ACX Check says RMS too quiet, have tried everything!

Hello! I’ve tried everything before posting here, for days and nights … now at wit’s end. Grateful for your advice please. I’m using an Aston Origin mic, an M-Audio 192-4 interface, on a Lenovo Yoga. I have the interface gain at nearly 90 per cent, Audacity gain at 9, and my mouth a fist and thumb away from the mic. My booth is all acoustic foam or soft fabric lined.

I have always felt when recording on Audacity that it was ‘quiet’ in playback. When testing this and recording like for like on Reaper, it’s way way louder on Reaper. When I record externally via Source Connect, they always ask me to turn my interface gain down to circa 82 per cent because I’m too hot. They’re using Pro Tools as the DAW. Problem there then being that my back up track I’m running on Audacity then records the session really low, so I’m hoping they never need it! But now the problem is looming large directly for me … I have to record an audio book, and I don’t pass the ACX Check for RMS … my noise floor varies from -70 dB to -62 dB, so I’m a thumbs up there, but an RMS failure … I’m tried everything, from re-installing Audacity, reloading the interface driver, doing an Irish jig … but alas. Here’s a recording sample and screen shot. Thank you so very much for your time and thoughts!!

ACX Check screen shot.png

Follow the [u]Recommended Audiobook Mastering Process[/u]. The RMS normalization step will make your RMS levels perfect. That will also boost your peak & noise levels. The peaks are no problem. If they are too high the Limiter will bring them down (with almost no effect on the RMS level). But, noise is often an issue.

they always ask me to turn my interface gain down to circa 82 per cent because I’m too hot.

You can’t go by the knob setting. You have to go-by the “meters”. It’s only too hot if you are hitting 0dB and clipping. Otherwise, your recording levels are not that important because digital recording gives you tons of dynamic range and you are going to adjust after recording anyway. It IS important to get a good-strong signal into the mic (for a good signal-to-noise ratio) but the digital levels are not critical.

I applied the Audiobook Mastering Suite exactly as written and the result is a technically compliant sound file that passes and sounds exactly like you.

Screen Shot 2020-08-14 at 12.31.53 PM.png
I don’t hear any obvious sound, voice, or theater errors, so you could submit that.

It’s “produced by contented cows” not with them. They have an intimate hand (so to speak) in the process.

You’ve already been using ACX-Check, so you’re only missing Audiobook Mastering.

That’s the long version. This is the abbreviated version.

Screen Shot 2020-08-14 at 12.45.16 PM.png
It’s three tools and it is a suite, a harmonious grouping. Use the tools in order, don’t add any and don’t leave any out. They work well and clean up after each other.

You should know there is an oddity in Loudness Normalization. It says RMS on the left but the number is wrong. Switch to Perceived Loudness and back and then put the -20 number in.


THANK YOU so very much for all of that, Doug and Koz! I will now spend time working with all that you have advised … I think I thought that the file needed to be better on e.g. the RMS before mastering. I’ll try all this out and report back! Have a dandy weekend, Rachel

I thought that the file needed to be better on e.g. the RMS before mastering.

The mastering tools will take whatever you have and solve Peak and RMS (loudness). If you announce too quiet, you will fail Noise. If you’re too loud, you will fail overload distortion or clipping.

Nobody fails overload because home microphones are designed to be quiet. Quiet recording and failing noise makes you think your production problems are your fault. Overload and clipping immediately sounds harsh and bad and makes you want to send the microphone back.

The fuzzy rule is to announce so occasional blue wave peaks make it up to half-way (0.5) or the bouncing sound meters start to turn yellow.