Please note: All new registrants to the Panopto Community Forum must be approved by a forum moderator or admin. As such, if you navigate to a feature that is members-only, you may receive an error page if your registration has not yet been approved. We apologize for any inconvenience and are approving new members as quickly as possible.
Audio Normalize
Caitlin McCabe
Administrator
Original post:
Jeff Cremeen, 4/27/2018 at 12:45 PM
It would be extremely useful to have an audio normalize function to raise the overall level of a recording once completed.
13
Comments
Ha ha - this has been on the roadmap since *forever* 😁
+1 for this! Looking forward to implamentation by the end of the year.
Is there a way for this to be enabled by default globally?
@Andy Denyer The feature is available globally, but I can't think of a situation where it should be applied globally. IMHO this feature should be used as an exception rather than a rule and that is how I am training my users with it. If we have a room or a building where every recording needs this to be applied, we need to address the microphone used there instead of the recordings being produced.
I know that we have a few users who would be upset if we applied this globally as they spend a ton of time making sure that their sessions sound how they want their students to hear the content.
I could see maybe a recorder option (think like a checkbox below the capture computer audio checkbox) where the normalization is done prior to the content being uploaded (that way it is reflected in the editor as well), but I don't think a site default for the current implementation is the answer here.
@Michael Espey I think you might be confusing "Normalization" with "Limiting". Normalising means running a program that looks for the loudest peak in the audio (eg, -12dB) and increasing the volume by that amount (so 12dB) so the new peak is at 0dB. So you increase the volume by exactly the correct amount to maximise volume without adding any distortion. Limiting on the other hand turns the volume down whenever it passes a certain threshold and squashes the sound. In our case, room recordings are generally of musical content to limiting the signal before it goes in to Panopto will have a negative effect on the content as it will reduce valuable dynamic range information. We need to keep the microphones at safe levels due to the range of instruments and content being recorded.
So I would say, I can't think of a situation where it SHOULDN'T be applied. If there is headroom available, why not turn it up to help our users?
Hi all
I was the one that advocated for normalisation as an option way back and it's good to see arrive finally but I can't help but think it's not quite as useful as it could be - certainly I'd be looking at an option to apply it globally or at least inheritable from a top level folder setting.
I don't know what algorithm they are using but to my mind I was thinking along the lines of ReplayGain where for any given recording a playback gain/cut would be applied commensurate with a typical average nominal speech level upon playback (for broadcast this is used to be typically -10dBFS) - the point here is that the player does the gain adjustment and it's not 'rendered' anywhere except the 'podcast' versions - that being the case then there's no reason why you couldn't apply this globally provided that there was a maximum amount of gain applied that would deem to be reasonable before you knew you were probably adding lots of noise. The aim here is to make all recordings 'sound' similar in level. In fact if you want to do this really properly then you need to be looking at loudness normalisation and not strictly level normalisation - TC Electronic | Loudness Explained
However, normalisation isn't the whole story since any peaks higher than 0dBFS after the gain would need to be limited but this is nothing that many audio/video conferencing apps aren't already doing in their apps and players.
Ultimately then the aim to to remove barriers to quality as simply and as non-destructively as possible and as such aim to be able to be used as a default for all content - ideally yes you need good quality mics setup with the right amount of gain set placed in an ideal position etc but it's difficult to point out to the lay person what the right amount of gain actually should be. At least normalising does provide a non-destructive, consistent and as such better listening experience to the end user and to that end I would support more work in this area...
TLDR - Do what YouTube etc already do - YouTube's loudness normalization algorithm - AI Mastering Blog