Page 1 of 3

[Sound] Whip me, beat me, critique me

Posted: Sun Jan 08, 2012 7:30 pm
by Gradius
My goal is to get believable impersonations of the original SC voices through sound editing. I use adobe audition. I'll post more about my progress. Any help/feedback would be appreciated.

I think this is pretty close, but it's missing the raspy voice effect, which I've looked everywhere for. I can't decide if it's more dynamics processing, or just screwing around with the graphic equalizer. The other challenge with Artanis seems to be the pitch. He is actually pretty low pitched, but sounds youthful, and I can't do youthful without setting the pitch higher.

Dark Templar:
I don't want too many echoes flying around, but this is clearly missing the "thickness" of the original, same as with the Artanis. Looking for the effects Blizzard used is like looking for a needle in a haystack, the only way to copy it is to stumble upon the exact effects they used. =/

PS: Do you guys recommend getting books on sound theory, or just experimenting? I know a bit about sound, but not enough to really know what I'm doing here.

Re: [Sound] Whip me, beat me, critique me

Posted: Mon Jan 09, 2012 9:18 am
by IskatuMesk
I've never read anything, much less anything about sound theory. So, no, I wouldn't say it's necessary at all. Learning will come with experimentation and experience.

Re: [Sound] Whip me, beat me, critique me

Posted: Mon Jan 09, 2012 6:38 pm
by Lavarinth
Artanis- I would recite that line again without the pauses, you got down how he elongates words rather good, that was definitely his key difference between other characters in the way he spoke. I'd find Artanis to be rather fluid with his voice, so the pauses are throwing me off.

The Dark Templar is also well down, but as you mentioned needs that little UMPH for make it greater.

As for what to do to improve your own skills- It's asking and experimenting from my experience.

Re: [Sound] Whip me, beat me, critique me

Posted: Thu Jan 12, 2012 3:41 am
by Legion
I absolutely love the echo/delay you gave Artanis. Also, you've a mean voice for Protoss.

My impression of Artanis.

I usually speak like a retard to get close to that weird idiolect he's got.

Quality is a bit low, but was only trying to put in my 2 cents.

Also, I can't do DT's, but I've been messing around with HT voices for ages...

Re: [Sound] Whip me, beat me, critique me

Posted: Thu Jan 12, 2012 9:42 am
by Falchion
It's been some time since I've been away from Audition or the sort, but I've been working on a separate stuff for BW, so I re-installed it and gave a try on two things.

The first is a concept Dark Templar Character named Cyrth, while the second is a placeholder Cerebrate.

Please ignore the voice acting, what matters for me is the effects, which I've been putting effort. Still, I think DT could use a re-work.


Re: [Sound] Whip me, beat me, critique me

Posted: Fri Jan 20, 2012 4:20 pm
by Gradius
Well here's an update from me:
Same thing but with change to voice acting. Haven't really been able to 1-up this one in effects yet.

This voice is damn hard to replicate. I just don't know what effects they used. This is the best I could get after several tries.

Legion - those are all great. Your Artanis voice is pretty much what I consider the gold standard for fan editing. I've been listening to this sample you posted a few years ago: ... ample5.mp3

Seriously, that's a really good Artanis. :P

Falchion - for a non-canon character the DT is perfectly acceptable. It's got pretty unique voice & effects. The cerebrate is alright, but again, it's hard to replicate Blizz's exact effects.

Re: [Sound] Whip me, beat me, critique me

Posted: Sat Jan 21, 2012 9:35 am
by Lavarinth
The only thing that stands out immediately is the pronunciation of "Executor." You say "ex-eh-cue-ter" while Artanis says "ex-eh-cue-tore."

Re: [Sound] Whip me, beat me, critique me

Posted: Sat Apr 21, 2012 7:39 am
by IskatuMesk

You have no idea who I am in most likelihood but in either event I will, despite the ancient date of this thread, offer some thoughts on at least the last provided samples. You wanted critique, so I will dip into all of the aspects I feel are of note. The fact you are using audition right off the bat is a boon to you. <---

I have seen many amateur voice actors go for a DT first and to me they always seem to make the same mistakes. I'll dissect this and kind of explain why.

The echo stands out the most. It is distracting and sounds very un-protoss-like. I believe a shorter reverb would suit this sample and make it 2x as good as it currently is. Avoid using echoes until you are very comfortable with reverbs, because echoey sounds often do not suit protoss voices very well, and it's very easy to make them sound "off". Likewise, there is no pre-reverb for the "psionic" effect, but I suspect this is intentional.

Following the echo mentality, the voice sounds somewhat metallic, mechanical. I could have a use of a voice that is like this - but not for a protoss, rather a gundam, let's say. The mechanical effect seems to come from a variety of natural reverb (your recording environment) and the echo.

Following this, it sounds like you altered the pitch of your voice. Pitch alterations, especially deepening your voice, results in a very flat sound. This is not as bad as many voices (like the immortal in sc2) for "flatness", but you may want to investigate using either a compressor (dynamics processing), wave hammer (directX plugin) or EQ to kind of try to restore some fidelity to the upper regions.

Personally, I use a DirectX plugin called Elastique Timestretch. I find it to be far more capable of producing good, believable pitch alterations than the stock stretch filter. Additionally, I've also begun to experiment with Autotune to this end, but I find it cannot give me a deep voice very convincingly (at least with my very limited experience). My common scheme for altering pitches and deepening them is usually like this;

Apply Elastique Timestretch with a max of -3 unless it's a heavily edited voice
Dynamics Processing for a outspoken character or Wave Hammer for a close-up, moderated spoken character.

Wave Hammer functions a little differently than the stock compressor and has less options than it, but can produce some interesting effects.

Lastly, I think you need to put some more emotion into the acting itself. It sounds like you're reading from a script. The first few lines, especially the formalities, followed by the dialogue, seem static and lack conscious thinking on the character's behalf.

To summarize, the things I think you could improve with this sample;

- Lighten the echo, focus on a smooth reverb that is shorter. Experiment with pre-reverbs to add that Dark Templar whispy sound. Tight echoes with specific settings, especially emphasis on higher frequencies, can work too. Check out Delay as well.
- Try to centralize your voice and avoid the "distant mechanical" kind of sound, this could be related to your environment, your distance from your mic, your mic itself, and the effects you're using.
- Consider experimenting with compression or EQ to prevent deepening your voice from sounding flat. Check out DirectX effects and you might find something cool.
- Practice your acting! You can never practice acting too much. Just ask my neighbors. They love my constant screaming and swearing. <---

Speaking independently of replicating Artanis' voice (Or Legion's which I feel is overprocessed);

I feel the recording quality of this is greatly superior to the second one, likely to the lack of a huge pitch shift. However, the trailing echo sounds very out of place for a Protoss - it reminds me somewhat of my Undead and Xul`Amon effects. To me, this is a much more solid sample overall than the second one. You should consider trying to tighten it or, again, attempting to use a reverb in its place.

I do not think you need to force yourself to speak so slowly in the beginning half. It sounds forced to me, out of character, if you will. This is more of a nitpick, but it's something I've been working on a lot myself - pacing. Entaro Tassadar also sounds a little "rehearsed", so to speak - it lacks the kind of empathy behind the words that I would expect from them.

Anyways, hopefully I don't come off as too much of a dick. I'm not so good with words. I know these are old, but I wanted to eventually get to them. I'm glad to see new people attempting voice acting and I will gladly see to any additional samples you wish to provide time permitting.


I made a very quick and dirty dry run on some random lines to show some of my simple reverbs. ... xample.mp3

Re: [Sound] Whip me, beat me, critique me

Posted: Sat Apr 21, 2012 9:07 am
by Gradius
That is exactly the kind of feedback I've been looking for. I've read some of your articles before as well as listened to your podcasts here, and I like your "Imrian" voice, so I appreciate you taking the time to look at my samples.

The reverse reverb is indeed something I stumbled upon myself in my own experimentation, and you're right, it is much better than the echo effect. The default echo doesn't seem to allow for much alteration, and the "room" echo filters are just overkill. I will take a look at some of those plugins you mentioned though. Hopefully they're free. :P

For the DT I try to copy the original and lower the pitch of my voice naturally, but it's something I can't ever seem to add emotion to. I'll keep practicing the acting.

Here is my latest sample since then:

Re: [Sound] Whip me, beat me, critique me

Posted: Sat Apr 21, 2012 9:27 am
by IskatuMesk
In these voices again, sometimes it sounds like you are forcing yourself to talk slower. This is useful - especially with heavy echoes and such, to avoid mashing the tails into each other - but for some it's not necessary I don't think (especially if you decide to go with shorter reverbs instead of echoes). Especially if you want to make certain voices more distinct. I think the weakest would be Meliak and the strongest would be Khraxus. I actually really like Khraxus' voice. Really, at this point, it's going to come down to constant practice to identify weaknesses in acting and strengthening them. For me, it's very hard, because other than ventrilo I have zero talking contact with people outside and have spent a decade in isolation. So, my speech defects are hard to overcome. I practice very very often, talking to myself constantly. But it's not always enough.

I don't think elastique is '''free'''', but honestly I can't even remember how I found it. Spaceman should be free, though, and is a great addition if you want to try Terran voices (for com static). A large number of directx effects become available if you install soundforge and turn on the directx effects in audition's view menu.

Here's some settings I use for voices like the one above.


Remember, the frequency settings in Echoes can be used to make them sound more whispy (using only high end frequencies) or more basey (only the low end), and playing with the balance of these can help remove the typical "cheap echo" effect.


This is one of the reverbs I made for the sample above. Playing with the high-frequency absorption, the Reverb (Wet), and attack times and such will yield very different results.

Re: [Sound] Whip me, beat me, critique me

Posted: Mon Dec 10, 2012 9:44 am
by Gradius
Alright I'm back. Finally approaching the voice-acting stage in my project and it's time to get these nailed down. Here is the high templar:

My biggest concern: is the reverb tail way too noticeable?

I've looked for elastique & other time-stretch plugins, but I have none that work. One will let me preview but freezes when I try to apply the effect. That aside, I'm not sure what such an effect does that's different from regular pitch shifting. Are you supposed to increase the speed while decreasing the tempo to get a lower pitch? I ended up just using auto-tune for shifting the pitch.

Also, noise reduction. I'm sorry, but this crap where you record a noise sample to calibrate your noise reduction filter doesn't do much for me. Using dynamics processing is really the only thing that actually eliminates noise.

Re: [Sound] Whip me, beat me, critique me

Posted: Mon Dec 10, 2012 11:29 am
by IskatuMesk
Dynamics processing really doesn't eliminate noise... in most instances it makes it worse. I guess if you have insane background noise... but even then, dynamics is a compressor.

As for Elastique, its just a more refined shift. The tempo is generally something you won't be changing unless you want to speed up or slow down.

As for your file;

- The reverb is much too echoey. The voice itself sounds oddly hollow, as though you have cut out the low-end of the frequencies.

Autotune causes noticeable distortion and while I use it for specific stuff I don't really use it for humanoid voice shifting.

Re: [Sound] Whip me, beat me, critique me

Posted: Fri Dec 14, 2012 5:29 pm
by Gradius
I'm going to have to stick with the regular pitch shifter. I haven't been able to get any other third-party pitch shifters to work.

Here's the best high templar I can do for now:

Not much of an improvement, but it's a hard effect for me to replicate. I tried increasing the higher-end frequency of the echoes like you said but I'm still failing to get the "whispiness". I'll keep experimenting. ... aceman.mp3

Trying to do a ghost. I used the spaceman filter for the second one. But I think the first one sounds better.
This is my rendition of SC2's executor/advisor:

IskatuMesk wrote:Dynamics processing really doesn't eliminate noise... in most instances it makes it worse. I guess if you have insane background noise... but even then, dynamics is a compressor.

So what does the "noise gate" preset do then? Reduce the volume of the noise right?

Re: [Sound] Whip me, beat me, critique me

Posted: Fri Dec 14, 2012 6:47 pm
by IskatuMesk
I have never used noise gate before, so I have no idea. All I know is that noise removal in any quantity warps the voice. If you're having trouble removing huge amounts of background audio, dynamics processing will no doubt mangle the voice in the process.

Editing is all about experimenting. I find my dynamics profiles don't work at all for other people's voices. I can't use my presets on Laconius' voice, for example. Spaceman in particular needs a few hours of work. Once Crainy's machinima comes out you'll see some of my more tuned effects. Spaceman is an old plugin, used widely in professional environments, but it's not as super in-depth diverse as newer plugins might be I imagine. However, I can't get many newer VST/DLL's to work on Audition 1.5.

The thing about effects is that they don't make the voice. They compliment it. The High Templar needs to be spoken in a whispy voice for the effects to support that. Your echo is too echoey and not smooth enough. I recommend trying reverbs instead. But foremost, practice making as close to the templar voice as you can without effects.

The Spaceman effect you are running is rather intense. Try playing with the Drv and hi/lo pass options. Otherwise the ghost does sound fairly good. By the way, com static would apply to the breathing effect. It completely breaks any kind of believably without that.

Again, the Executor sounds like a pipe-y echo. This masks the voice rather intensely and it's hard to figure out.


Although it might come too late to make a significant difference in whatever warcraft 2 campaign you're making, I will eventually make an MGEC on these voices. I'm sure my youtube already has one or two on protoss voices if you look hard enough.

Re: [Sound] Whip me, beat me, critique me

Posted: Tue Dec 25, 2012 6:58 am
by Gradius
Saw your work on Crainy's trailer. Nicely done.

Here is my latest tries at a high templar with better voice acting:

I think the first sounds good, but has too much of a robotic sound which I tried to fix in #7.

Executor again:

Not sure what else I'm missing here. It seems that Blizzard's version is just a chorus effect with some fancy mastering, but I can't match their version.


Terran voices aren't my thing, but there's the goliath.