I am going nuts on something that I've spent too many hours on now. Sometimes audio levels drop to about half on export or even playing in PP on rendered files. I can't figure out why as if I scrub back and try again, many times it plays fine.
I have 11 Layers of HD footage with clips linked to after effects etc. and 6 layers of audio. Everything is rendered and exporting gives me the same wierd audio level drops. It always happens between clips (in the beginning of the next clip) but not always. It drops to about half for a few seconds and always to the same level.
I should say that have exported this very project several times during making it without changing problem areas "audio" but as it builds up I am getting problems - is it too heavy to handle or what it going on? This is not the first time I have encountered this problem. Problem areas have only 2 layers of audio tracks playing at the same time.
Exporting audio without video gives same result. Anybody have any suggestions?
MAC, i5 12GB RAM, CS5
Can you post a screen-cap of your Timeline, highlighting the points where the Audio drops? Might be something obvious there.
Also, can you give the specs. of those Audio files?
Included a screen dump. It happens in either one of these clips at the start. Like I say - when I experience this during playback in the timeline, scrubbing back and listening again will "fix it". No really, but then I know that there is nothing wrong with the sound. It has been through Sound Booth as well. Plays beautifully. 48khz - nothing out of the ordinary. The reason for the many layers are some sci-fi balls with transparency and "words" being spelled out inside of them. Later in the timeline I have another excuse :-) Even reducing this to one .mov through AE as a combined export doesn't change anything. Keeps happening. Last change was a color adjustment through SA in AE. Exporting without video - same problem. Playing from start in the timeline recreates the problem. Playing again, no problem. Exporting in any size - same problem.
This thing that leaves me wondering is that it drops to about half. It doesn't cut completely off. But there is simply no way to export this for me right now. I'm stuck with a project inside Premiere :-) Or no sound :-)
Thank you for the screen-cap.
Does Enya drop down in Level too, or just the narration WAV files?
Also, is there any chance that you have some Track Keyframes in that Audio Track, though you should hear those, in playback.
When you do playback with the Audio Mixer open, do you see anything out of the ordinary, with the sliders, or the Levels meters?
Thanks for suggestions - it makes me look other places than I would focus.
Something does happen - sound drops below -30db and only on those WAV files and for about 4 sec. There are no keyframes anywhere, but I did have a DeNoiser set on the sound clips. I have removed that now to see what happens. Playing the clip individually (as in the source window) plays perfectly. Playing on the timeline, however, gives a problem.
It does seem to happen on the exact same places each time. Sound files are in this respect identical and all had the same DeNoiser (preset) active. Also the sound is kind of "distorted". Now - I tried to render and replace files and that resulted in the effect being "burned" into the files so that not even playing them as source files worked anymore. Something is really wrong. The DeNoiser filter have been on the whole time - also in the earlier exports that worked, but I suspect that this might be involved somehow.
I will let you know when it is out.
No, all is as it should be, and additionally indicates that you do not have any Track Keyframes set.
This one really has me baffled. Hope that others have some useful ideas. Thanks for the info and screen-caps.
[Edit] On, and Enya's Level's are fine in that area, correct?
Maybe not very useful info here, but an advice...
Being a native user of æ, ø and å when it comes to talking and writing, I have learned long time ago that using such (æ. ø, å) in file names (or in anything purely computer related) is something that causes hick-ups now and then.
PS! "Half" when it comes to sound is -3 or -6 dB (dependent on sound power or sound level)
Yes - Enya is fine. Thanks.
Audio is not my speciality - and "half" was just what it felt like to me, dropping from around -12 db, so you are absolutely right Dag. -6 Must be half :-) It is much more than half.
Dag: Though I agree with you on the native letters in general - I don't believe this has got anything to do with it in this case. I still have a feeling that the DeNoiser might cause an issue though it doesn't make any sense to me right now. It is, however, the only thing that make these audio files different from the rest. This is the best indicator right now. Why it didn't give me any problems previously is where things start to be spooky, to say the least.
Of course, this is purely guessing until proven otherwise.
OK, thanks. So the other Audio Track is fine, just this one?
As a test, check that you have clean, unused Audio Track of the proper channel-count (even adding one if necessary), place the CTI at the beginning of that series of Clips, and then lasso them. Move them to the blank Audio Track, with Snap ON, and test.
Sorry that nothing is just jumping off the screen for me.
I still have a feeling that the DeNoiser might cause an issue
I would agree. Remove the effect and clean up the clips in Soundbooth.
I'll triple up on that. The DeNoiser effect works (near as I can tell) by having some "run in" audio to work with; basically, it's analyzing on the fly, trying to establish a noise floor. When you have silence before the clip, the effect thinks that's what the level of the following clip should be, so it clamps down the levels in an effort to crush what it construes as noise. A lot of times, the effect will work just fine if you start playback somewhere after the beginning of the clip, but anywhere before, and you get the unwanted attenuation.
That's been my rather unsuccessful experience with that particular effect...
Thank all of you for participating. I would not have solved this without your suggestions - Bill led me on the track and took me in a direction of the audio files themselves where I was focused on system capability and timing....and Colin put the nail on it. It was in reality a combination.
It happens to be the DeNoiser filter. That it calculates on the fly makes sense when looking at my time table - the problem appeared after the project reached a certain complexity and my system couldn't handle "on the fly" anymore. It also makes sense that the problem almost (but not 100%) always happened in the same spots, which confused me and also explaines whey I could scrub back and listen again - without the problem! Moving a layer ever so slightly could also change the result.
I also backtracked and checked my ealier projects where I had the problem. Those places the DeNoiser filter was active! So that determines it. My project now exports with no glitches and I am a happy guy.
That filter should come with a warning (or maybe I didn't read the manual) as it is completely unpredictable. Maybe on a very simple level, it could be trusted - but I will never use it again. Too bad because it is a time saver to just drop a preset onto (in my case) 16 identical files.
However, spending 30 hours struggling with this it was definitely not worth it at the end. Combining files, rendering for hours etc.
Thank you for your assistance and let this be a reminder to use the filter with caution.
Good thinking Colin!
For any noise removal, I always head directly to Audition, sample that "noise floor," and then process the whole stream. I can see where the ramped Volume could well throw an "auto" Effect off.
Way to go!!!!! I think that you nailed it.
Thanks for the question, and all of the discussion that followed. I have something new in my "toolkit," and can now help out better, thanks to you and to Colin.
If you do much of this sort of work, and do not have Adobe Audition (soon to be ported for the Mac too), I highly recommend it. General noise reduction is really quite easy, and beyond the "simple stuff," it offers a fabulous array of tools. Though I am anything BUT an Audio Guy, I could not live without it. Have not seen SoundBooth, since beta, so it might have added many of the features of Audition, but the latter is my go-to program.
PS - I am going to add a link to this thread, for my article on "Noise Removal," as it will likely help others - again, a big thank you!
I started out an audio guy by trade, and it's been my experience that "auto" denoiser filters never work well. I started out using ProTools, but sold it along with my studio several years ago when I moved exclusively to video and web production. I've not used Audition, though I've heard and seen good things about it. I'm very unimpressed with the capability of SoundBooth, however. My go-to program now, for doing audio work of any kind, is a free open-source program called Audacity. It has a very good denoiser filter (called "Noise Removal" IIRC) and would probably serve you well if you can't afford Audition...
I wonder why, with everything else that's included in the Master Collection, they don't include Audition too?