typical noise floor?

[canadave]

Hi,

I'm finally starting to review our technical stuff at our small community radio station. I notice that when we have an open mic and silence in the studio, recording into Adobe Audition, we have a fluctuating noise floor at around -48 to -45 dbfs. I'm more a computer person than an audio broadcast engineer...does that range sound a little high? It seems a bit high to me, but I'm not sure.

Any suggestions on how to lower it? We've got the console going directly into an integrated motherboard sound card on our PC. Would buying a PCI sound card, even a cheap one, improve things noticeably?

 

[frankberry]

What is the noise level with all microphones closed?
This is the figure which will dictate whether you need to replace your sound card.
Adobe Audition is not the proper way to measure your noise floor.
How does the studio sound on-air?
If it sounds good, relax.

 

[canadave]

What is the noise level with all microphones closed?
This is the figure which will dictate whether you need to replace your sound card.
Adobe Audition is not the proper way to measure your noise floor.
How does the studio sound on-air?
If it sounds good, relax.

The noise level with the mics closed is around -57 to -51.

Although things usually sound okay on air, the noise floor has become an issue because we've introduced a Behringer mixer to get some extra mic capacity. There's just enough extra noise from the mixer, on top of the relatively high noise floor in general, that it can introduce an audible hiss. I just was curious if switching to a PCI sound card rather than integrated sound card would lower the noise floor enough to make a difference.

 

[frankberry]

A good USB audio card might be a better performer.

 

[bradgoehl]

One of the best, easy audio interfaces is the Henry Matchbox USB.
http://www.henryeng.com/USBmatII.html

 

[[email protected]]

Motherboard Integrated Sound device will give lousy S/N performance and -40dbm is common for that configuration. AS has been suggested by others, a PCI professional audio card or an outside USB audio box will give much better performance. The consumer type PCI audio cards are a bit wimpy on S/N specs.

 

[Goat Rodeo Cowboy]

What is the noise level with all microphones closed?

Adobe Audition is not the proper way to measure your noise floor.

For "community radio stations" and for people operating studios for voiceover, podcasting, etc, what is the preferred and yet practical (affordable) way to measure your noise floor? I always wonder about the accuracy of readings on Audition, but since a number of people use it, it does give us an apple-to-apples comparison.

CANADAVE: What catches my attention is the lack of difference between MICS OPEN and MICS OFF.

For those of us working with recording studios, there is a disconnect in trying to quote numbers. For on-the-air audio chains, ZERO LEVEL is pretty much that so maybe a noise level at -45 to -48 is acceptable. In recording we often set our chain to achieve peaks in the -9 to -15 range to leave some headroom. If program content FULL BORE is -12 for recording and noise is -45, then you may need to look for improvement. In this case you are not so much observing system noise.... but ROOM NOISE LEVEL.

AND if FULL BORE is -12 for recording and noise MIC CLOSED is -51 to -57, that is the number that would concern me. That would be noise internal to your audio chain.

A ratio I have picked up from the people who do church sound work is: Program Content (what is coming out of the speakers) has to be at least 25dB above the ambient noise of the room (Papers rustling, kids murmuring, feet shuffling and program content still reverberating around the room from a few milliseconds ago and maybe some traffic noise.) If the gap is 25 dB or less, listening becomes a bit unusable.

"Back in the day" my memory of broadcasting is that we were shooting for a ratio / difference between program content and noise content (measured at the transmitter output) was 50 dB or better... and back in the days to tubes you had to hustle to exceed that.

I remember walking into a station in Pine Bluff, AR one day back in the 50's and the owners (both engineers) were just beside themselves. They had stayed up all night tinkering and had uninstalled the transmitter power supply, put down some rubber padding, turned a large "bread pan" upside down on top of the padding, and then mounted the power supply on top of the pan. They had reduced their noise level by by 1.5 dB or something and you would think they had just circumnavigated the globe in a Piper Cub. And I think it got them down to -57 or something.

I've always assumed when we take the transmitter out of the equation... a free standing studio should exceed that.

But that leaves us with the comments and observation of frankberry. Measuring this things with Audition is not the same as measuring things with professional audio measurement equipment...... or is it? ;D

 

[TomT]

The accuracy of Audition probably can be questioned, but gives you a good idea of the relative difference in levels. It would be interesting to just record the noise floor to see what it sounds like--is it just hiss or is there hum or buzz which might indicate grounding problems.

Ideally, would be best to see if you can borrow a noise distortion test set, such as the Potomac Instruments sets, to get an absolute set of numbers.

However, that is somewhat academic, replacing the built in sound card with either a better grade internal card or outboard USB card or mixer would still be the best step.

 

[Goat Rodeo Cowboy]

I'm curious to know if some one with good test equipment and the ability to measure known standards has actually done a "shake down cruise" to see how trustworthy the later versions of Adobe Audition are when it comes to metering.

If I tell it to reduce the selected area by 10 dB, is that what it does, or does it change the level by 7.8 or 9.2 or 11 or maybe 12.3 dB... or does it actually do 10 dB when asked?

So I do something like you just described. I can see that in record with NOTHING going on in the studio I am getting a reading "jiggling" around between -56 and -58 dB. So I amplify the whole file up 48 db and now I have something in the -8 to -10 range and I can hear everything including that clock about six feet from the mic that runs off a AA battery and goes "tick tick tick tick" along with the other buzzes and belches in the room that our ears normally ignore... or are below the threshold of our ears. Are these numbers believable?

And what is the yardstick by which we set up our chain at a given set of gain. If I talk loud... doing radio production of some "hot" promo announcements, my gain will be set at one place and we call that ZERO. But if I am narrating and audio book and using a quiet conversational voice, backed off from the mic some distance to minimize lip-smacks, mouth clicks, and as one person described them recently... the bubbles bursting in the saliva on the end of your tongue, my gain will be cranked up much higher, making my noise appear to be much higher.

For finished product we want a noise level appropriate for the finished output... announcing a wrestling extravaganza.... or a Texas cowboy sitting under a mesquite talking about the last words his mother shared before she died.

 

[[email protected]]

Back in the day....studios were constructed for low noise within the room. Walls, floors, celings were thick, air ducts were large and low velocity, sound absorb material was on every surface....and walking into the room actually made yours ears feel weird.

Not anymore...noisy rooms are "fixed" with mic processors and noise gates....though the "fix" isn't nearly as effective as the treatment described above. And the room noise rides on top of the voice audio even with a noise gate, thus making the voice a bit dirty.

If you're doing voice over or commercial work from home, the cheapest and quietest place to set up your microphone is in your bedroom closet. Load it with clothes for sound absorbtion ;D

Don't forget to use a quality sound card in your PC too.

 

[Goat Rodeo Cowboy]

Back in the day....studios were constructed for low noise within the room. Walls, floors, celings were thick, air ducts were large and low velocity, sound absorb material was on every surface....and walking into the room actually made yours ears feel weird.

I was in the industry for 15 years. During that time I wanted to see everything there was to see, and learn everything I could. I visited the studios of 300 radio stations. (I sat down and made a list one day.) Since I got out of the business, I have visited another 300 radio stations.

Even "back in the day" very few stations had studios that were constructed for low noise within the room. I worked for stations that had room air conditioners and when we did those 15-minute newscasts "back in the day" you turned the darn thing off and sweated it out. The most impressive studio I ever visited was WREC in Memphis in the lower level of the Peabody Hotel circa 1960. In those 15 years I worked for ONE station that attempted to build to the kind of low noise standards you described.

 

[[email protected]]

Goat Rodeo Cowboy.....I must be spoiled, having worked in Chicago radio for 30+ years...prior to the year 2000 there were a lot of stations in the Windy City built to those specs ( and I built a few ) and a few built with floating floors and walls. On the other hand I've been at a few stations in Florida that are similar to the "window air conditioner in the studio" type you've described.

 

[Goat Rodeo Cowboy]

You and I worked at opposite ends of the "playing field" of this industry. My dream would have been to be the owner/operator of a county seat single-station market. A place where everything has to be done with gracious amounts of frugality... but a place where you could have personal interaction with sponsors, public officials and garden variety listeners. These were markets where on the best day of your life, the funding available (if you are really good at sales) only permits modest facility.

I've listened to a lot of Chicago radio through the years and from pictures I have seen, there was a lot of facility to be admired. The only Chicago station I was eve in was over at Moody Bible Institute. I attended a two-day seminar there for people involved in religious broadcasting so we weren't really focused on facility as such.

I have some other forums where I am going to post a question I will pose to you. There are meters for measuring sound levels. Cities pass ordinances that specify what the maximum signal level can be in the area surrounding a concert venue. (Amphitheaters present a real challenge when people at the concert want to rock, and people living nearby want to sleep.) I own a modestly price Sound Pressure Meter which can gauge that type of sound. But such meters are not designed to measure "available silence". (One of my early mentors in photography went beyond taking pictures with "available light". What we did had to be called photography with "available darkness".)

So in these well built studios that you have had the pleasure of working in and constructing, IF I had a suitable meter, what would be the level of the ambient noise inside a well designed and built recording or broadcasting studio?

My current plan as soon as I get a few spare moments, is this: Generate some sound that is loud enough for my meter to capture and measure. Do a recording while that sound is present in the studio. Then kill the sound. I realize the calibration of what I can measure in the recording with Adobe Audition may not laboratory grade measurements. but I will be able to measure the difference between the TEST SOUND and the room tone / standing noise. So, what number should be my "dream goal"?

 

[maineengineer]

Adobe Audition is actually a very useful tool in determining a studio noise floor, for broadcasting or recording purposes. The "old" FCC standard for FM broadcasting of 60 dB below (-60) normal level is still a good target for noise level. The corresponding FCC standard for AM broadcasting was 45 dB below (-45), but noise could easily be discernible on voice-only elements at that level. If you can't achieve the -60 dB level due to nearby computer contributors (fans, hard disks, etc.) which cannot be shut off, Adobe allows you to sample the steady-state noise sources and effectively remove them through software, at least for recording purposes.

 

[Goat Rodeo Cowboy]

If you can't achieve the -60 dB level due to nearby computer contributors (fans, hard disks, etc.) which cannot be shut off, Adobe allows you to sample the steady-state noise sources and effectively remove them through software, at least for recording purposes.

Yes, I've been using that technique for several years. Works great if you are doing broadcast style recordings where "Presenter" voice... a full throated projecting voice style gives you some advantage of overcoming the noise, and since broadcast bands also have some noise built in, who can hear that there is some "noise floor" in your recording.

This same technique works well if you are doing 'house of worship' recordings. Pastors have their picture in the dictionary to illustrate "presenter voice style". And people expect to hear an occasional cough, some movement of children, the dropping of hymnals and coloring books during a sermon.

Adobe 'Noise Reduction' polishes up those kinds of recordings real well.

Now, hold on! I'm going to "shift gears and pop-the-clutch' on you. You decide to do voice recordings... training narration of corporate videos... audio book narration... e-learning guides for the corporate and educational market. Sometimes the client says: "Ditch the radio-announcer presenter-voice'. You are narrating a tender, romantic scene from a novel which requires a very soft, slower paced voice. If your studio is a bit noisy and you tighten down-the-screws on Adobe "Noise Redudtion" you can get a recording full-of-worm-holes. I submitted an audition recently and the client pointed out to me the 'artifacts' which I knew came from being too aggressive with noise reduction. The client came back and said: I love your voice style for this task. Make me a new audition and get rid of the strange and obnoxious sounds and we can cut a deal.

So. Here is the question I have posed. Sound Pressure Level meters will measure louder sounds. (concerts, lectures, worship music, etc.) Is there a tool that measures sound levels down at the lowest noise level you will ever achieve in a studio setting. When do I know that talking louder, getting closer to the mic is all I have left to improve that ration with the ambient noise level. What meter will tell me: "Come on Dude! Work a little smarter. Spend a little more money. You should be able to squeeze the noise level down another: 4 dB? 6 dB 8dB?" Or the meter may tell me: "What's your problem. You are already 3 db better that we tell people it is possible to achieve with rational methods."

One more question. When calculating the signal-to-noise ration, is it common to measure the program content and the noise content at 'average peak voltage' or at 'average RMS power" or does it make any difference as long as you use the same standard on both?

I was up a 2 A.M. this morning so I could record with: The HVAC fan OFF, my little fan than brings downstairs cool air upstairs OFF, sound plugged into a very quiet laptop buried a long ways from he mic location so the computer in the room could be turn off, AND- the little recirculating pump that makes sure that hot water is instantly available at the distant sinks and tubs was turned off. That little pump that grinds water through the first floor SLAB adds between 2 and 3 dB to the ambient noise of the upstairs room!

As you can see... I'm am searching for tiny, tiny miracles. I just want to know how to calculate when there are probably no more miracles left in the pot that I bought one day in a yard sale at the end of the rainbow.

 

[maineengineer]

When you use the magic word "client," it is then presumed that you are charging money for your services, and thus representing yourself as a professional, in which case -60 dB would be the absolute worst-case-scenario noise floor. Adobe Audition typically measures peak levels, so that if you adjust your record input level to peak at 0 dB on your voice, then the background noise level (ambient studio noise, equipment noise, etc.) present when you are not talking (with microphone still "on") should be -60 dB or lower (or even -70 to -100 for agency work). Rather than try to put a Band-Aid on the situation through artificial noise reduction after the fact, you'd be well-advised to consider investing in changing your "studio" environment to reduce or eliminate the noise(s) at the source in the first place. It would also be advisable to consider taking an audio recording course which is typically offered at local colleges, to acquire more knowledge and related skill sets than can be offered here as a casual "Engineering Tech Tip."

 

[TomT]

Some noises will always be a problem, but sound proofing may help.

About 3 1/2 years ago we bought a station that had been LMA'ed by a group. Had to set up new studios in 2 months (they wanted us out of their complex). We picked up a split level that had a clear path to the transmitter for the STL. However, it was located along a busy road.

Studios went into the back bedroom, but the production studio had to be located in a front bedroom, with considerable road noise.

We used several different types of acoustic treatments, both the foam and fiberglass panels. This quieted the room down considerably.

The fiberglass panels can be hung on the wall like a picture. They come in several colors, are fairly attractive and are fire resistant:

http://www.bswusa.com/Acoustic-Treatment-Acoustics-First-AFR124-P64.aspx

The foam is more effective, but it isn't fire resistant and usually you have to glue it to the walls. Not recommended if your are renting. Also is a dust-catcher. Limited choice in colors besides the usual charcoal grey. We used tan foam and light green fiberglass panels in our control room; charcoal foam and dark red panels in the production studio.

Our production studio project cost around $1K, but that's cheaper than trying to rebuild a room from scratch.

 

[Goat Rodeo Cowboy]

Rather than try to put a Band-Aid on the situation through artificial noise reduction after the fact, you'd be well-advised to consider investing in changing your "studio" environment to reduce or eliminate the noise(s) at the source in the first place. It would also be advisable to consider taking an audio recording course which is typically offered at local colleges, to acquire more knowledge and related skill sets than can be offered here as a casual "Engineering Tech Tip."

I work very hard when asking questions to NOT create a "Geek-2-Geek" closed circuit. I go out of my way to ask questions in such a way that maybe answers will come back in a style the beginners and people with closed minds can understand what is going on, and EVERYONE can benefit from the conversation.

I have a habit in these forums. I just went back and read EVERYTHING you have ever posted in these forums. It helps me not to insult people when I respond to them.

I'll save you some reading. I made my first recording about 1949 or 1950. It was crude. On a "wire recorder". Passed the test and earned my FCC First Phone early January 1961. I've worked 15 different radio stations and set foot in about 600 others. Helped build from ground up or remodel four radio stations.

I own a modestly priced SPL meter but the manufacturer assumed everybody would be measuring the levels of rock concerts and today's contemporary gospel-babes-a-dancing church music. 40 dBA or 40dBC are it's lowest measurable ranges. To measure conversational level voice work and to measure ambient studio noise, it is useless.... without some ingenuity. ;D

I have a daughter who is an architect.

I have a consultant friend who helped design and install the sound systems in the U.S. capitol.

I have a consultant friend who at one time was in charge of all outdoor sound at Disney World.

Not at college level, but I have taught seminars on how to get good recordings in your church.

We can have productive and instructive conversation, or we can engage in typical sports-bar male behavior and see which one of us can out-arrogance the other.

Now, back to my question. If I can find a comparison technique OR a uniquely sensitive meter, what will be the measure of the ambient noise in a properly designed and constructed recording space. At what point do I say: "Aha! Trying to reduce the noise any lower than this will not be cost productive. If my signal-to-noise ratio at that point is not satisfactory, then I must focus on raising the level of the spoken content.

My effort at a commercial venture (which I guess indeed makes me a professional) is audio book narration. In a radio station production room, you just tell the announcer to get louder. After all, the hallmark of radio up until this point has been LOUD and LOUDER. But when you are narrating a scene in a book where a mother is whispering to a child to keep quiet lest the intruder hear them and know their location, getting LOUD and LOUDER like a radio announcer is NOT an option.

Then here comes the finger-poke-in-the-eye: You may be dealing with a publisher who has set what may be unrealistic standards for such ventures. They want the Average RMS level of the audio content up in the -20 to -22 range, When you introduced enought compression and limiting to meet that standard, the noise floor comes up like an express elevator in a high-rise office building to "bite you in the butt!"

By the way: I just found a claim in the Internet that "absolute zero".... the quietest space ever measured and certified to the satisfaction of the Guiness Book of Records is --09.4 dBA. (They didn't specify Fast or Slow. Maybe a STANDING level as opposed to program content with dynamics doesn't need a Fast or Slow designation.

So if -9 is as quiet as it gets, and doing a concert in the park where your sound level measures 109 will get you ticket by the police, I have a hard time visualizing a radio announcer or a book narrator sitting in at studio and getting his/her program content 100 dB above the ambient noise.

I have seen some indications via Google that maybe about 12dBA is as quiet as you will get a reasonable studio to measure. But I have no way to determine if the person(s) publishing information like this are qualified to evaluate and publish. Somewhere in this range you reach the point where the only way to get a lower measurement involves a structure with concrete walls 10 feet thick and other impractical criteria that exceed the parameters that most of us can put together.

Now watc'ha got to say?

 

[maineengineer]

A 60 dB difference between the desired (audio) and undesired (noise) levels is the minimum signal-to-noise ratio that I would recommend to you as even marginally acceptable for your specific application of professional recording for audio books. A sound pressure level meter is not the proper instrument to employ, as it only measures the external noise in your studio physical environment and not the internal noise also contributed by your recording equipment (in the form of hum, hiss, etc). Adobe Audition is suitably acceptable in making this determination, provided that you are proficient in its application. With a broadcast quality microphone, mike preamp, and an audiophile sound card (typically external USB), you should easily be able to attain this standard even in a quiet living room setting (if not, then it's time to consider wall treatment, etc.). Once you achieve this you can then simply amplify the entire production recording to the level the "publisher" requires, and the signal-to-noise ratio will remain exactly the same (i.e. "quiet"). Compression and limiting, by definition, will only make a noise situation worse, not better. Audio books are especially critical with regard to noise, even more so than broadcast applications, which is why those I've worked with have even had separate sound booths constructed entirely within their recording facilities. Your mileage may vary.