I agree. You shouldn't normalize at this stage, even the one from FL that is pretty "soft" on sonical changes compared to other hard normalizers. Especially if you dont control/limit ISPs. And I think even if you do, some streaming codecs will do nasty things to a 0db normalized export. You should control your final output during final mastering process and adjust levels for export at this point, not after. Not 100% sure it will cause issues (actually for streaming codecs, it's quite possible), but definitely not 100% sure it's a safe way. I would avoid "experiments" at these production stages.

(I also master on FL12 most of the time, if you need more ideas/tips)

 

I've read the info on mp3gain and it looks like it's based on Replay Gain algo. Basically it writes id-tag information that media players read and change the volume accordingly. So it's not a direct form of audio processing.

All in all, better not to mess up with lossy re-encoding, rather encode once targeted on the destination platform.

 

"MP3Gain does not just do peak normalization, as many normalizers do. Instead, it does some statistical analysis to determine how loud the file actually sounds to the human ear." this part is the same as replaygain.

"There is no quality lost in the change because the program adjusts the mp3 file directly,"

It directly changes the gain of individual MP3 frames, so the gain will be heard in all players irrgardless, because it's not completely a metadata thing. This part is not same as replaygain. The metadata it writes is in case the user wants to restore the original gain - because it changes the mp3 data.

 

Now that part is clear. It's probably a similar algorhythm that streaming companies use.

 

Steinberg has a free loudness meter and its ready for pro work.
https://www.steinberg.net/en/support/unsupported_products/slm_128.html

 

Last edited by a moderator: Jun 15, 2017

...and Melda


https://www.meldaproduction.com/MLoudnessAnalyzer

 

Streaming MAY use RMS, but I'm pretty sure they reencode everything.

 

They use LUFS for sure, including algos that do not take into account silent moments in the songs (it's threshold based) as the silence may look that the song is more quiet. All in all, it's pretty smart (and welcome) wha they are tryting to accomplish with LUFS normalization.

And obviously they need to reencode since they have to change the gain of the source file (and various bitrates in case of youtube).

 

rms is a variable prameter on socity habits .

 

The reason is called intersample peaking. Read this article: https://www.saintpid.se/en/isp-true-peak-limiters-test/



To avoid it, either leave your limiter lower (-0.3dB has always worked for me, I don't even know what sort of music you guys are doing that you need more), or better yet, use a proper limiter that catches intersample peaks. Look for option "ISP". The post also has a list of plugins that catch these peaks. Try AOM Invisible Limiter, Izotope Ozone or T-Racks Stealth Limiter. Also, use oversampling to help the limiter see the peaks.

 

Mp3 conversion adds up to 1.5dBFS of increased peaks depending on the source. Hence why you might need up to 1.5dBFS of headroom before conversion, to avoid clipping/distortion.

For Redbook CD audio (44.1/16) you might want to add 0.3dB of headroom to avoid clipping and weird behavior in some DA converters (in CD players), if anyone is still using CDs that is.

 

Last edited: Jun 18, 2017

Man you are mixing things up.. we are talking about mastering for conversion

Quote (I like copy & paste - saves time ;)):

Is there such a thing as Mastering for SoundCloud or YouTube?

I personally do not set out to master a track specifically "for SoundCloud" or "for YouTube". I follow the recommended practices for Mastered for iTunes which - in my view - apply to any lossy format (such as avoiding clipping and distortion, and keeping peaks below -1 to -1.5dBFS).
Rob Stewart

Is Rob Stewart a Mastered for iTunes certified mastering engineer?

Yes. I am on Apple's list of approved MFiT mastering engineers, which means that if you choose to have me master your project, I will follow Apple's published guidelines, as summarized in this article, which assures you that A) your release will be labelled as "Mastered For iTunes" and B) more importantly, that the resulting AAC compressed files that are sold via iTunes will sound as close as possible to the original 24-bit masters.

source: justmastering.com

..or/and what Baxter already wrote.

 

It looks as though Waves WLM plug in is the ticket.

Thanks for all the info people! This music community is always full of answers! And I love how diverse they are!

 

Last edited: Jun 19, 2017

I have one more question. SO I used Waves WLM Plus plug in and it took everything down to -23 LUFS. My peak TP level is now capping at around -24db......is this right? Or did I do something incorrect? Do radio stations and the like, want the volume level this low? I must have done something wrong here......hmmmm

Edit: I used the standard setting of EBU R 128 LUFS 18

 

So what? Conversion doesn't fuck your sound if you don't act stupid and do a proper job in the limiting, which is what I'm talking about

 

Hi @Trevor Gordon
Seems there is something wrong. It is impossible to get -23 LUFS if your peaks are always under -24db.. Did you already solve the mystery? (let me know) If not, I can give you a quick explanation what you should do /w WLM. You make nice beats that deserve best quality possible..

 

Hey Matt,

I think I've figured it out. I have been doing some comparing with other club tracks sold on Traxsource and such - and they always seem to be around -9 lufs so I'm sticking with that and leaving 1 - 1.5 db headroom. I'm understanding the WLM tool a little more now...I wasn't using the reset button enough to check the LUFS.

 

Exactly.. you must go to the beginning of your music RESET everything and let the piece go through from the beginning to the end (not pushing reset in the middle) and readout the long term LUFS - that's the number the service will be looking at. And if you are above that it doesn't help (anymore) because they will normalize everything to their standard.

So check with the service and try to comply (yeah -9 is the upper limit for most of them this days ..and that -23, -24 is just a guide for the future). Your peaks will at some point get higher or lower (but that are not LU - loudness units) as I see you already figured out.

What is really important - when I uploaded some material to YT respecting their standard it really sounded better, clearer ..even louder ( so it's good to test). Always enjoying your production - hope to hear something new asap ;)