Are you recording both L & R channels or adding them to produce a mono track?
If adding them, maybe you have swapped the phase of the stereo audio channels somehow. Mono sounds like voice will cancel out but any background music that might have a stereo componente will get through.
Does the music sound exactly the same or does it sound hollow or strange as though some instruments are missing?
I'm assuming you've actually used keyframes to change the volume levels of each track as needed.
If that's the case, go to the Window drop-down menu and open the Audio Mixer. As I say in my books, in this Audio Mixer you can monitor the audio levels of every track.
When voice and music are playing at the same time, you usually want your voices peaking at about zero and your music betwen -6 and -10.
Just to make sure that I am clear of what you are doing, and what you are ending up with, I will restate what I think that you are saying. Please correct me, if I am wrong.
You have downloaded some AV files. These already have muxed (multiplexed - combined Audio and Video) Audio already in them. The muxed Audio consists of both music and vocals. When you play these AV files in VLC, or WMP, the levels of both the muxed music and vocals are fine. When you Import these files into PrE, suddenly the vocals, only, drop in level, but the background music retains the perfect level. Is that correct?
I can think of no reason that this would be happening, unless you have added EQ (Equalization), which is not easy to do in PrE, or other Effects to the Audio, as PrE will use the levels in the muxed Audio, and will not arbitrarily lower just certain frequency ranges, like human voice. Without input from the user, the levels are just the levels from the muxed Audio stream.
Now, are you applying any Audio Effects?
Are you adjusting the Volume with either Clip, or Track Keyframes?
As a test, Open Audio Mixer (Window>Audio Mixer), and play your Timeline. Do the levels of the Audio look good? Do the sliders on your Audio Track move on their own?
Because the muxed Audio stream is the combined frequencies of both the background music and the vocals, adjustments will affect ALL frequencies, unless one adds EQ, Filters or Gating, that affect just certain frequency ranges. Keyframes on Volume, whether Track, or Clip, will affect ALL frequencies, though the human ear might find that some sound lower, or higher, than the others. Without your intervention, the level adjustments should affect ALL frequencies equally, so that is why I am at a loss to explain how just the human vocals are being lowered.
One thing that you can do is to add EQ to the Audio, bumping up the 125Hz to 250Hz range. This is where most human speech will be located. You might hedge your bets, by boosting everything from 100Hz to 500Hz. Unfortunately, PrE does not offer a full Equalizer, unless one finds a VST plug-in that will work, or if Adobe has added EQ, like a Parametric, or Graphic Equalizer. You can do this in other audio-editing programs, like the free Audacity, and there are many VST EQ's available, and many are free.
One possibility is that the muxed Audio is mixed for a 5.1 soundtrack. If so, then most of the human speech will have been mixed to the Center channel. As you will be editing for a stereo (2-channel) Master, without the benefit of the location of 5.1 SS, the Center channel signals will then be spread to the L & R, and can diminish, as they were originally located almost exclusively in the Center Channel, which no longer exists.
I also find that for most commercially produced videos, even with full DD 5.1 SS, the mix favors the music and SFX over the human speech, which focuses to the Center Channel. For that reason, I have a boost in my home theater systems for the Center Channel. Hollywood is too heavy into boosted music and SFX for my tastes, as they want to "immerse" the viewer/listener in their Audio tricks. I have to alter their mix, to clearly hear the human vocals. When I mix, I favor human speech, and boost that, but I am working with separate individual Audio Tracks, and not combined, muxed signals.
When setting up the project options I notice there are options for choosing normal or 5 channel audio.
Whatever you choose should be compatible with your original material.
Referring to the previous poster, maybe most of the voice is on a channel that has been 'discarded'?
Thanks everybody, this is for Bill Hunt...
1st of all thank you for your great effort and time. I must admit that I am a complete newbie to all of this so most of what everyone says may not make sense you know? Not familiar w/ the lingo yet I suppose hehe.
Basically, this is what I just did/tried.. I brought in a 720p DLed movie into Elements 9 in both Stereo and 5.1 settings and all that, still voices are pretty much not there and there is also a quality issue w/ the overall audio, soundtrack, effects and all. I don't know what this is all about. I've uploaded smaller things like Dled t.v shows and those worked just fine, it's always these movies that are causing the problem.
I have a 5.1 sound system but I do most of my work at night when ppl are sleeping and I cannot be noisy. Do you think hooking that up and trying this over again would help and that I would actually be able to hear something?
I tried using headphones when uploading on both stereo and 5.1 but no change.
Tried to do a Smartfix but says it couldn't because it's missing audio in foreground or background tracks.....huh?
I looked @ the details on all my movies that didn't give me clear audio, for one movie under audio it said 6 channels @ 48kHz. 6 channels? I'm assuming 5.1 or something, *shrugs*. Anyways it works on my 5.1 speakers just fine when viewing in a player and using either the speaker or headphones.
The files that did work said 2 channels (stereo) @ 48kHz.
Sorry if I digressed Bill. I did that test you told me to do, I opened up the movie in timeline and looked at the audio mixer, the bar is set at zero on audio 1 and when they talk, it almost peaks at -20 and during loud battles it can get from -12 to close to 0. Keep in mind this is all still lower overall then when in WMP or VLC.
I think what you said here Bill is what I'm suffering from: "One possibility is that the muxed Audio is mixed for a 5.1 soundtrack. If so, then most of the human speech will have been mixed to the Center channel. As you will be editing for a stereo (2-channel) Master, without the benefit of the location of 5.1 SS, the Center channel signals will then be spread to the L & R, and can diminish, as they were originally located almost exclusively in the Center Channel, which no longer exists."
But like I said earlier, I've only tried this w/ headphones and not my 5.1 SS yet, but will that be the fix of all this? Does that mean I'll only be able to edit things like this properly using the 5.1 SS instead of headphones ever? *SHrugs*
Hope I made sense look forward to hearing from you, thank you once again everybody!
If the original Audio is 6-channel, there is one big drawback to using PrE - it will only do a 2-channel/stereo mixdown. There is no real 5.1 (6-channel) editing, and no Export to 6-channel. PrE will Import 5.1/6-channel Audio, but that is where its capabilities end. [PrPro will allow 5.1 editing, BUT one must purchase the Minnetonka Audio SurCode DD 5.1 SS encoder plug-in, about US$250, to have the capability to Export to DD 5.1 SS AC3.]
One option might be to do a multi-track Audio-only editing in a program, like Audacity, though I do not know the full capabilities of that freeware program. I use Adobe Audition, so only know its capabilities. There, one could do the mixing, boosting the Center Channel, and then doing a mix-down to 2-channel. I am not sure what other Audio-editing programs might offer full 5.1/6-channel editing, but would assume that Sonar, Cakewalk or ACID would, but they are not freeware. An option from a multi-track Audio editing session, would be to Export as 6 discrete, individual 1-channel (mono) WAV files, Import those into PrE and adjust, as is needed. [This is similar to what I do with DD 5.1 SS streams, when I also need to create a DTS Audio stream - my SurCode DTS encoder will only work with the 6 discrete streams, producing a DTS mixed stream.]
With a good 5.1/6-channel mix, one probably ends up with human speech at ~ 80% Center Channel, and then ~ 10% L and ~ 10% R, with only a tiny bit of Ls and Rs signal. If you are working with the mixed-down sources, then you end up with 50% L and 50% R, at very best, and that depends on the settings of the mix-down to 2-channel.
Wish that I had an easy solution for you.