Hissy/Fuzzy audio fix?

Noiz2 said:
I know this is going to confuse you but maybe that is because you are supposed to be listening, not watching.
Click to expand...

I wanted to address this by itself. Of course for audio things, listening is important. Even with using simple filters, the effects of the filter may mangle the audio, and so, even with the 'noise' eliminated the results are so poor quality that one needs to give up, or find another way...

The point is there may be visual tools that allow some one to tune filters and processing more quickly and more acturately, rather than spending some amount of time with sliders. And of course, perhaps after years of doing so, may not need the 'crutch' of such visual aids... unless they loose portions of their hearing as they age... and then they might as well make silent films any way...
 
using visuals for sound is like
"listening" to video to color correct.

Too many things are going to be happening at a particular freq. Dialog, background noise,..etc.

All voices are different and would tell you nothing looking at it visually compared to
whats going on behind the dialog.

blah.etc...blah...
 
In the original poster's case... finding the 'noise floor', in terms of what I would call 'hiss'... 'room tone', 'background noise', or just 'noise', would allow the poster to adust something say, a 'noise gate' filter, again a threshold based filter to sort of 'chop' out unwanted audio signal... or perhaps get away with some form of band pass or the 'shelving' filters...
Click to expand...

Actually No. Filters are, well I won't say never used for noise removal. But are certainly not the best choice. It's a little like carving a turkey with a hatchet. Noise reduction apps like the one you mentioned (RX) are not using filters to get rid of the noise they are using a noise print to set a very complicated set of expander gates with filtered keys. RX's spectrograph is useful for finding freq specific "events" to use the spectral repair on but isn't much use for broadband noise removal.

I also completely disagree about the visual vocabulary. If I say I'm going to dip a freq. there is probably nobody over the age of six or so that would not understand that I was lowering it. Would anybody be confused about what a "notch" did? OK you seem confused by all such complicated "jargon" terms, but even my mother knew what a notch was.

Now a spectrograph is one big color painting. How do you describe that. Try to set your filter to filter out "the red stuff that's sort of in the center". I find it useful in only a few situations. One is the freq specific events IF you can surgically select on the spectrograph. Another is sometimes when you looking for freq. specific thing. Sometimes the spectrograph makes it easier to see them. That's about it. I would never use one to set an eq. It's a lot easier to look at something like a spectrum analyzer if you need a visual. But you can't, or shouldn't, mix with your eyes.
Cheers
SK
 
The point is there may be visual tools that allow some one to tune filters and processing more quickly and more acturately, rather than spending some amount of time with sliders. And of course, perhaps after years of doing so, may not need the 'crutch' of such visual aids...
Click to expand...

Oh that would be a crime. Imagine asking for experience in those you hire. I know personally I would never consider hiring a sound editor who had ever seen a mixer before. The first word of "jargon" they use and it's "out the door with you". Let something like that slide and your sliding down that slippery slope toward requiring... talent.


unless they loose portions of their hearing as they age...
Click to expand...

Now your talking. As long as my mixer is deaf I won't have to worry about their non-lack of experience.
 
Rippie said:
did you say something? I could'nt hear you over the genny
and the traffic ;)
Click to expand...

There isn't any traffic. I'm looking at my scope here and I don't see any traffic, your just making excuses again.


On a slightly more serious note. A friend was working at a major post facility on a big budget Hollywood film. She walks on to the mix stage and hears this buzz. So she goes to engineering and tells them about it. They pull up some screen and look at it and tell her that she's wrong there is no buzz. After a fair amount of hassle she gives up on trying to get one of them to come to the stage. They keep pointing to what ever box they had to "prove" there was no buzz. The director walks onto the stage and the first words out of his mouth are "what the hell is that buzz". Now engineering finally comes to the stage and guess what. There was a buzz, the mix is on down time and people are not smiling.
 
Noiz2 said:
Actually No. Filters are, well I won't say never used for noise removal. But are certainly not the best choice. It's a little like carving a turkey with a hatchet. Noise reduction apps like the one you mentioned (RX) are not using filters to get rid of the noise they are using a noise print to set a very complicated set of expander gates with filtered keys. RX's spectrograph is useful for finding freq specific "events" to use the spectral repair on but isn't much use for broadband noise removal.
SK
Click to expand...


Well... it is pretty obvious we speak different languages of signal processing... filtering is filtering... whether simplistic single pole or more complex with various thresholding, weight taps. correlators, whatever, to extract out 'desired signal' from 'unwanted signal'...
 
Noiz2 said:
Oh that would be a crime. Imagine asking for experience in those you hire. I know personally I would never consider hiring a sound editor who had ever seen a mixer before. The first word of "jargon" they use and it's "out the door with you". Let something like that slide and your sliding down that slippery slope toward requiring... talent.
Click to expand...

Were I in the position to hire a sound person, I would pass on someone who could not explain in reasonable signal processing terms what they do, especially if the 'job' is taking wild sound and there may just not be any recourse but to extract out audio from an abysmal setting. I may even find it with in myself to pay big bucks up front for such a person.

If they couldn't I would consider them, no matter how talented, to be knob tweedlers who have most likely been more lucky than talented, and after such knob mechinations, tell me their price has just gone up.

The original poster's setting was not a 'nice' sound stage with a set gang of workers, and actors, and the ability to take for ever to set up, or take multiple takes of the shot.

I'd look for some one who can get a fart out of buzz saw background, and explain how they could claim to do that.
 
j1clark@ucsd.edu said:
Well... it is pretty obvious we speak different languages of signal processing... filtering is filtering... whether simplistic single pole or more complex with various thresholding, weight taps. correlators, whatever, to extract out 'desired signal' from 'unwanted signal'...
Click to expand...

All I can say is that your "jargon" differs from that established signal processing. So I'm guessing you .EDU email is as a student. You are using the video /FCP / Avid terminology of anything that changes something is a filter. That is not the wider definition. Expanders are not "filters" in the general, or audio context.
 
Last edited:
j1clark@ucsd.edu said:
Were I in the position to hire a sound person, I would pass on someone who could not explain in reasonable signal processing terms what they do, especially if the 'job' is taking wild sound and there may just not be any recourse but to extract out audio from an abysmal setting. I may even find it with in myself to pay big bucks up front for such a person.

If they couldn't I would consider them, no matter how talented, to be knob tweedlers who have most likely been more lucky than talented, and after such knob mechinations, tell me their price has just gone up.

The original poster's setting was not a 'nice' sound stage with a set gang of workers, and actors, and the ability to take for ever to set up, or take multiple takes of the shot.

I'd look for some one who can get a fart out of buzz saw background, and explain how they could claim to do that.
Click to expand...
well thankfully you have nothing to do with the industry. I have far more background and theory than the big name sound designers/ mixers I have worked with. They are better than me at what they do but as a rule they could not "pull a fart out of a buzz saw" and if they could they certainly could not tell you (in the terms you want to hear) how they did it. They can however produce great tracks, win awards and make a living. Three things I'm guessing you don't.
 
Noiz2 said:
well thankfully you have nothing to do with the industry. I have far more background and theory than the big name sound designers/ mixers I have worked with. They are better than me at what they do but as a rule they could not "pull a fart out of a buzz saw" and if they could they certainly could not tell you (in the terms you want to hear) how they did it. They can however produce great tracks, win awards and make a living. Three things I'm guessing you don't.
Click to expand...

You are right that my income is not in the audio world... the most recent round with a signal processing was to detect small RF signal in the presence of various forms of distortion and noise. I'll tell you my method for achieving the goal... peak detection, and long term averaging with exponential decay performed in the frequency domain.

There were other methods proposed, and in some cases may have detected specific signals, but in general, the stupidly simple method was more reliable.

Quite a few years ago I learned the hard lesson of 'getting your name on patents'. If one creates 'noise' in the form of describing one's algorithm in obscure terms, invent 'jargon', a patent may be more likely issued, than if you simple state, in simple forms, what the process is. Since I didn't describe my contribution of The Device in obscure terms, I didn't get on the patent... the person who did describe the process in such terms did... then the lawyers between two companies began their process, on the one hand to show that The Device was based almost entirely on prior art... and the other hand, that The Device was a novel innovation.
Both companies eventually went under. Lawyers got paid though...
 
I used to work for some inventors and did a lot of patent searches (we happen to have one of the two repositories locally). Don't feel too bad only about 1 in 10 US patents survives a challenge. Part of your problem might have been that you can't patent an idea, only a device or process. "Simple" terms may have made it sound too general and not specific enough. You can appeal the offices decision, very few patent's make it through on the first pass.
 
So what happened to the guy who's wedding was ruined by the audio issue? did you ever save it? Did your client get upset and most importantly, did you still get paid? LOL
 
get a room.

get a room.

i love how the internet devolves every discussion into a ridiculous ego match on semantics and one-ups-manship.

can anyone just f*****g tell me how to reduce a massive nasty hiss or hum from my poorly recorded amateur audio so i can finish a lousy video?

other than using a 3 band equalizer, which isn't cutting the mustard for this one.

a bit of BG info - i was stuck shooting man on the street style stuff with nothing but a shotgun mic in a giant warehouse with about as much air condition hum, ambient shuffling, and muddy din as you could ask for.

can someone please give me a couple of ideas on what to do to minimize the BG din?
 
ambulantic said:
i love how the internet devolves every discussion into a ridiculous ego match on semantics and one-ups-manship.

can anyone just f*****g tell me how to reduce a massive nasty hiss or hum from my poorly recorded amateur audio so i can finish a lousy video?
Click to expand...

That's a brilliant way to ask for help.
 
You must log in or register to reply here.
Back
Top