you sure it's not the limiter on the MM-1?.

I may hate troubleshooting, but the result is usually worth the hassle.
 
I was thinking more in terms of using the Oculus in place of your MM-1 connection just to test the MIC input of the computer with everything else as before. Does the same thing happen? What I'm getting at is the MIC input jack of your computer is meant for consumer devices - i.e. consumer input levels. The signal from the MM-1 may be too hot, so a limiter of some kind kicks in and takes a while to find the proper attenuation level. So one solution - if you do not want to go with a mini-pin adapter from the XLR of the MM-1 - is to lower the headphone output level. Maybe start low (2 or 3) and work you way up until the limiter kicks in. Oh, are you sure it's not the limiter on the MM-1?.

I may hate troubleshooting, but the result is usually worth the hassle.
Hey I really appreciate you and unfortunately the mic for the oculus is part of the usb/hdmi connectors. I am at the final day of nothing but editing for a month straight so I didn’t get to that test yet but definitely on it
 
But definitely track the problem down. It's a part of the audio learning process, and an important one for low/no/mini/micro budget types.

Learning how to troubleshoot an audio signal chain is a crucial skill.

But when I read this, I cringe:
I have a sound devices mm-1 preamp and I am plugging the headphone monitor directly into my computer's microphone port on the front of the tower.

There are two things that strike me here. First is that the 3.5mm mic input on a PC tower is already a festering cesspool of awfulness. Those inputs are typically not well isolated from RF noise from the rest of the tower and are not very clean as far as signal quality. Anticipate anything from hum and buzz and hiss to weird companding artifacts. Never expect this input to be worth your while.

Second is using the headphone output from the MM-1. Headphone outputs are weird since they’re not really line-level and they’re not really mic-level. They live in their own world. Sure, you can use headphone outs for little Bluetooth speakers and even feed the unbalanced line inputs of a mixing console, but the unpredictability of headphone output in general, paired with the expected trash of a PC’s mic input, just spells a recipe for disaster.

Unfortunately, neither the DR-60DmkII nor the DR-70D are able to function as USB interfaces like many of the smaller recorders are. But if you’re looking to record voice to your computer, I’d highly recommend picking up a small Focusrite USB interface. Either the Solo or the 2i2 will be enough, depending on if you want/need to be able to record a single mic or two mics at a time. A single mic is usually enough for most VO work. ADR really needs two (shotgun and lav). It’s a minimal investment either way. Use something like Audacity or Audition to record, or use the free software that comes bundled with the interface.

Bonus for the Focusrite: it gives you proper monitor outputs for good monitor speakers.
 
Last edited:
I was thinking more in terms of using the Oculus in place of your MM-1 connection just to test the MIC input of the computer with everything else as before. Does the same thing happen? What I'm getting at is the MIC input jack of your computer is meant for consumer devices - i.e. consumer input levels. The signal from the MM-1 may be too hot, so a limiter of some kind kicks in and takes a while to find the proper attenuation level. So one solution - if you do not want to go with a mini-pin adapter from the XLR of the MM-1 - is to lower the headphone output level. Maybe start low (2 or 3) and work you way up until the limiter kicks in. Oh, are you sure it's not the limiter on the MM-1?.

I may hate troubleshooting, but the result is usually worth the hassle.

Both limiters are off on the MM-1 and i dont get this behavior when I record to the tascam

I plugged my phone into the audio input jack and played a youtube video into voice recorder - It didn't have the behavior but it was a lot quieter too so might not have triggered it. I recorded with the USB audio of the oculus and also didnt get the behavior. I am okay with buying a relatively cheap component if it means i have a proper hookup i suppose.

Okay so i hooked my mm-1 back and tried to do it with a super low level, super high level, etc - i cannot trigger the behavior at all anymore.
Everything is working perfectly now and I didn't change anything lol. great it'll probably break just as randomly again next week or something 😄
 
Glad things are working now. And you're right, it'll probably peep out it's ugly head again, and probably when most inconvenient.

I am okay with buying a relatively cheap component

Don't buy cheap, buy INEXPENSIVE. There is a difference.......

There are a number of decent audio interface units out there for under $250, even passable for under $150. I'm partial to Focusrite, ART and PreSonus, although there are other solid brands with which I have no personal experience.
 
Back
Top