I am very new to audacity and this is my first go on ACX specifications.
I am working on a few files and normally I edit and have someone master them. However, I am running into more frequent few needs of “sample” files that need to be cleaned up for ACX last minute
I am working with the preset files from the narrator of
noise reduction dB 15- etc
sensitivity dB 3.62- etc
frequency smoothing Hz 600-etc
attack/delay time (sec.) 0.10-etc
I am trying all different things to get it to the specs for ACX but I keep running into issues. The original files come in at
Peak -2.9
Offset 0
RMS -20 (using Stats) https://forum.audacityteam.org/t/acx-check/37562/1
When I do the suggested configurations of
Effect> Amplify I am getting 1.1
Analyze> Contrast I get -24.5
and somehow the entire segment of Analyze> Contrast ends up being around -17
(and the recommendations suggested to another poster on >ACX check) of the Noise reduction along with the Effect>Normalizer at the suggest settings.
(not in this sample)I have ran several different chains so I am trying to make sure I am keeping up with the different edits and outcomes. I feel like I am seriously missing something…
I have done the Highpass filter/Low Passfilter
I included the original copy as well the edited one with
Neither of those is the “raw clip.” I can hear processing in there. We’d like to hear the real raw capture.
Or make one up.
Record a clean, 10 second mono clip. Hold your breath and don’t move for the first two seconds. Export WAV and post it. I’ll tell you what I think you should do in my opinion, and why, to meet ACX compliance.
Unfortunately, that is what I have to work with because that is what comes from the narrator using the settings I specified above (my original is what I get from him.) Hence my dilemma of trying to fit this to ACX and not being a Pro in the software, or if I should ultimately go for a different software ie protools or twisted (I can get a stack made for the specs on acx), I hate doing that because I love working in the files and being able to tweak things as needed.
In my opinion, you’re stuck even with advanced software. Once you burn processing like Noise Removal or Compression into a performance, that’s the end of the world. Further processing is just going to dig you a hole deeper and deeper.
We urge strongly that the performer export and save raw performances. That’s the “Camera Negative” of the sound world. Lock and key, armed guard. Without that you could get what you have. No show.
“Something went wrong. Send me the raw performance. I’ll process it here and send it out.”
I can hear noise pumping in that clip and there may be other errors. A too loud presentation is very unusual. Almost all common errors give you a performance that’s too quiet. So there’s something else evil going on. Too aggressive compression?
You’re beyond what I can do. Maybe the other forum elves have some ideas.
~~
ACX expects you to shoot the piece, do minor adjustments to it and out the door. They say so in their videos. As an experiment, I shot a voice test myself and I got ACX conformance with one sound filter and Normalize. We keep telling people that and I wanted to see if I could do it. We have other posters who can do that, too.
This is a screen capture from the third video in the ACX instructional series.
And the silence is suspiciously dead. I can’t shoot a voice clip with -80 noise and I really doubt you can, either. LF_rolloff_for_speech.xml.zip (326 Bytes)
If you go with Steve’s LF_Rolloff filter, it’s used with the equalizer “Length” slider all the way up. All that sub-audible rumble and trash that screws up other filters goes away.
Koz
