Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
Sound mixing, and matching tracks recorded in different ways
  • 37 Replies sorted by
  • I tried spectral denoise and voice denoise worked better for my clip. The noise/wind was all over the dialogue in every frequency and db range. I tried many combinations and my combo sounded the best with the least distortion. I hope this answers your questions.

  • I can't say I've done a great job with the soundtrack, and we're near deadline, but I have one last question.

    I'd like to keep most of the vocal levels within the same decibel range. I obviously can't simply raise or lower the entire track. But is there a way to boost the low levels, and keep the louder levels more or less the same, and have everything within a certain decibel range? Sort of like boosting the gamma in visual data? I think this is what compression's used for, but I'm not sure if I'm using PP's tools for that very effectively.

  • @hardimpact trying to follow your 10 steps guide, but i'm completly noob in sound. Any video tutorial of these or similar procedure i can play (and pause) for noobs?

  • audition would be a better tool for this. if your voices are widely dynamic even after a -23 RMS, then apply multiband compression till it sounds nice, then add bs.1770 last. the adobe youtube videos by 'videorevealed' are really, really good.

  • Thanks a lot @hardimpact !! now viewing his video tutorial "The easiest audio editing for non-engineers in Adobe Audition"

  • Although more academic than immediately practical, this channel is good for learning about various audio engineering concepts in a succinct manner https://www.youtube.com/channel/UC9KI12liJIeXLpcNqDN4QNQ/videos

  • My 3 cents: I would not use Audition for anything, you should upload some samples for people to look at, have them take a whack at it and then pick the one you like the best.
    NB: you can definitely "boost the whole track", not sure why you couldn't, and if you want to boost the good bits and not the bad bits, you do this with micro edits. Don't do this in Audition, but if you must, just keep splitting the audio, lower the chunks that have the bad bits and raise the chunks that have good bits. There's a million other ways to do this, including multi-band compression, noise pattern reduction, and so on, but if you aren't going "chunk by chunk" and instead filtering the whole enchilada your results will be worse. IMHO the typical YouTube videos on audio are mostly junk, but there are some good ones by Kraznet on Samplitude.
    The other thing to ask is why you aren't paying a skilled audio engineer for a few hours work.

  • I wish i could pay a skilled audio engineer (and a colorist too) but my work is almost no budget. We are two people band right now. However, thank yow guys for those advices.

  • I thought this would be useful for frugal filmmakers who don't own an audio editing studio or calibration equipment. . .Free DIY eq calibration and volume for audio editing.

    "Play a pink noise and apply graphic eq to just one channel(left or right, doesn't matter). Make all sliders zero except one one at time and write down the number where the sound matches the same apparent volume as the other speaker. Finally, play all freqs to set master gain against pink channel on other speaker because eq'ing changes volume."

    Example: If you compare a 20 band eq only one slider at a time, the pink noise of your left speaker will match the apparent volume of a single right speaker's freq. You can now model the freq response of poor headphones against a Sony MDL so they don't sound so tinny. Finally, with eq complete, you now can set comfortable talking level with pink noise -23db in os's volume control.

    this can also be applied to video players and even your web browser as a equalizer plugin. the professional way is sonarworks but its not free ;)

  • @hardimpact

    What is it about? Matching speakers to individual ears?

    Calibration is not about it.

  • The first part is balanced frequency response across the whole range with speakers/headphones that don't respond equally like pink noise. i.e. some speakers/headphones are stronger in bass or treble. I'm looking at you beats headphones :)

    As for loudness calibration, some people actually own a SPL meter but are recommended to lower its setting if in a small room. At that point, you might as well calibrate to theatre settings. The sonarworks also does loudness too, I forgot to mention that.