I use MP3Gain for some time now, I set it to 90 db and it has always worked well.
But I have a song setted at 90 db like the others and I have the feeling it sounds the same volume (but I can be wrong), but the waveform displayed in Traktor is completely different from the others:
On the left, a “normal” song (the waveform is thin but it’s logic as I set my songs at 90 db). On the right, the song that still has a huge waveform after MP3Gain. According to the VU meter, it also sounds a bit louder but this could be due to the fact that the song has a constant loud bassline… Anyway, MP3Gain confirm the 90db volume, so it shouldn’t sound louder nor look louder on the waveform, right ?
I also recorded a “set” containing little parts of different songs, and uploaded it on Soundcloud. I’ll let you guess where the weird song is:
Does anyone have a clue about this? I really want to understand that behaviour. I’m pretty new in DJ stuff and I want to understand what I do and learn good practices as of now.
If you only have a controller (no hardware mixer that has hardware VU meters) you can only trust your ears.
You might as well stop using mp3gain and just enable Traktor’s autogain and adjust the gain at same time as you do your Traktor prep.
I should have precised that I use MP3Gain in order to not use Traktor autogain as I don’t want to be used to that feature because it does not exist on CDJ or with vynils. And my controller (Numark Mixtrack Pro II) does not have knob for gains… I can still use the mouse but I really don’t want to.
So, MP3Gain was for me a solution for both Traktor and CDJ Djing.
Maybe the best thing I can do is to buy a little extra controller (like Akai LPD8) to have more knobs (or to buy a better controller of course..).
However, I’d like to understand the problem (how can two 90db songs have waveforms that different!?) because MP3Gain should work well and it does except in the case I described. Until now, I’ve been pretty satisfied by that solution
mp3gain doesn’t say exactly how it does it, but it’s not a peak normalizer. It’s doing some kind of average level normalization.
One song that’s much more compressed (dynamic range, not data) than another will wind up with a smaller-looking waveform just because the peaks aren’t as loud once the average levels are equalized.
You could get an BS.1770/EBU R-128 program loudness value using some software (Reaper and RX come to mind, though I know there are free options…Reaper’s demo is functionally unlimited, FWIW) and see if it did it “right”. But…the sound is more important than the visualization. So, if it sounds good, it’s probably right or at least close.
Well, you’re definitely right, it does not sound bad to me so I shouldn’t worry about visualization.
That said, I don’t feel very comfortable uploading a set with a waveform like the one I got with Soundcloud (cf. my first message). I mean, it just looks wrong, it’s like if I had fucked up something haha…
Yeah I know that’s very weird. As far as I remember, I never saw this before in Soundcloud so that’s why it looks wrong to me…
I put it Public so you can check if you’re interested. In my opinion it does not sound louder than the previous song nor the next one (there are 3 differents songs in the record). I mean, the bass are very powerful but overall I cannot say that it’s louder.
Perhaps there was a lot of compression on that track when it was recorded/mastered? That’s what it looks like to me. Anyway, the recording sounds fine, so don’t worry about it.
I almost don’t have any knowledge in mastering but I thought the same thing. This was an “unreleased remix” that Kerri Chandler recently shared on his Facebook page. So, maybe no one really mastered it because it was not supposed to be released.
But I’m surprised it’s the first time I see this. Some old tracks or tracks from independant artists should be bad mastered as well I guess..
It’s more likely someone who didn’t know what they were doing squashed the crap out of it and called it mastering.
Indie vs. big doesn’t really matter…just who did it and what they were going for. And while going old enough will find quieter masters, it’s very much possible they were better before the DIY era when every idiot with a webcam started putting out guides and a bunch of producers with compromised monitoring started thinking they could do it themselves.
Thank you everyone, that really helped me to have a clearer view about all of that stuff. This is probably just a song that was not mastered following usual standards. And, in the end, the most important is how it sounds!
slight_smile: