I have a live music clip with decent clear sound but recorded from closer to the drum kit than the other instruments. The cymbals are too loud (but without distortion) and I’d like to get them more in balance with the rest of the instruments. Preferably using free Windows software ‘cos I’m a cheapskate.
Anyone here done this and can explain how?
ATM: How To Save Audio From An MP4 Clip, Process The Audio And Replace It In The Clip.
by 29 Comments
I don’t have a definitive answer to your specific question, but from when I have been messing about with audio / video files (albeit on a Mac) it always seemed that FFMPEG was the free tool that could do pretty much everything.
The only realistic way is through careful EQ. If you start attenuating the frequencies above about 20khz, you’ll see some results as that’s where most cymbal energy is expended, but be careful as a lot of the life of a recording exists in the top end and rolling too much off will throw some baby out with bathwater and make it very sludgy and dull. Basically, don’t expect wonders without multiple tracks of audio to play with: the best you’re going to get will be a compromise.
@fentonsteve
I’d change your password, you’ve been hacked.
I was going to say “but nobody can hear above 20kHz so you can filter it all off and it will sound like the best recording ever, The Nightfly”. But then I stopped myself and made another cup of tea instead.
Ha! Yeah but any eq you do will have a reasonable curve around the centre point anyway, unless you ramp the Q to a narrow spike. If you start there, you’ll probably cut quite a bit of top cymbal crash. Dullness abounds.
Also you spelled “Rumours” wrong.
We’ve all been told recently that The Nightfly is the pinnacle of recording technology, didn’t you get the memo?
I got a copy of that memo. It’s a pity it was printed on Izal as that seriously impacted upon my full enjoyment of it.
And there goes today’s tea…
Multiband compressor is your friend.
There is software available called Ripx. This uses AI to separate songs into their component instrumental tracks. It is not perfect but enormous fun! I find it very useful for working out and practicing guitar parts.This would seem your best bet for what you want to achieve. It does not fit your brief in terms of being free. However, there is a trial version available. https://hitnmix.com/
Audacity is free (and great) audio editing software, though a bit heavy-duty. As with most audio manipulation, to avoid unwanted artefacts, adjust until you can hear it… then turn it back a notch.
To read in the mp4 file, Audacity needs the FFmpeg plugin installed (simple ‘how to’ is on the Audacity support site).
Replacing the audio track in a video file is not my area of expertise, though.
Unwanted artefacts…? “Parrrrp!” “Somebody open a window….”….etc
Audacity allows for comprehensive EQ (look under the Effect menu for Filter Curve or Graphic EQ) but a rule of thumb is ‘the rule of 3’.
More than 3dB of EQ is obvious. 3dB of EQ across 1 Octive is acceptable. 1dB of EQ across 3 octaves is acceptable. Notch filters of more than 3dB are acceptable but they should be less than an octave wide. EQ boosts are more obvious than EQ cuts.
Think of the audio spectrum as a balloon – if you squash one bit, another becomes highlighted.
I think of the audio spectrum as an isolated piece of space-time fabric…
…but sometimes it’s more like soup: lots of wee bits floating in a viscous medium…
…I need a lie-down, really.
Have you thought about contacting Peter Jackson? He’s apparently quite useful in this sort of area although your three-minute video may come back eight hours long
Arf!
I was going to say that the person who would advise on the ATM if he could is Lou Reed.
I use a free video editing tool called Shotcut which is quite good at allowing you to process the audio separately – you can even process it outside the package and reimport it very easily. It has all the plugins you might need – EQ, compression, balance, fade in/out etc. I’ve used it quite successfully after trying a number of others which were either completely unintuitive (looking at you, iMovie) or too basic (come on down MovieMaker) plus various others.
I was gonna mention Audacity because it has nice graphics and bar charts an’ everything, but I only use it for actual edits, and the occasional fade in/out. Good job I didn’t because there’s some interesting stuff on it from fentonsteve above.
I didn’t realise there’s an EQ filter on it. Now at last I can remaster the Nightfly FLAC’s to make them sound like a human was present at the recording sessions.
Here’s a couple of half hour exercises in Audacity, (although they took me much longer to make) ; Thirty Minutes Beyond The Blue Horizon, and Thirty Minutes To Slumberland. They’re not playlists, but heavily edited “soundscapes” (if you will) using too many sources to remember – some may be familiar. They’re seamless, *immersive*, and a bit trippy. Best listened to on ear goggles to get the detail. It’s a workupload link, no monetising, no ads, no captcha codes, no sneakiness. Go on, force yourselves.
https://workupload.com/file/p9GYyKWGwAB
¯\_(ツ)_/¯
Downloaded. Better be good!
Thank you, Ainsley, for exceeding the number of comments I expected to get.
As a deal-breaker (I hope in a positive way) for those unwilling to make the physical effort involved in a few mouse clicks, here’s a little screed:
Thirty Minutes To Slumberland is themed around dreaming, as the exquisite cover design may suggest to you. The music is a mostly suggestive of the dream state, mostly peaceful, and if you nod out to it that’s a good thing – although there’s a lot of granular editing, overdubbing, and effects happening to keep it a long way from health spa muzak. Features the Cocteau Twins, and a supernaturally alchemical melding of Clarence Clemons (from his universally ignored and meditative solo album “Peacemaker”) and Paul Horn (from “Inside The Great Pyramid”).
Thirty Minutes Beyond The Blue Horizon was inspired by reading James Hilton’s Lost Horizon, and the first movie adaptation, dialogue snippets from which drift in and out. It’s themed around the search for a mythical or actual refuge away from the stressful world in which in we live in. “Somewhere Over The Rainbow” (here represented by IZ’s gorgeously simple version, and fractional samples from “The Wizard Of Oz” movie) is the same song as “Beyond The Blue Horizon” (Mike Nesmith’s version) and the longed-for destination has many names – Mu, Shangri-La, Xanadu. Like Slumberland, this ends with a blissed-out finale featuring monks chanting, Polynesian love songs, and sound samples building the mood.
They’re both very carefully assembled sound collages, with no one song source exceeding a couple of minutes, often being edited to a few seconds. They are NOT playlists, neither are they remixes. They’re mood pieces to pass half an hour (the ideal length of a dedicated listen for me) in pleasant reverie.
If there’s no Megan Thee Stallion on it, I’m out.
You can hear her snorting and her hooves a-clippin’ and a-cloppin’ on Slumberland, so you’re in.
Yebbut, is it as great as ’30 Seconds Before The Calico Wall!’?
Hint everyone: these are great.
So is the Calico Wall collection, but that’s a non-relaxing mind-fuck noisefest!
…and , yes, they ARE good.
Well, thank you very much. I know from experience how difficult it is giving stuff away to you cynical, depressed, suspicious-to-the-point-of-paranoia miserable bastards, but I like to think it’s worth the effort. I have other Thirty Minute exercises in Audaciousness, but these are probably the most *choke* “Afterword friendly”, and anyway I haven’t had my due praise for these two yet.
NOTE TO GARY: THESE ARE NOT FOR YOU. TAKE A STEP BACK THERE, FELLA!
While you’re here, you might be interested (heavy emphasis on that “might”) that the process used to create (okay – “do”) the covers is essentially the same. Using pretty basic technology (no Photoshop here), images are snipped from the internet and placed on a background from the same place. They’re sized, treated by filters and distortion and such, but that’s essentially it. Scrapbook technology. Which is what is used for the music – snippets of music and sound, all sourced from my friendly internet, and collaged together with Audacity. I can thoroughly recommend Audacity as a kind of Lego Technic plaything. The simple effects (cross-fade being a good one) are easy to master, but as you get into it, you’ll discover new combinations and possibilities. And like Lego, if you get it wrong you can remove that piece – Audacity allows as many un-do’s and rethinks as you want.
If, like me, you’re an elderly person in the twilight years of your life, in the “sans everything but my record collection” stage, you might want to have a go at this. It’s harmless, keeps you off the streets, and if you come up with something you can try to share it – rots o’ ruck with that one!
Do you have a friend with a (newer) Akai MPC? They are releasing their ‘Stems’ application which does exactly what you’re asking for.