by date
Recording Voice Overs that DON
Recording Voice Overs that DON
| Name: | John Seguin |  |
|---|---|---|
| Date Posted: | Mar 01, 2006 | |
| Rating: |  4.8 out of 5 | |
| Public: | YES | |
| Comments: | YES | |
| RSS Feed: |  or Subscribe with  . | |
| Profile Page: | View profile page for John Seguin |
Blog post
How to produce voice-overs that DON'T SUCK: Part II, The Tools
Greetings! If you have not read the first in this series, please review that section now before proceeding.
Now, the tools! In this article I will lay out for you the minimum tools that I would recommend to record a quality voice over (VO).
Computer
Mac, PC or Linux... choose your poison. Although I profess to be a Mac guy myself, I have no beef with any OS. The bottom line is that you need two high-quality pieces: a good sound interface and a good audio editor. These abound by the dozen. If you are going to be doing this sort of thing for the long haul it would of course behove you to investigate the pros and cons of all of them before making your purchase.
In general, I would recommend looking at a multitracker as opposed to a 2-track editor. The reason being that two-track editors offer a more simplistic environment and often offer advanced sound design and editing features (such as those found in Peak Pro, a favorite sound effect/VO editor of mine. For example, I found the ability to loop a sound with a cross-fade built in extremely easy and very convincing for taking seemingly random sound scape materials (such as crickets in the forest) and slicing it down to a meer 20-30 seconds without it ever appearing to "loop". This is easy enough when you are dealing with sounds that have a very finite beginning and end (foot steps, for example), but becomes extremely complicated when you start working with sounds that are continuous (a train flying down the tracks, or crickets, or a waterfall, to name a few).
Logic Pro 7
However, I have also had lots of success recording straight into Logic Pro, a multi-track editor. The advantage here is that you can then easily mix in additional sounds, music, voice all at once and adjust their balance, pan, EQ, effects, etc. all in real-time.
There are a variety of multi-track editors to choose from. I have compiled an incomplete list of some of the better-known and well-used packages here:
1) Logic Pro (Mac) - www.apple.com/logicpro
2) Pro Tools/Pro Tools LE (Mac and PC) - www.digidesign.com
3) Digital Performer (Mac and PC) - www.presonus.com/
4) Nuendo (Mac and PC) - www.steinberg.de/ProductPage_sb190e.html?Product_ID=2442&Langue_ID=2
5) Cubase (Mac and PC) - www.steinberg.de/ProductPage_sb51ba.html?Product_ID=2442&Langue_ID=2
6) Vegas (PC) - www.sonymediasoftware.com/products/showproduct.asp?pid=965
Sound Interface
Many computers come with built-in sound capabilities. However, the quality of these sound interfaces often varies dramatically. The standard for these types of devices that DO support sound-in is usually stereo-in using a 1/8" jack, digitizing at 16 bit, 44.1khz. This can be considered fairly basic stuff. The 1/8" jack presents an assortment of problems. First, there is no way to provide phantom power throught this jack (which we'll discuss soon when talking about microphones) and the preamp (if there is any) is often very low in quality resulting in poor quality electric guitar/bass direct-in.
In addition, the standard for professional quality sound is 24-bit, not 16-bit. Sampling rates may still start at around 44.1 khz but are often done at 96khz or even 192 khz. These extremely high sampling rates however are probably more than you really need just to record a professional sounding voice over. However, a discussion of sampling rates and their effects on sound quality really could fill an entire article so I'll leave that for another time. (Hint, hint...)
Finding a reasonable quality sound interface that supports 24-bit and phantom powered XLR connection for a condensor microphone should do the trick.
MOTU 828 MKII
The rig that I use is a MOTU 828mkII firewire interface. This is probably overkill for just doing a voice over, but since I have many different sound sources in my studio, the additional channels are a necessity, not a luxury. Here are some other well-known brands in the industry that offer a range of products that should fit the bill in the form of PCI card, USB or Firewire:
1) MOTU - www.motu.com
2) M-Audio - www.m-audio.com
3) Pro Tools - www.digidesign.com
4) Tascam - www.tascam.com
5) Edirol - www.edirol.com
6) Mackie - www.mackie.com
Microphones
Microphones are a very tricky subject. They are very, very difficult to choose because each of them has very much a characteristic "sound". Sure, they can be generally lumped into a handful of categories, but in the end, each has a unique sound. Unfortunately, from a catalog it is often very hard to judge what that "sound" will be. My usual recommendation is to go by word-of-mouth from folks you know, follow message boards the mic you are thinking of and stick with well-known brands.
There are tons of brands in microphones -- most of them fighting for the sub $99 spot. If possible, go to a local music store and try out a few of these. Let your eyes be you guide. There's some GREAT stuff in the $75-$300 range, but make sure you do your homework first.
Rode NT2
The microphone that I have now that I prefer for voice over is the Rode NT2. (Around $200, last time I checked) It does a very good job of capturing the warmth of the human voice and the nuances within it. For ANY voice over job, I would highly recommend using a p-pop filter placed a few inches in front of the mic to stop high-speed air from hitting the diaphragm of the mic. These high-speed air produceds the "pops" heard when saying a sibilant sound (such as the "t" in "tomb" or "b" in "bob") Anytime you hear these booms you can immediately pick out a voice over as amateur.
Compressor/Limiter/Expander
Now we're getting serious. Sure, these types of DSP functions exist in software recording tools, but nothing beats the quality or response of a hardware unit. These can now be bought quite cheaply, and even the lower-end products offer decent performance. However, read a review or two to make sure that no one has complained of excessive noise from the unit you are looking at.
Without getting into too much detail, this unit will "smooth out" the dynamics of the signal you are picking up from the mic. They will drop the noise to zero based on a threshold you set (thereby elimnating the hum of your computer when no one is speaking - called a "noise gate") as well as capping the maximum signal (limiting) and bringing up the level when it is too low (expanding). Additionally, the most important part of these units is the compressor, which takes the lowest signal and highest and squeezes ("compresses") these values towards a smaller dynamic range, giving the impression that the performance was more even than it was. The also accounts for minor variations in the distance between the microphone and the sound source. This is especially valuable in recording a voice over!
Behring Composer PRO
The hardware unit that I use, though admittedly "low end" seems to do a nice job. It is the Behringer Composer Pro , picture above. (an older model than what is available now).
Sound Proofing
There are many different ways to approach this, ranging from the extremely professional....
(From Auralex.com)
... to the extremely "guerilla".
What you need is based largely on the environment you are working. Nothing looks cooler or more professional than to have tons of professional acoustical foam set up in geometric patterns all over the place perfectly balancing the acoustical environment in which you are working. However, for many of us, throwing blankets and carpet up on the walls, though shabby looking, will probably have 70-80% the same effect.
For myself, I noticed that my computer's fans were pretty noisy and that I could hear this in my recordings. So, I built a sound isolation box using pro-grade auralex acoustical foam and fiber board, hinges, etc:
(Coincidentally, if there is interest in the community to learn more about how I did this, please feel free to comment as such and I will do a more thorough write up about this...)
For recording the actual voice, you do NOT want to position your mic and voice talent in line with any hard object that would cause a splash back reflection. Have them talking into a thick comforter or blanket (or acoustical foam!). Surrounding them with soft surfaces really helps. Here is one such set up in my studio:
Notice the towel on the music stand to make sure that didn't reflect back.
Above is "Dr. Willoughby" from "Wildlife Tycoon: Venture Africa" recording a part for the voice over in the retail version of the game. (Available very, VERY soon at a store near you!)
Up next...
Coming next in this series is working with the voice talent. You are all set up -- now how do you give these actors direction? Coming soon!
John Seguin
Composer/Sound Designer
jseguin@seguinsound.com
www.seguinsound.com
For older blogs about sound design and composing for games visit John's Blog at www.seguinsound.com/blog
Greetings! If you have not read the first in this series, please review that section now before proceeding.
Now, the tools! In this article I will lay out for you the minimum tools that I would recommend to record a quality voice over (VO).
Computer
Mac, PC or Linux... choose your poison. Although I profess to be a Mac guy myself, I have no beef with any OS. The bottom line is that you need two high-quality pieces: a good sound interface and a good audio editor. These abound by the dozen. If you are going to be doing this sort of thing for the long haul it would of course behove you to investigate the pros and cons of all of them before making your purchase.
In general, I would recommend looking at a multitracker as opposed to a 2-track editor. The reason being that two-track editors offer a more simplistic environment and often offer advanced sound design and editing features (such as those found in Peak Pro, a favorite sound effect/VO editor of mine. For example, I found the ability to loop a sound with a cross-fade built in extremely easy and very convincing for taking seemingly random sound scape materials (such as crickets in the forest) and slicing it down to a meer 20-30 seconds without it ever appearing to "loop". This is easy enough when you are dealing with sounds that have a very finite beginning and end (foot steps, for example), but becomes extremely complicated when you start working with sounds that are continuous (a train flying down the tracks, or crickets, or a waterfall, to name a few).
Logic Pro 7
However, I have also had lots of success recording straight into Logic Pro, a multi-track editor. The advantage here is that you can then easily mix in additional sounds, music, voice all at once and adjust their balance, pan, EQ, effects, etc. all in real-time.
There are a variety of multi-track editors to choose from. I have compiled an incomplete list of some of the better-known and well-used packages here:
1) Logic Pro (Mac) - www.apple.com/logicpro
2) Pro Tools/Pro Tools LE (Mac and PC) - www.digidesign.com
3) Digital Performer (Mac and PC) - www.presonus.com/
4) Nuendo (Mac and PC) - www.steinberg.de/ProductPage_sb190e.html?Product_ID=2442&Langue_ID=2
5) Cubase (Mac and PC) - www.steinberg.de/ProductPage_sb51ba.html?Product_ID=2442&Langue_ID=2
6) Vegas (PC) - www.sonymediasoftware.com/products/showproduct.asp?pid=965
Sound Interface
Many computers come with built-in sound capabilities. However, the quality of these sound interfaces often varies dramatically. The standard for these types of devices that DO support sound-in is usually stereo-in using a 1/8" jack, digitizing at 16 bit, 44.1khz. This can be considered fairly basic stuff. The 1/8" jack presents an assortment of problems. First, there is no way to provide phantom power throught this jack (which we'll discuss soon when talking about microphones) and the preamp (if there is any) is often very low in quality resulting in poor quality electric guitar/bass direct-in.
In addition, the standard for professional quality sound is 24-bit, not 16-bit. Sampling rates may still start at around 44.1 khz but are often done at 96khz or even 192 khz. These extremely high sampling rates however are probably more than you really need just to record a professional sounding voice over. However, a discussion of sampling rates and their effects on sound quality really could fill an entire article so I'll leave that for another time. (Hint, hint...)
Finding a reasonable quality sound interface that supports 24-bit and phantom powered XLR connection for a condensor microphone should do the trick.
MOTU 828 MKII
The rig that I use is a MOTU 828mkII firewire interface. This is probably overkill for just doing a voice over, but since I have many different sound sources in my studio, the additional channels are a necessity, not a luxury. Here are some other well-known brands in the industry that offer a range of products that should fit the bill in the form of PCI card, USB or Firewire:
1) MOTU - www.motu.com
2) M-Audio - www.m-audio.com
3) Pro Tools - www.digidesign.com
4) Tascam - www.tascam.com
5) Edirol - www.edirol.com
6) Mackie - www.mackie.com
Microphones
Microphones are a very tricky subject. They are very, very difficult to choose because each of them has very much a characteristic "sound". Sure, they can be generally lumped into a handful of categories, but in the end, each has a unique sound. Unfortunately, from a catalog it is often very hard to judge what that "sound" will be. My usual recommendation is to go by word-of-mouth from folks you know, follow message boards the mic you are thinking of and stick with well-known brands.
There are tons of brands in microphones -- most of them fighting for the sub $99 spot. If possible, go to a local music store and try out a few of these. Let your eyes be you guide. There's some GREAT stuff in the $75-$300 range, but make sure you do your homework first.
Rode NT2
The microphone that I have now that I prefer for voice over is the Rode NT2. (Around $200, last time I checked) It does a very good job of capturing the warmth of the human voice and the nuances within it. For ANY voice over job, I would highly recommend using a p-pop filter placed a few inches in front of the mic to stop high-speed air from hitting the diaphragm of the mic. These high-speed air produceds the "pops" heard when saying a sibilant sound (such as the "t" in "tomb" or "b" in "bob") Anytime you hear these booms you can immediately pick out a voice over as amateur.
Compressor/Limiter/Expander
Now we're getting serious. Sure, these types of DSP functions exist in software recording tools, but nothing beats the quality or response of a hardware unit. These can now be bought quite cheaply, and even the lower-end products offer decent performance. However, read a review or two to make sure that no one has complained of excessive noise from the unit you are looking at.
Without getting into too much detail, this unit will "smooth out" the dynamics of the signal you are picking up from the mic. They will drop the noise to zero based on a threshold you set (thereby elimnating the hum of your computer when no one is speaking - called a "noise gate") as well as capping the maximum signal (limiting) and bringing up the level when it is too low (expanding). Additionally, the most important part of these units is the compressor, which takes the lowest signal and highest and squeezes ("compresses") these values towards a smaller dynamic range, giving the impression that the performance was more even than it was. The also accounts for minor variations in the distance between the microphone and the sound source. This is especially valuable in recording a voice over!
Behring Composer PRO
The hardware unit that I use, though admittedly "low end" seems to do a nice job. It is the Behringer Composer Pro , picture above. (an older model than what is available now).
Sound Proofing
There are many different ways to approach this, ranging from the extremely professional....
(From Auralex.com)
... to the extremely "guerilla".
What you need is based largely on the environment you are working. Nothing looks cooler or more professional than to have tons of professional acoustical foam set up in geometric patterns all over the place perfectly balancing the acoustical environment in which you are working. However, for many of us, throwing blankets and carpet up on the walls, though shabby looking, will probably have 70-80% the same effect.
For myself, I noticed that my computer's fans were pretty noisy and that I could hear this in my recordings. So, I built a sound isolation box using pro-grade auralex acoustical foam and fiber board, hinges, etc:
(Coincidentally, if there is interest in the community to learn more about how I did this, please feel free to comment as such and I will do a more thorough write up about this...)
For recording the actual voice, you do NOT want to position your mic and voice talent in line with any hard object that would cause a splash back reflection. Have them talking into a thick comforter or blanket (or acoustical foam!). Surrounding them with soft surfaces really helps. Here is one such set up in my studio:
Notice the towel on the music stand to make sure that didn't reflect back.
Above is "Dr. Willoughby" from "Wildlife Tycoon: Venture Africa" recording a part for the voice over in the retail version of the game. (Available very, VERY soon at a store near you!)
Up next...
Coming next in this series is working with the voice talent. You are all set up -- now how do you give these actors direction? Coming soon!
John Seguin
Composer/Sound Designer
jseguin@seguinsound.com
www.seguinsound.com
For older blogs about sound design and composing for games visit John's Blog at www.seguinsound.com/blog
Recent Blog Posts
| List: | 09/30/07 - When Orcs Attack! Audio Post-Mortem 08/29/07 - AGDC and Organizing your Sound Files 12/21/06 - Scoring the series "Hey, Shipwreck"... 11/15/06 - Music Postmortem for Venture Arctic 06/12/06 - Sound Design: A Primer 06/01/06 - How to Produce Voice Over's that Don't Suck Part IV: Editing and Post Production 03/10/06 - How to Record Voice Overs that Don 03/07/06 - Announcing the SeguinSound Forums for Game Music and Audio |
|---|
Submit your own resources!| Anders Linder-Noren (Mar 01, 2006 at 20:05 GMT) |
| David Montgomery-Blake (Mar 01, 2006 at 20:28 GMT) Resource Rating: 5 |
GREAT BLOG!
| Ajari Wilson (Mar 01, 2006 at 20:35 GMT) |
| Rubes (Mar 01, 2006 at 20:47 GMT) |
| Clint S. Brewer (Mar 01, 2006 at 20:58 GMT) Resource Rating: 5 |
You are much more pro than me, I wonder what you think of my software and cheap microphone set up?
in particular, what are your thoughts on the software noise removal? It sounds ok to me in practice, but I havent' played the effects on many different systems. I've noticed that some sound quite different comming through my little laptop speakers, way too much trebel.
Quote:the question on everyone's mind now!
But are those winged flying hot dogs on that cover in your recording area?
Edited on Mar 01, 2006 21:06 GMT
| abc (Mar 01, 2006 at 20:59 GMT) |
http://www.silentpcreview.com
| John Seguin (Mar 01, 2006 at 21:04 GMT) |
@Ajari: Ha! No, they're supposed to be eagles. :)
@James: Agreed -- having a computer that's mostly silent to START with is a good way to go!
| Phil Carlisle (Mar 01, 2006 at 21:17 GMT) |
I'm trying to find the right operations to do to process my source sound into a radio chatter effect. A google search on the subject returned nothing. Actually, thinking about it, I should have asked the SSR guy (our digital music studio partner), but I forgot :)
Thanks,
Phil.
| Ajari Wilson (Mar 01, 2006 at 21:24 GMT) |
| Jon Frisby (Mar 01, 2006 at 21:50 GMT) |
-JF
| MorpheusX3K (Mar 02, 2006 at 00:26 GMT) |
| Ben -Djaggernaut- Chavigner (Mar 02, 2006 at 09:04 GMT) |
| Mike Fosker (Mar 02, 2006 at 13:02 GMT) |
It comes with a lite version of Pro-Tools which is pretty good, and I'm pretty pleased with the quality of the audio.
Having a decent microphone makes a huge difference like you mentioned.
| John Seguin (Mar 02, 2006 at 21:01 GMT) |
I would also challenge you to try to get a better microphone if at all possible. The computer mic you are using probably has the frequency response to that of a cell phone. So far it may be working, but a little investment in this department will go a lonnnng way.
Also, props on using Reason as your effects engine. Reason is extrodinarily flexible and has very reasonable DSP possibilities built into its samplers. I may have to try to using it in a similar manner that you did. Althought the tools I am using now seem to work fine for me, it doesn't hurt to be able to expand one's sonic pallete!
@Phil: World War II era radio chatter, eh? The would be a culmination of a few things. First, I'd apply some cuts in EQ to the low and high end so that most of the signal you are getting is right around 1-2Khz. (Some manipulation and guesswork will be needed to find the exact poing). This will very accurately emulate the sound they were able to transmit on these ancient sets. I would also add a distortion filter next in my signal chain -- just to make sure your signal is not TOO clean. I would also compress the signal heavily, probably as a last step, as this is also done in radio technology to make things more understandable. If you want to overlay this with some sweeping static/white noise to give it a "bad connection" sound, this is possible, but definitely a bit more involved. Writing a custom alogirithm fo this in MaxDSP might be a good way to start. Hope that helps!
@Ajari: Multiple PC setups, eh? This used to be my big focus -- though after much research I've decided that the better way to go is to probably just buy a very sweet single PC and add a bunch of hardware DSP cards to help with load, along with using a sequencer that allows you to "freeze" tracks (like Logic Pro and others) so as to temporarily bounce the real-time DSP work to a static audio file. (modern disks like serial ATA can accomodate quite a few simultaneous audio streams in playback). When I investigated the multiple machine approach I realized there is no way (to my knowledge) to control "everything" from a single system. So you'll always be bouncing between two systems. Making sure you keep files together, etc -- like opening project "cool game" on system A with the sequencer and also project "cool game" on system B and making sure you save both and then you must deal with infrastructure for MIDI transportation, audio input from one system to the other, latency issues, "jitter", etc.... I decided this probably was not worth it. Maybe in the era of 800Mhz G4 systems, but now with crazy 2+ Ghz G5 dual processors its not nearly the issue it once was. My two cents -- others may certainly disagree. :)
@Mike: The MBox is a nice setup. I had one of these but upgraded chiefly for the additional inputs. In many wayts the mbox/pro tools LE setup is PERFECT for voice over work. Just add a decent mic and (highly recommended, anyways) outboard compressor and you are good to go!
| Chris Labombard (Jun 24, 2006 at 00:14 GMT) |
Edited on Nov 30, 2006 15:08 GMT
You must be a member and be logged in to either append comments or rate this resource.
