Modifying Recorded Dialouge

[TotalRookie]

Ever since I figured out how to get NPC's to have a basic conversation with me I have been wondering how I can get someone else's voice besides my own in a recording. Sure it's cool that my NPC can talk to me, but its always bothersome that it's my own voice, and it definitely does not help with immersion. Not being able to find alternatives to my own voice in dialogue has discouraged me from messing around with the GECK too much. So my dilemma is: what programs can I use to edit my voice and make it sound like somebody elses? My own voice is also distorted a little bit for some reason and I would like to know why. I think it may be due to the fact that I have a $20 headset. If anyone could answer my two problems it would be GREATLY appreciated. I see this as my only roadblock in creating some immersive mods. Thank you again for any advice you may have.

[pkleiss]

How are you making your recordings then? Are you using the Audio tab in a Quest dialogue topic info? If so, then the $20 headset is only half your problem. But anyway, get Audacity and install it. Its free. You can use this to make your own recordings and then just overwrite the files you made in the geck with the ones you made with audacity. They will sound a whole lot better. Audacity also has several effects that you can apply to your recordings to change the way they sound. You can change pitch, tempo and a myriad of other things. I have used it for the same purpose and it works well.

[TotalRookie]

Yeah originally I was recording it through the Audio tab on the GECK. Have you gotten your voice to sound much different using this program? I would also like to know what kind of mic/headset you use and if it will make much of a difference in terms of audio quality. Thanks for your help by the way.

[pkleiss]

I found that the recording made with the geck were barley audible in game. Using Audacity will improve them a lot. I also use a $20 headset, so the quality of my recordings is limited. I have a friend who has a $300 mic, and his recordings are even better, but he lacks a good studio. Just download Audacity like I said and make some random recordings, then play with the effects. That's how I learned. Just make sure that your 'peaks' don't hit the top or bottom on the recording level graph otherwise your recording will get distorted. One effect I use is 'Amplify' which will maximizes the gain (loudness) of any recording without distorting it. If your audio card does not have built in noise reduction (I have the Realtech HD Audio Manager in my system tray that I use for noise reduction and to mute certain channels - like the mic when recording from my PC), Audacity can do that too. You just select a section from your recording that is blank (just the background noise) select get noise (or what ever its called) from the noise reduction effect, then select your entire recording and hit the noise reduction button (or whatever its called) in the same effect, and voilà, you have a much clearer recording. You may have to play with the noise reduction level to eliminate any 'tinny' sounds, but with practice, you can get very clear recording even with a $20 headset.

 

Another benefit of Audacity is the ability to capture whatever is playing on your PC from any website or game or whatever.

[TotalRookie]

That's a lot of good advice there thank you. I played around with audacity and I couldn't really clean up the audio too well for some reason my voice sounds distorted anyways. I blame the mic because the whole Get Noise thing does not remove the weird distortion effect on the voice. I did manage to get my voice sounding like another person which is cool. As of now I created a new sound file and exported it to the existing file being used for my greeting as a .WAV but when i preview it the GECK it still sounds like my original voice? Now if I play it just by finding the directory it sounds like the newly edited one. Yet GECK still appears to be using the newly edited file. I may just be "over-writing" them wrong. It's frustrating me :/ ... What a coincidence I think we have the same Realtech headset lol.

 

Edit: Um nevermind? Apparently using the preview button in GECK once a file has been edited is useless you have to click under path to get the updated one. All I need now is either a better headset or I'm just not cleaning up the my recording right. Seeing as though we (probably) have the same headsets I'm wondering exactly what you do to fix this.

Edited July 6, 2010 by TotalRookie

[pkleiss]

I don't use the audacity noise reduction as my audio card has a control panel with noise suppression already on it. I use that as its automatic and thus, requires no editing. If you have the Realtech HD Audio Manager (mine is by ASUS) in your system tray, double click on it and go to the 'Microphone' tab and click on the noise suppression button. There is a limit to how good the recording will sound with a $20 headset though, but it should be a huge improvement over your Geck made recordings, right? If you can get the room where you make your recording as silent as possible, then you might be able to get away without noise suppression which will sound even better (if there's no background noise). The problem really occurs when you use the Audacity Amplify effect to maximize the gain. It also maximizes the background noise. If you fore go the noise reduction and just speak really loudly, you can get the recording to fit the recording level graph from top to bottom without using Amplify. I would guess this would be the best way to do it, but I can never get up enough gumption to talk that loudly and wake the neighbors. :tongue:

 

Another thing my Audio Manager does is allow me to mute the other recording channels while I'm making recordings with the mic. My settings allow me to mute: CD Volume, Line Volume and Stereo Mix volume. I don't know enough to know if this makes a difference, but I do know that when I am recording from my PC, I must mute the mic (and un-mute the 'Line Volume') or I will pick up ambient room sounds from the mic on a recording from digital media (like a CD or from a website thats playing some audio I want to capture).

 

If you are going to use the Audacity noise reduction and Amplify, make sure you do the noise reduction first before amplifying it.

 

Good luck!

[TotalRookie]

Somehow I managed to get rid of that weird tunnel-like effect on my voice by using another program called RecordPad to record my voice. It had an option to set the microphone as the record device and for Input Channel I selected master volume. It sounds much much clearer now I'll be happy to mess around with some recordings for the next couple days. Thank you for all your help! Kudos given : )

[pkleiss]

Hmm. I just dl'd that recordpad myself and will try it out later (after the World Cup). Thanks for the info. One thing you should know. WAV files are huge, MP3 files cause stutter in the game, so OGG is the way to go and what Beth used for their voice files. You can get an OGG converter for Audacity. I don't know about recordpad. So you may still have to use Audacity to convert your WAV to OGG. For example, I made a radio station that plays 16 songs. The archive of all those wav files was 270 MB. The OGG format seems to be about 1/10 that if WAV files, so its worth it to convert to OGG. But you can use WAV if you want, just stay away from MP3.

[TotalRookie]

Oh wow yeah I can imagine WAV files are big thanks you'll save me a lot of space. I just hope by converting the format to OGG it won't lose its quality. Thanks for the tips I hope recordpad works as well for you as it is for me, let me know.

[BadPenney]

I've been having some fun with sound files, but I don't know about Audacity and its functions. I like to use Sony SoundForge for recording. I suspect that many of the functions are similar.

 

Your microphone is important. Condensor mikes are good. Headsets are low fidelity units and should be avoided. I don't recommend using noise suppression when recording. It is an automated, generic process that removes information before recording it. I would rather record all audio data first and then use a noise gate filter to remove undesirable noise by setting the decibel levels manually, preserving as high a signal to noise ratio as possible.

 

A quiet recording environment is necessary too. You can make acceptable recordings without a sound studio but there is only so much that noise reduction can do. I'm lucky enough to live next to a park instead of an expressway. Using line conditioners on your power supply and turning off electronic devices and fluorescent lights nearby can reduce the amount of electrical hum in your recordings.

 

I use effects in SoundForge to help me get some extra variation when using my own voice. Pitch contouring can make you sound much different; older or younger. I'm even finally having some success altering my voice to sound female.

 

It is also good to be a ham actor and to have a recording space that is not only quiet but private enough that you can make some noise, even shout if necessary. As long as you set your recording input levels such that you don't get hard clipping on the peaks then your recordings are usable. You can't mutter words and then just increase the volume to generate dramatic dialogue though.

 

It is important to have fun when you are recording dialogue. If it feels more like a job than playing like a kid then your recordings will probably be boring no matter how good your gear and software.

 

If you don't know any friends that are silly enough to come over and play, then you might still be able to get usable vocal files of other people from the net. I like to use Marcel Duchamp's "found object" approach for dialogue sometimes, which in this case involves searching through open source creative license recordings for interesting bits, and then finding a use for them as mod dialogue. There are a surprising number of audiophiles recording all manner of high fidelity files and uploading them for others to use. Splicing and mixing the source files can provide even more material.