12 Replies Latest reply on Oct 26, 2012 7:43 AM by josephs51576386

    How to efficiently separate audio track

    philip242

      I'm new to Premier Pro and to video editing in general. 

       

      I have a single camera interview, audio from on-board mic.  The questions are all too loud and the answers too soft.

       

      I started keyframing the volume but this seems tedious and error prone.  I'd have to be careful to lower or boost each section by the same amount.  I'm thinking if I split out all the questions onto a separate track, I could adjust each track's volume in one feel swoop.

       

      The way I'm doing this is:

      1) Select Audio 1 only

      2) Mark In/Out points

      3) Lift

      4) Unselect Audio 1, Select Audio 2

      5) CTRL-V

      6) Drag pasted clip to align it into the hole

       

      I'm wondering if there is a faster way?  Seems like I'd like to mark In/Out then just drag the portion of the clip down to Audio 2, keeping it aligned to the same point in time,  is something like this possible?  Or just a way to streamline my steps? 

       

      Thanks.  Just getting started so any pointers welcome.

        • 1. Re: How to efficiently separate audio track
          josephs51576386 Level 3

          Is the audio stereo? Also is one person in the left side and another person in the right side? If one person is left and the other right you can simply copy and paste the audio and put the exact same track on 1 and 2 then pan one to the left and the other to the right. Then turn down the one that is too loud and add gain to the one that is too quite until you get the desired volume levels. If adding gain adds a lot of "hissing" usually a denioser will help remove this a bit. If the background noise of AC or cars passing becomes louder or any deeper tones you can use a highpass and set it to around 140 if needed. Although it will take the lows out of the persons voice so it's best to avoid using it if possible.

           

          If they are both running on both sides one thing you can do is a multicam edit and set it to audio follows video. This way you it will make the cuts for you because you'll be selecting right as they talk back n forth. Although truthfully I'd recommend just using a compressor (dynamics effect). You can set your compression ratio to around 4:1 or 6:1 if needed and then set the threshold to w/e you want it to start reducing the volume at. Then add some makeup gain so that the lows are made louder. This should most likely correct the problem. You'll need to set the attack as low as possible, then also set the realease as quick as possible.

          • 2. Re: How to efficiently separate audio track
            Stephen_Spider Level 3

            Here's a tip for you. If you don't have an external mic for interviews, move the camera right up into the subject's face, as close as you can get while keeping good focus. That will at least make the best out of the onboard mic's capabilities. If you don't record good audio, you can't make it good, just better.

             

            I agree that the dynamics effect is likely your best bet here. You might add a highpass filter set at 75 to 150hz and light noise reduction with the denoiser.  ComputerNovice25 is steering you in the right direction.

            • 3. Re: How to efficiently separate audio track
              Felix CL Level 3

              Since you recorded with an onboard mic, you will probably not have the 2 people on different channels, so opanning might nopt be an option for you.

              I will try to give you a fast solution for what you are trying to achieve.

              First of all: if you want to fade questions and answers, you might want to keep them in the same audio track, but split them up.

              a quick way to do so is

              • with only the Audio track, you wish to edit selected (deselect all other video and audio tracks) play back your sequence.
              • use Ctrl+K (PC) or Cmd+K (Mac)  [this is the default shortcut] to add edits to the audio clip between Questions and answers as you play through
              • it helps to then select all Answer-Clips and  go to "Edit > Label > (color)" so you can give them another color to find them more quickly later.
              • now you can adjust the audio of one of them and then copy the audio effects to the others (select all of them and paste)
              • select all audio clips and go to "Sequence > Apply Default Transition to selection" ( you can set default transition and length in the settings)

              Done.

               

              ---

              ...and keep smiling

              • 4. Re: How to efficiently separate audio track
                philip242 Level 1

                Thanks for 3 good ideas. It is mono audio. I haven't even looked into the audio effects yet: compressor, high pass, denoiser etc.  Sounds promising.  And using CTRL-K to chop up the clips in-place seems faster than what I was doing. 

                 

                So no one recommends splitting Questions or Answers out into a separate audio track altogether: what I was doing!

                • 5. Re: How to efficiently separate audio track
                  Felix CL Level 3

                  So no one recommends splitting Questions or Answers out into a separate audio track altogether: what I was doing!

                   

                  it depends, how much difference there is in the audio level.

                  since it's an onboard mic, you may have some background sound so you will want to fade the clips to avoid hard changes in overall level and it's faster to fade them all if they are in one track...

                  ---

                  ...and keep smiling

                  • 6. Re: How to efficiently separate audio track
                    philip242 Level 1

                    Anyone tried this "speech volume leveler" in Audition:

                     

                    http://podcasts.creativecow.net/adobe-premiere-tutorials-podcast/audio-problems-balancing- speech-levels

                     

                    Certainly the name sounds promising, and in this tutorial he gets nice results.

                     

                    Maybe it is just the equivalent of doing compression, denoise, high pass manually in Premier Pro itself?

                    • 7. Re: How to efficiently separate audio track
                      Allynn Wilkinson Level 2

                      philip23342342 wrote:

                       

                      Anyone tried this "speech volume leveler" in Audition:

                       

                      http://podcasts.creativecow.net/adobe-premiere-tutorials-podcast/audio -problems-balancing-speech-levels

                       

                      Certainly the name sounds promising, and in this tutorial he gets nice results.

                       

                      Maybe it is just the equivalent of doing compression, denoise, high pass manually in Premier Pro itself?

                      I **love** the "speech volume leveler" in Audition!  It's saved me a million times.  I'm actually editing three tracks of directional mic audio from a lecture right now and this tool is really teasing out the student questions. 

                       

                      I think this might be the best way to fix the problem here as well.  Just a couple of words of caution.  Do use the option to "boost low signals" but also make sure you put a hard limiter on it *afterwards*.  This ensures no peaking because the speech leveler will boost the questions too.  The "boost low signals" will also pick up other ambient noise and boost this so an Adaptive Noise Reduction on the whole thing is a good idea.  You should be prepared to tweak this live as you go.  Rememeber that order counts in Audition (and Pre effects).

                       

                      Here's a clip of my rack (*blush*)

                      audition_rack.jpg

                      • 8. Re: How to efficiently separate audio track
                        josephs51576386 Level 3

                        Basically the audition method will achieve the same result as compressing/limiting the audio inside Premiere Pro. Also the hissing you noticed that got added in his tutorial will be removed inside Premiere by using the denioser in my own experince anyways. You just have to simply set the reduction to around -10 or so. Main difference is that one method uses audition and the other method allows you to just stay inside Premiere Pro. However for certain tasks I do feel audition is the better option. For your specifc issue though, I assure you using the premiere pro dynamics effect along with denioser is plenty. I generally only use audition if I'm needing a remove/isolate a specifc noise. For instance it does a great job removing AC noise because of it's capture noise print feature. But for simply leveling audio I personally prefer just using Premiere pro's dynamics effect with the compressor/limiter/makeup gain.

                        • 9. Re: How to efficiently separate audio track
                          philip242 Level 1

                          Wow this is great, thanks all.  I tried both Audition and the compressor in Pro.  Both equalized the volumes in the two speakers very well. 

                           

                          I agree staying in Premiere Pro is nice in this case, so I'd be inclined to do it that way. But Audition looks super powerful, I'd like to learn about it later on when I really need it.

                           

                          With the denoiser I watched this video:

                          http://www.youtube.com/watch?v=WxvUANeFrj8

                          And the guy nails all the noise with about -10db reduction and 1.7db offset.  I'm not sure what offset is but I had to crank that up to 10db to really get rid of the noise.  Not sure what other damage I was doing to things, but it sounded okay to me.

                           

                          To a newbie one difference I see is Audition modifies the sound file, so the waveform looks equalized or levelized or whatever the term is.  Where the Pro way the waveform stays unequal looking, but it sounds fixed. Not sure if this matters, but it was a difference.

                           

                          I amazed you can fix audio like this without ever needing to mark each speaker transition in some way, pretty cool.

                          • 10. Re: How to efficiently separate audio track
                            josephs51576386 Level 3

                            That is because the waveform it shows you inside Premiere pro isn't updated to reflect the changes that the dynamics effect causes to the file so basically the waveform just isn't updated when using the dynamics effect. It won't matter unless you just want to look at the waveform after it's been updated, instead of just looking at the audio meter. However I do want to mention that audition is indeed a very powerful and necessary program. Because for a lot of audio tasks Premiere just doesn't cut it, so in those cases it's best to use audition. For the task you're currently wanting to perform though Premiere is more than enough in my personal opinion.

                             

                            Also in my own personal experince if you're using the denioser on simple dialog with just people talking and no music or anything it doesn't ever cause any notable issues or damage to the sound.  If you use it on music though it can sound really strange and really change the the overall sound of the music in some cases.

                            • 11. Re: How to efficiently separate audio track
                              Allynn Wilkinson Level 2

                              It's also true that Audition (like Premiere) can do non-destructive audio editing. 

                               

                              If you use the rack in the Multitrack of Audition, any changes you make are non-destructive (i.e. they don't effect the original files and I don't think they change the waveform).  If you are working in the Waveform Editor changes can be either destructive of non-destructive (which is confusing!).  You can't do things like adaptive noise restoration in non-destructive mode.  Premiere makes backups of all the audio and sends those to Audition.  I'f I know I'll be doing some serious "heavy lifting" as I am with the current project, I tend to make backups of the backups and work from those.

                               

                              Coming from FCP 7 (where the audio tools are naf) I naturally took to doing my final audio editing in Audition.  Heck, I used to port aiffs from FCP to Audition on a Windows box just to fix the audio!  In waveform mode you can do crazy things like "heal" audio to remove cell phone rings (I'm not kidding - it works!).  It is a massive program to learn but well worth it since good (or at least non-annoying) audio is vital to good video

                              • 12. Re: How to efficiently separate audio track
                                josephs51576386 Level 3

                                The waveform updates inside audition after you apply the effect. Once it's removed from the effects rack the waveform updates (since you have clicked apply). In Premiere it doesn't update your waveform, even when you render the audio with the dynamics effect applied. That is what I was referring to, I wasn't meaning it was destrutive, sorry for not explaining it clearly.

                                 

                                As far as I know the only time I've ever seen Premiere update the waveform for any effect is if you use the audio gain feature via the right click option inside the timeline.