Skip to main content

Wasnt Shure where to post this, maybe in the daw section. Anyways, I'm listening too songs i have been recording and exporting over the last years and i find that the volume level is very different for each of the songs. Its kind of annoying. Is there a way to deal with and correct this without spending several hours?

In general i find it very challenging to keep the same persistent volume level for songs i have made a year ago vs last month. Is there some special tools that can help with this?

Thanks!

Topic Tags

Comments

Lofty Whitaker Fri, 07/29/2016 - 14:04

Can't think of a way to do this in a batch and I'm willing to bet you wouldn't really want to. Too much room for file errors. If you have two track masters in digital format you should be able to set up a template in Wavelab or Soundforge for example and process each song to the same volume level. Do you have any mastering software?

Earthless Fri, 07/29/2016 - 16:33

I just got Adobe Audition. Not much experience with it yet. The tracks are just home recordings and not any "commercial product".
It will take too much time to do this without a batch process. The recordings have been made in Ableton Live and i havent paid enough attention
to get the songs at the same level when exporting or to be honest im not really sure how to achieve it. Anyways, for future recordings i need to pay more attention to this when mixing/mastering the songs. A mastering template is what i need. Ill see if i can set that up in Audition.

dvdhawk Sun, 07/31/2016 - 20:23

You might want to look into WavePad. It will batch process files with the script of your making, in this case with a "Normalize" function. It will also ask you if you'd like the files to retain their file format (wav., aiff, mp3, whatever), or convert them to a single format of your choice. Bear in mind, just because all the tracks would then have the same peak volume, it's no guarantee they'll be perceived as having the same level of loudness. In fact it's pretty unlikely, especially if your recording techniques have changed that drastically over the years. But, I don't know what else you could do to get any better results with less investment, or less time and effort.

I would definitely recommend making a duplicate folder for the original recordings that remains unaltered incase you don't like having all your tracks maxed out volume-wise.

pcrecord Wed, 02/15/2017 - 06:21

free version of virtual dj does adjust volume while listening.
Although correcting the files would be a better idea for the future
By using a LUFS meter with any DAW you could make wave files ready to burn on CD or put online without having those irregular volumes,..
There is some free LUFS to download.. just do a search

Kapt.Krunch Fri, 02/17/2017 - 04:28

The main problem is the "dynamics" of different songs. For an example, load up several songs in your program, and compare at their waveforms.

If you have some that have a lot of spikes thinned out at top and bottom, with a more-narrow "thick" average waveform, you have a lot of dynamics....difference between loud and soft.

If you have some that the "main meat" of the waveform is really thick, with very little spikes at top and bottom, you less dynamics. It has likely been mixed/compressed/whatever to death.

If you have some that don't peak at all NEAR the peak of others (thick OR thin), then the level of those is way low. THOSE may be able to be amplified to the peak of others. But, that may also increase any noise that may be present.

So, anyway...what happens if you try to "batch process" a bunch of files where some are brickwalled-compressed-thick, and some have the main meat more in the middle, with more dynamics?

You'll set the "target level", and if any that are compressed to death are squashed BELOW that level...they'll just get EVEN LOUDER! The ones with the "thinner middle, more dynamics" will only be brought up to the very highest random peak in each of the two stereo tracks. The problem with THAT is if you have, say, the left track with even ONE INSTANTANEOUS PEAK that is higher than ANY peak on the right channel, it's going to bring up the HIGHEST (but still lower) RIGHT peak further, to match the target level. That just imbalanced the stereo image. AND, it still didn't get that more-dynamic file up enough.

IF all the tracks APPEAR to have "close" to the same dynamics (which is doubtful), whether they are lower volume or not, you COULD try a very light compression batch process, to try to smooth out all the peaks to be at a more consistent level, first. Then, you could try to batch process the overall levels, by some means. Even "normalizing", which isn't the best. What that will do is raise the ones that are lower, overall, in volume to the same peak levels as the louder-overall-volume ones, and may either slightly raise OR lower the louder ones.

Bottom line is, if you try to do "dynamic" files along with "compressed to the max" files, it won't work. The more dynamic ones will never reach the average loudness of the less-dynamic ones.

And, I haven't seen a convincing way of of both lowering one level while raising another, and then "loudness-matching" all, within one single batch process.

Good luck,

Kapt.Krunch

DonnyThompson Fri, 02/17/2017 - 08:05

dvdhawk, post: 440295, member: 36047 wrote: Bear in mind, just because all the tracks would then have the same peak volume, it's no guarantee they'll be perceived as having the same level of loudness.

Yup. And when you figure in the dynamics to each song, as Kaptain (@Kapt.Krunch ) mentioned, you're throwing even more variables - and potential problems -into the situation.

You didn't mention what it is you want to do with these songs - other than tell us that they aren't commercial quality recordings - and presumably, this means you have no intention of releasing them, so my thought is, if you aren't planning on doing anything with these, then you have a couple options:

1. If this is simply for your own listening enjoyment, and you don't want to take the time to adjust the songs individually, then use your volume control on your music player.

2. If it bothers you that much to have to constantly use your volume control, then take the time to manually level the volumes of all these songs. There is no batch command that will do this for you - and get you satisfactory results at the same time.
But you're gonna need some kind of base-line standard that you can match all the songs to, or you're going to be right back where you started. This base-line could be using one of the songs as your level to match the others to, or....

If you are thinking about a release of some sort for any of these, then this will require that you adhere to a certain standard of perceived volume level. As Marco ( pcrecord )mentioned, get yourself a LUFS meter. If your DAW program doesn't have LUFS metering built in, then get one. Most are very affordable, there may even still be a few freebies out there. Standards for streaming and broadcast have changed, and will require that you meet certain volume criteria. Do your research for streaming services such as YouTube, or music download services like iTunes. They each stipulate the LUFS levels that they accept. There's no way around you having to do these changes manually.
Then, before you do start changing the levels - do exactly what Dave ( dvdhawk ) mentioned, and make copies of your original files.

Then... start learning as much as you can about perceived volume levels, dynamic range, limiting ( and other mastering basics).
My personal recommendation is to check out Ian Shepard's website; there's a lot of great free info there.
http://productionadvice.co.uk/mastering-basics/

Good Luck.

pcrecord Fri, 02/17/2017 - 09:42

Kapt.Krunch, post: 447564, member: 27145 wrote: The main problem is the "dynamics" of different songs. For an example, load up several songs in your program, and compare at their waveforms.

If you have some that have a lot of spikes thinned out at top and bottom, with a more-narrow "thick" average waveform, you have a lot of dynamics....difference between loud and soft.

If you have some that the "main meat" of the waveform is really thick, with very little spikes at top and bottom, you less dynamics. It has likely been mixed/compressed/whatever to death.

If you have some that don't peak at all NEAR the peak of others (thick OR thin), then the level of those is way low. THOSE may be able to be amplified to the peak of others. But, that may also increase any noise that may be present.

So, anyway...what happens if you try to "batch process" a bunch of files where some are brickwalled-compressed-thick, and some have the main meat more in the middle, with more dynamics?

You'll set the "target level", and if any that are compressed to death are squashed BELOW that level...they'll just get EVEN LOUDER! The ones with the "thinner middle, more dynamics" will only be brought up to the very highest random peak in each of the two stereo tracks. The problem with THAT is if you have, say, the left track with even ONE INSTANTANEOUS PEAK that is higher than ANY peak on the right channel, it's going to bring up the HIGHEST (but still lower) RIGHT peak further, to match the target level. That just imbalanced the stereo image. AND, it still didn't get that more-dynamic file up enough.

IF all the tracks APPEAR to have "close" to the same dynamics (which is doubtful), whether they are lower volume or not, you COULD try a very light compression batch process, to try to smooth out all the peaks to be at a more consistent level, first. Then, you could try to batch process the overall levels, by some means. Even "normalizing", which isn't the best. What that will do is raise the ones that are lower, overall, in volume to the same peak levels as the louder-overall-volume ones, and may either slightly raise OR lower the louder ones.

Bottom line is, if you try to do "dynamic" files along with "compressed to the max" files, it won't work. The more dynamic ones will never reach the average loudness of the less-dynamic ones.

And, I haven't seen a convincing way of of both lowering one level while raising another, and then "loudness-matching" all, within one single batch process.

Good luck,

Kapt.Krunch

Your answer suggest you never used an LUFS meter