Source Filmmaker

Source Filmmaker

LunarSC Feb 10, 2013 @ 2:44pm
Extract Phoneme's not working
I'm using scout and heavy, both are HWM's the sound clip is scouts no03.wav
Here's a pastebin of the error:
I have ONLY the sound clip I want as show waveform. However, if I select the bladed shot instead of the the sound clip (which is on show waveform) and try to extract phoneme's without having the sound clip selected, it will say "At least one sound clip must be selected."
I have only the face selected in the animation set editor.
I also have no phoneme tab.
Please help?
< >
Showing 1-12 of 12 comments
Pte Jack Feb 10, 2013 @ 3:04pm 
Lunar, can you post a picture of your SFM session with things selected the way the program is causing the error? The scout you're trying to give the sound clip to, is he a scout from a recorded, game mode, scout? Which shot is active in the blade when you try to assign the wav, the one where the scout is talking or the other?
LunarSC Feb 10, 2013 @ 7:15pm
Sorry that it's small, with dual monitors, alt+prntscrn would take away the right click window.
R234 Feb 10, 2013 @ 8:05pm 
Try right clicking the Scout's animation set > show in element viewer > animationSet, then in the element viewer check if there is an array called presetGroups, which when expanded contains three DME preset groups, namely emotion, phoneme and viseme. If not, try creating a new project and importing an HWM Scout, and do the same to see if he now has the preset grops. If he does, then it is your project that got messed up, as some people have reported happens sometimes. If he doesn't, you should try deleting your sourcefilmmaker/game/tf_movies/models/player folder and validating your SFM cache files to redownload the models, since they're most likely what's causing the issue.
Pte Jack Feb 10, 2013 @ 8:16pm 
I think I see your problem,

Photo 1: you've selected your shot that you want the model to talk in and moved the playhead to the shot (correct). The models for that shot appear in the Animation Set Editor (correct), you've selected the scout (correct), but it appears that you have only selected the lower face and the tongue (incorrect, select the FACE heading)

Photo 2: You've selected the sound file and made the waveform visible. I'll assume you've turn the waveforms off (hidden) in the other sound clips. (Correct) But you still only have the lower face and tongue selected on the model. (incorrect) This is the point you should extract the Phonemes.

Photo 3: You've reselected the shot in the Clip editor (incorrect) and tried to extract phonemes. (incorrect). As soon as the sound clip turned green it is no longer selected, therefore when you try to extract the phonemes, there is no sound file selected. SFM is right and tried to tell you that. Also, again, it appears that only the lower face and tongue are selected. Select the FACE heading.

Solution: Select the shot in the clip editor that your character will speak, move the playhead to that shot
Select the character and expand his controls in the Animation Set Editor,
Select the face heading
Ensure that all the other sound file have their waveforms hidden, then show the waveforms only on the clip you want spoken, recalculate the waveform, just to be sure.
with the SOUND clip selected (yellow) right click the character's face control and extract the phoneme.

See if that works...
Last edited by Pte Jack; Feb 10, 2013 @ 8:18pm
LunarSC Feb 11, 2013 @ 12:38am 
Hey Pte Jack, thanks for trynig to help but I'm still getting the same problem, even though I did everything I was told. I hope these screenshots help
Pte Jack Feb 11, 2013 @ 2:00am 
Lunar, I am thinking that you don't have the hwm models in your shots like you think you do...

I put a player model and a HWM model into stage.

I attempted to extract phonemes to a player model and received the missing preset group error.

I then extracted the phonemes to the hwm model and it worked properly.


Here is what I want you to try.

Right click your character and and show the MODEL in the element viewer.

Look for the element named "modelName" and read what is says in the box beside it.

if it says models/"PLAYER"/(character).mdl then this is your problem. It's not the hwm model.

if it says models/"PLAYER/HWM"/(character).mdl then you do have the hwm character in the scene and for whatever reason, it looks like your hwm characters have lost their phoneme sets.

I'm trying to find a quick solution for you if you find that it happens to be the wrong model.

If it is the right model, I would suggest you save everything (and make a backup as well), then validate the local application cache file. It looks like the hwm models or SFM did not install properly. Once the validation is complete, restart SFM and try to extract your phonemes again.

Pte Jack Feb 11, 2013 @ 2:42am 
Here's the test and how to check your model type...


I attempted to change the player model into a HWM model in the element viewer. Because there are bone differences between the 2 models, the hwm skewed in some places and did not look right. SO...

If required, there are only two ways I can suggest that you might be able to switch between a player model and a HWM model make it work in your animation.

The first is to create the HWM model animation set in the shot. Then in the motion editor with ALL OF TIME selected, select the player model. Make sure you can see the entire clip. Copy the animation samples.

Switch to the HWM player, again make sure all of time is selected and paste the samples. This will (should) move the hwm into the same position as the player model and give it the animation you've created.

Turn off the visibility of the player model and play the animation. Watch for any skewing. If everything looks good, delete the player model... You'll have to do this for each shot in your clip that the character has to speak.

The other way is lock every single bone between to player model and the HWM (you would drag the bone from the player model to the HWM to make the lock) and then select the hwm and zero out the the Zero to move the hwm into the correct position and then zero out the playhead to give it the animation. After that go through and unlock every bone, then delete the player model.
Last edited by Pte Jack; Feb 11, 2013 @ 3:12am
Pte Jack Feb 11, 2013 @ 7:14am 
Looks like you two are having the same problem, I've referred him here
LunarSC Feb 11, 2013 @ 2:52pm 
Hey buddy thanks for the help, I started out with regular mdl's, used google to find out they needed to be hwm, but apparently my replacement models didn't work. I started a new load and it works fine now. However, could you help me with my microphone? It's VERY quiet in SFM when I play it back, and there's no way to adjust the sensitivity to go higher. I have turtle beaches x12. Thanks so much!
Pte Jack Feb 11, 2013 @ 4:23pm 
Make an actual export of your movie before you go playing with volume, you may find quite a surprise. If you are doing your own voice overs (vo(s)) I would use something like audacity. That way I can control levels and add the effects I want (like echo, reverb, etc). Do not exceed a 16bit wav recording when you export the vo, 24bit and 32bit wavs don't work in SFM.

As R234 pointed out, MP3s work for adding sound effects and music but don't work for phonemes. The recording for phonemes has to be in .wav format.

If you're using SFM to record your dialog, here's the developer page on the topic.

If you find you have to adjust the volumes of your .wavs in SFM, right click the sound file,right under Play Sound and just above Show Waveform is Set Volume. Standard volume is 1, 0 is off, so you can ramp up the volume by going greater than 1 and ramp down the volume by using .99 to .01.

Please don't ask how to make a sound ambient, I can't answer that one. I'm not there yet.

LunarSC Feb 12, 2013 @ 6:04pm 
I'll use audacity, thanks! I'm not sure if this is making a sound ambient, but how can I change my voice to sound like heavy or scout? thanks in advance if you can answer it :3
Pte Jack Feb 12, 2013 @ 6:58pm 
That my friend is you... You can try the various effects in audacity, but voice acting is just that, either try to sound like the scout or find someone who does...

You can post a request in the discussions here that you're looking for an actor, then provide your script to the people that respond. Now whether or not they request some kind of compensation, well, that's between you and the actor.
< >
Showing 1-12 of 12 comments
Per page: 15 30 50

Date Posted: Feb 10, 2013 @ 2:44pm
Posts: 12