Inconsistent audio levels in podcasts

[RuralRon]

I've mentioned this before but it's been awhile, and this is the first time doing so on the forum I believe.

 

My preferred way to listen to podcasts is audio only via rss feeds I play in iTunes. I listen to almost every podcast FDR publishes and occasionally a video.

 

I have noticed a considerably large variation of audio levels in the podcasts published over the last year to 16 months. The differences are pretty easy to notice. "Normal" levels can be heard on any podcast where Stefan is the guest speaker (Peter Schiff for example). I consider the audio level of those podcasts to be normal because it is very very close to 4 or 5 other podcasts I listen to regularly.

 

Has anyone else noticed this? Why do you suppose the levels are so low on so many of the FDR podcasts?

[RuralRon]

Nobody wants to weigh in on this? Why not?

 

I was just listening to a podcast in the chat room (in the player that appears below the chat input window) and it was quite soft. If I don't plug my Mac into an external amplifier it's hard to hear over the window fan which is in another room.

 

I'm beginning to wonder why nobody has bothered to respond, even if it's just, "It sounds OK to me" or "What are you talking about?"

[shirgall]

Nah, you aren't alone, I've pointed out some of the ones that were really noticeable. I think we're all waiting to see how the new studio works out, as it may improve the mastering workflow.

[RuralRon]

Thank you for your response Shirgall. Sometimes I think I must type with invisible ink or something

 

It seems kindof strange actually, that whoever has the responsibility of putting these podcasts together (MMD I presume) wouldn't normalize the sound levels through Audacity or other sound processing software. Sound level normalization is a pretty standard thing to do even by amateur standards.  Could it be that its one more step that takes more time to do than it's worth? I really doubt that's the case, especially since I don't see this problem in podcasts with far smaller resources and audiences.

 

Another reason it seems odd, given how Stef is obviously so aware of how important it is to have a quality product and be in touch with his audience, that this issue has persisted so long and seems to be overlooked.

[shirgall]

 

It seems kindof strange actually, that whoever has the responsibility of putting these podcasts together (MMD I presume) wouldn't normalize the sound levels through Audacity or other sound processing software. Sound level normalization is a pretty standard thing to do even by amateur standards.  Could it be that its one more step that takes more time to do than it's worth? I really doubt that's the case, especially since I don't see this problem in podcasts with far smaller resources and audience.

 

Another reason it seems odd, given how Stef is obviously so aware of how important it is to have a quality product and be in touch with his audience, that this issue has persisted so long and seems to be overlooked.

 

The reality is that we don't know how the podcasts are recorded. My suspicion is that previously the entire Skype call was recorded as a single track, and as a result it is extremely difficult to separately level the various participants. The previous way to get around this is to have each participant send the audio of the call in separately, and then try to match it up. This requires some delicacy.

 

This is why I wasn't sure if the new studio also meant moving to multitrack recording, like this:

http://lublog.tuttoeniente.net/archives/399/record-multi-track-skype-group-call-on-os-x

 

I'm a Linux guy, so I mess around with Jack instead of Soundflower, but it's the same basic idea. And, of course, I'd prefer mumble to skype for multiparty calls.

[RuralRon]

The reality is that we don't know how the podcasts are recorded...it is extremely difficult to separately level the various participants.

 

I'm a Linux guy, ...

 

That is absolutely true, we (you & I) don't know what method & apparatus is used to make the recordings. However, I contend that is mostly irrelevant to the issue of the final audio level, as I know I can pass any MP3 through software that will allow me to set an absolute level or to normalize the overall level in that MP3 to a consistent level. Your statement, "it is extremely difficult to separately level the various participants" is incorrect, as that can be accomplished after final edit.

 

The technical term for how that is done is automatic gain control. BTW, I'm a linux guy too, as well as Mac & Windows; I use them all almost every day.

 

This would be applied after all editing and splicing is done so all elements to be normalized are present. It's a post processing step in the production process.

 

As I mentioned in my OP, this is not rocket science and the techniques to accomplish good and consistent audio levels are well known. That's why I am so puzzled about why this problem persists and why Micheal or James or anyone else on the staff hasn't responded to the feedback provided to FDR about this issue.

 

Knowing it's not just me (I was 99% confident of that but until you responded I remained open to some obscure problem on my end) and that I'm not the only one to bring this to the attention of FDR leaves me to conclude:

 

1) FDR is in denial this is a significant quality problem worthy of consideration.

2) FDR is aware of the problem but it is a very low priority concern.

3) There are technical issues related to the tools being used (during the of 12 - 16 months I have noticed this problem have the same tools been used or have they changed or evolved?) in FDR's production methodology that are frustrating for FDR to overcome, similar to the recent opening rant by Stef complaining about Bluetooth and USB technology problems.

 

I lean towards a combo of #2 and #3 above.

 

However, if I were to be more deeply analytical, I might apply RTR principles looking at the relationship between FDR and the FDR audience, and ask why is FDR not being more communicative to the audience about the "feelings" the audience has expressed about this issue? Is FDR trying to protect a bruised ego or admit incompetence b/c they can't resolve this problem?

 

OK now listen, I get that sounds rather "bitchy", condescending and rather negative, even to me. I'm struggling to communicate without that tone but in the absence of a response (yours is the first after several attempts) my emotions escalate.

 

So perhaps I can turn this discussion in a different direction and use it as a vehicle to explore my emotions and RTR in an online context, since I have noticed I tend towards this negative tone in my online communications.

 

Anyone game to address that instead?

 

RTR by definition isn't possible in a forum context like this, so scratch that thought.

Edited June 10, 2014 by RuralRon

[shirgall]

The technical term for how that is done is automatic gain control. BTW, I'm a linux guy too, as well as Mac & Windows; I use them all almost every day.

 

Yeah, Automatic Gain Control can indeed normalize the stream somewhat, but at the expense of dynamic range (and usually noise in the silent bits, unless you also gate). Do you want to lose the efficacy of a Stef rant by effectively compressing/limiting the audio?

For what it's worth I believe FDR is working to improve things, but they are likely competing on the priority list against new content.

[RuralRon]

Yeah, Automatic Gain Control can indeed normalize the stream somewhat, but at the expense of dynamic range (and usually noise in the silent bits, unless you also gate). Do you want to lose the efficacy of a Stef rant by effectively compressing/limiting the audio?

For what it's worth I believe FDR is working to improve things, but they are likely competing on the priority list against new content.

 

I agree with your assessment in terms of the tradeoffs, but neither of us know whether any kind of normalization is being done right now. We can 2nd guess all day but that would be pointless.

 

What we do know empirically is:

 

1) There are many podcasters that don't have these type of level issues. Do they have the same dynamic range requirements Stef imposes due to his speaking style? I've heard some pretty wild rants elsewhere, but I can't say objectively that's not a relevant factor.

 

2) FDR chooses not to acknowledge the issue, publicly or privately (at least not to me, have they to you shirgall?).

 

I guess I've spoken my piece about this. Since you don't seem to have an interest in moving the discussion away from technical considerations and we both generally agree on that front, not sure if there's much else to say.

 

I do wonder why this sticks in my craw though. It's difficult to understand why. If I were able to afford therapy it would be a good topic to explore.

[shirgall]

Since you don't seem to have an interest in moving the discussion away from technical considerations and we both generally agree on that front, not sure if there's much else to say.

 

I do wonder why this sticks in my craw though. It's difficult to understand why. If I were able to afford therapy it would be a good topic to explore.

 

Well, as you said, at least I replied. But i have made my comments on other posts and I left it at that. I know plenty of podcasts where audio is better, but the audio isn't the message, unless it detracts from the message. Heck, I listen to the shows at 125% or 150%. I'm expecting some distortion.

 

We have no idea what they are up to, sure, but I merely was acknowledging that I was hearing what you were hearing.

[RuralRon]

And I thank you for doing so. Helps to have some interaction to affirm that I'm not alone in my observations. Thanks for your basic human kindness and interaction.

[Mishelle]

I get a bit frustrated because some podcasts, even the most recent ones, I have at full volume and have to strain to hear Stef, not the callers, but other podcasts come off just fine.  There's nothing I can do about this, speakers are good, other sites come off loud and clear and I have to actually turn them down.  Not sure if anything can be done about this, but my wouldn't that be nice!

[shirgall]

I have to say some of the recent podcasts have been very good, including improved lighting setup.