Profile cover photo
Profile photo
Paul Figgiani
Audio Post Production and Loudness Compliance Support
Audio Post Production and Loudness Compliance Support
About
Posts

Post is pinned.Post has attachment
Noise Reduction

For Broadband Noise Reduction I use the Denoise module included in the stand alone version of iZotope's RX4 Advanced Audio Repair and Enhancement suite.

When working in a DAW, it is possible to insert an instance of their Dialogue Denoiser plugin into the signal chain. The plugin works really well when used to apply a second pass of moderate noise reduction, where  audio was initially processed with the Broadband Denoise module.

-paul.
Photo

Post has attachment
LKFS vs. LUFS

In Apple's recently disseminated Authoring/Loudness best practices, they make reference to "LKFS" (Loudness Units K-Weighted Referenced to Full Scale) - instead of LUFS.

https://help.apple.com/itc/podcastsbestpractices/#/itcd55a9646a

In the past LKFS "specs." lacked a relative gate. In essence threshold based gating prevents low level audio and/or silent passages from misinterpreting the foreground loudness measurement.

Note this differential no longer exists. LKFS and LUFS are synonymous.

[-paul.]
Photo

Post has attachment

Post has attachment
upTimer 3.0

I recently released a significant update to my upTimer utility for Mac. Use it to track Podcast, Broadcast, or Live Stream air and/or tracking time.

-paul.

Post has attachment
In-The-Box (Multiple Source) Tracking Session

New version of a discrete, In-The-Box tracking session: Host, Skype + Hijacked Source.

The session I/O Device consists of LoopBack Virtual Devices wrapped into an OSX Aggregate Device. The Hijacked Source (in this case Safari/YouTube) may be replaced with any in coming source, such as a discrete Line Input - fed thru an external audio interface. Mix-Minus feeds x2 (Host -> Skype + System -> Skype) are implemented in the DAW. All sources are recorded on discrete tracks.

The key here is the engineer has in-the-box control of signal routing and session levels without any dependency on a hardware mixer.

Notes: The Audio Hijack pass-thru chain is purposely simplistic, consisting of an application source module (Safari), a protection limiter, a virtual output, and level meters.

Downward Expanders are inserted in the DAW on the Host and Skype tracks. Their purpose is to attenuate (or mute) audible low level signal which may occur when a participant is inactive throughout the duration of the session. Further plugin deployment is of course possible and subjective.

You can add support/routing for external backup recording by way of additional sends/outputs feeding an interfaced solid state recorder.

[-paul.]
Photo

Post has attachment
Noise and Downward Expanders

The attached graphic includes three frequency analysis plots. At the far left the analyzer reflects high level (isolated) broadband noise. The most prominent aspect is a loud, persistent 120Hz hum. This noise is constant throughout active and inactive speech.

The middle frequency plot displays full Downward Expansion attenuation of the noise. The attenuation occurs during prolonged passages of inactive speech. In essence the combined signal level (noise + speech) has dropped below a defined Threshold thus initiating signal attenuation (Gate closed).

The frequency plot located at the far right displays active speech with it's signal level above the defined Threshold (Gate open). Notice how the inherent noise is once again persistent (and audible).

Please note a Downward Expander is incapable of distinguishing speech from noise. The process solely relies on combined signal level and it's relevance to a defined Threshold. Basicilly in this example passages of active speech will (for the most part) include audible noise.

Keep in mind the source signal level of inherent noise will determine the seamlessness of the transition between persistent noise and attenuation. If the level of the noise floor is very low, transitions will be less prominent and therefore acceptable to the listener. However if the signal level of the noise floor is elevated - transitions between audibility and attenuation will be jarring. This is certainly not recommended.

-paul.

Photo

Post has attachment

Post has attachment
Spoken Word Intelligibility Optimization

Can you interpret each of the 7 reference points?

Post has attachment
Dual Mix-Minus in Audio Hijack

On May 16th I documented how to record Multiple Skype Clients on a Single Host System using Pro Tools or Logic Pro X. The workflow requires running two independent (simultaneous) instances of Skype on a single machine. Each participant (+ Host) may be recorded in RT on a discrete track. Two DAW implemented Mix-Minus feeds make it all work with no dependency on an external mixer’s Aux Sends.

http://www.producenewmedia.com/recording-multiple-skype-clients-on-a-single-host-system/

A few people asked if it was possible to implement a similar workflow in Audio Hijack. It is. The attached image displays the necessary Audio Hijack configuration.

You will notice I use the same virtual (“Aggregate”) Audio Device designed for the Pro Tools/Logic implementation. Also notice I am not using Skype Application modules in Audio Hijack. Instead, Input Device modules are used for source audio assignments and Mix-Minus signal routing.

Please note concerns regarding problematic latency have been mentioned. The fact is monitoring the session through a supported Audio Interface’s headphone bus (via hardware) will eliminate latency. Of course any in the box software processing applied to the host mic signal will be absent from the monitored audio. If you prefer to monitor a processed signal - use external gear such as the dbx 286s or the Wheatstone/Vorsis M1. Additional processing may be applied in post.

-paul.
Photo

Post has attachment
Dual Mix-Minus in Audio Hijack

On May 16th I documented how to record Multiple Skype Clients on a Single Host System using Pro Tools or Logic Pro X. The workflow requires running two independent (simultaneous) instances of Skype on a single machine. Each participant (+ Host) may be recorded in RT on a discrete track. Two DAW implemented Mix-Minus feeds make it all work with no dependency on an external mixer’s Aux Sends.

http://www.producenewmedia.com/recording-multiple-skype-clients-on-a-single-host-system/

A few people asked if it was possible to implement a similar workflow in Audio Hijack. It is. The attached image displays the necessary Audio Hijack configuration.

You will notice I use the same virtual (“Aggregate”) Audio Device designed for the Pro Tools/Logic implementation. Also notice I am not using Skype Application modules in Audio Hijack. Instead, Input Device modules are used for source audio assignments and Mix-Minus signal routing.

Please note concerns regarding problematic latency have been mentioned. The fact is monitoring the session through a supported Audio Interface’s headphone bus (via hardware) will eliminate latency. Of course any in the box software processing applied to the host mic signal will be absent from the monitored audio. If you prefer to monitor a processed signal - use external gear such as the dbx 286s or the Wheatstone/Vorsis M1. Additional processing may be applied in post.

-paul.
Photo
Wait while more posts are being loaded