• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

Auto Bleeping - does it exist?

W

wadio

Star Participant
I'm listening to a live radio talk show yesterday when the host let fly the expletive, "f***ing."

He's an experienced broadcaster and I was listening on the Internet so I assume it was bleeped on air but I'm wondering who would have bleeped it given the sparse staffing these days? If the host himself hit the button he would have lost the entire sentence and would probably have re-stated it.

So now I'm wondering if software exists to identify and remove the seven deadly words and their variants. Voice recognition is extremely good these days so it seems this might be possible.

Anyone know?
 
I've never heard of any software like that. I would guess that, if there is a delay, he would be the one to hit the dump button.
 
wadio said:
I'm listening to a live radio talk show yesterday when the host let fly the expletive, "f***ing."

He's an experienced broadcaster and I was listening on the Internet so I assume it was bleeped on air but I'm wondering who would have bleeped it given the sparse staffing these days? If the host himself hit the button he would have lost the entire sentence and would probably have re-stated it.

So now I'm wondering if software exists to identify and remove the seven deadly words and their variants. Voice recognition is extremely good these days so it seems this might be possible.

Anyone know?
Click to expand...

No, there isn't anything like that. As you mentioned, it could be the broadcast side had a delay/dump system to protect the licensee side, but they just let the online stream fly. No regulations about profanity on the public Internet.
 
With recent strides in voice-recognition, AI, etc. I'll bet the dump function could be automated. Someone should patent it. :cool:
 
wadio said:
No sarcasm necessary, thanks. BTW, sometimes small market = big unit price.
Click to expand...

And exactly the reason for sarcasm. For the sake of argument assume there are around 10,000 radio stations in the U.S. Of those stations, let's assume 10% might use a digital profanity delay system, so that's 1,000. Now let's assume that 60% are small market stations with equally small capital budgets. That means whoever that would be able to develop a REALLY good speech recognition system that translates every word, compares it with an internal database of words, processes what decision, then makes the call to cut bleep those words out? Existing profanity delays cost around $3,000.00. Assuming you could create the speech recognition software that can recognize random voices, associated word database and articifical intelligence to make decision to cut that audio? Yeah, I bet the price point would be probably more than the radio station value.
 
Further research shows there is a device you can buy for your TV. It is also called "The TV Guardian" and is available through Amazon for about $80. Google that for more info.

There are also more expensive units available.

If I had an extra dollar or two I'd try one out of curiosity...
 
wavo said:
Further research shows there is a device you can buy for your TV. It is also called "The TV Guardian" and is available through Amazon for about $80. Google that for more info.

There are also more expensive units available.

If I had an extra dollar or two I'd try one out of curiosity...
Click to expand...

No self respecting owner or engineer would trust something like that to determine what should be cut out of your audio.
Manufacturers haven't been successful in making speech recognition software work reliably for closed captioning on TV yet. Doing it for audio only that would be reliable enough to keep you out trouble with the Commission? Not going to happen.

The way it is now, should some bad language fly by, the first form of concession the lawyers make to the FCC is the Producer or Board Op gets fired once management found out. If the engineer or management was stupid enough to trust speech recognition software to protect themselves? They should be the ones fired.
 
Kelly A said:
No self respecting owner or engineer would trust something like that to determine what should be cut out of your audio.
Manufacturers haven't been successful in making speech recognition software work reliably for closed captioning on TV yet. Doing it for audio only that would be reliable enough to keep you out trouble with the Commission? Not going to happen.

The way it is now, should some bad language fly by, the first form of concession the lawyers make to the FCC is the Producer or Board Op gets fired once management found out. If the engineer or management was stupid enough to trust speech recognition software to protect themselves? They should be the ones fired.
Click to expand...

You're preachin' to the choir, Kelly! I don't think many engineers asked to demo the Enco device...

Our first priority is to protect the license.

But the question was asked if anyone tried to develop such a critter and it turns out there's a couple of companies who have taken a shot at it. I would love to have one to experiment with but not on the public airwaves.
 
Kelly A said:
No self respecting owner or engineer would trust something like that to determine what should be cut out of your audio.
Click to expand...

There are so many pitfalls here...

First, recognition is bad with accented speech, ranging from American regional accents to those who are ESL speakers. Big risk.

Second, there is a latency in interpreting speech, which can be longer than the time needed to perform the deletion. By the time a bad word is recognized, it may already have become RF.
 
"Television won't be able to hold on to any market it captures after the first six months. People will soon get tired of staring at a plywood box every night." -Darryl Zanuck, executive at 20th Century Fox, 1946

Voice recognition synthesized speech are growing better exponentially:

https://www.youtube.com/watch?v=D5VN56jQMWM&feature=youtu.be

If it's not good enough today it will be very soon.
 
wadio said:
"Television won't be able to hold on to any market it captures after the first six months. People will soon get tired of staring at a plywood box every night." -Darryl Zanuck, executive at 20th Century Fox, 1946

Voice recognition synthesized speech are growing better exponentially:

https://www.youtube.com/watch?v=D5VN56jQMWM&feature=youtu.be

If it's not good enough today it will be very soon.
Click to expand...

Given the inherent latency in interpreting "in context", it can't be fast enough for profanity control, even were the recognition to be better than it is today. As stated, no licensee will take that risk.

And, in general, speech recognition puts those who do not speak pure Midwestern English at a disadvantage. In my family there are a range of accents as most are ESL speakers, and recognition goes from fair (tedious to use) to awful (useless) whether it is contacting someone by phone or using the voice to text feature on a smartphone.
 
David, I don't know what voice recognition you're using but I find Google to be extremely accurate and it's getting better all the time as machine learning improves it. I'm told that Apple is not nearly as good but not having used it myself I can't say.

If a swear word isn't clear enough for a computer algorithm to flag it, is it clear enough to engender a complaint? Maybe, but stuff like that slips through now without each instance generating a fine.

As for latency, that's not an issue at all. Computers are very fast. If not fast enough the stream can be extended to give the software time to do it's job.

Unfortunately there are many things computers can do better and faster than humans. Look at the trading floor at the Stock Exchange. I predict that in a few years not only will profanity filters be better and more reliable than minimum wage board ops, I'll bet they'll me mandated by the FCC! That is, if anyone even cares about profanity by then.
 
The problem with automatic profanity recognition is the wide range of regional accents that the software must interpret without error.
Even the best voice recognition software can't do it with 100% accuracy.
 
You must log in or register to reply here.
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.
Back
Top Bottom