[From a submission on another thread: https://www.houzz.com/discussions/6251855/vent-a-hood-liner-vent-for-sale#27666135]

Perfect is the enemy of good enough, but so also is too simple (excessive avoidance of complexity). On the basis of the comments in the thread referenced above, I concur that it might be best to just establish a table of rows by specimen measurement, with columns by manufacturer, model, blower rated CFM, duct length and diameter (if known), use of a silencer (if known), MUA type (if known). Then users will have to find cases closest to their intent and try to extrapolate to their specific case. Otherwise, we need a lot of fan curves (flow vs. pressure) vs. power level vs. sound level (dBa at 1m or whatever). The required complexity is mind boggling if one wants to try to determine a close result -- not unlike the complexity of estimating adequacy of a given hood over various cooking activities without running a full computational fluid dynamics analysis of the hood and MUA systems (plus extraneous room drafts and turbulence).

Good luck and best regards @opaone!

From: https://www.houzz.com/discussions/6251855/vent-a-hood-liner-vent-for-sale#n=25

"It isn't likely to be sones/1,000 CFM either. Discuss with @kaseki to understand why. You are still presenting something as linear, which is unlikely to be linear."

In the first chart I presented dB(a) which is log. There is nothing we can do about that being log because dB(a) is indeed log. As far as I know the data in that chart is as accurate as possible with the limited data that we have.

Presenting dB measurements to an audience unfamiliar with log measurements is problematic. I was a touring FOH engineer and studio engineer for about 10 years. I have a generally good concept of what the difference in 66 dB(a) and 69dB(a) is. I know how much louder that is. Most consumers though think it only 5% louder - relatively imperceptible.

So, in the last chart I presented Sones. Sones is, IMO, a horribly inaccurate number but it does do a much better job of conveying noise experienced to people who are not accustomed to log scales.

"Whoever made the chart manipulated the data and likely did so incorrectly. I suspect the they ran each hood on its highest setting and took a measurement. Then divided that measurement by that CFM and then multiplied by 1,000 CFM. That doesn't work unless the slope is linear."

That is indeed the calc. Please tell me why in this instance that doesn't work. FWIW, this is exactly how the data is calculated and presented in the EU except that they have more accurate base measurements than we have.

"The best thing to do is present the actual unmanipulated test results and let people see the raw data rather than trying to adjust to hold CFM constant.

That is exactly what is presented in the first chart above and here: https://bamasotan.us/range-exhaust-hood-faq/

FWIW, marketing people love to take advantage of consumers lack of understanding of dB log scaling. The raw data 'our hood is ONLY 6 dB louder but much more powerful than competior B' sounds great to unsuspecting consumers who don't have a concept of just how much louder 6 dB really is.

Or you can find the actual logarithmic sound curve of each hood and then adjust them to some common CFM's to see quietest, but with only two data points on some models that doesn't seem likely."

I think that would be great if you would like to do it or fund it. IIRC it's about $12k per hood measured for EU compliant measurements.

-------

@bry911, that I have seen you like to swoop in and offer a lot of critique, some useful/accurate, most not, but you never or rarely offer anything of positive constructive value. You waste a lot of my and others time replying to your innane posts.

You like to sound like you know what you're talking about (I'm guessing that you are a lab tech?) but as soon as you get backed in to a corner or asked for detail to support your positions you disappear.

This is all being done with $0. We have very limited data of questionable accuracy to work with.

Likely the best noise data is my hood since we know exact actual CFM measurements for it and exact sound measurements.

But that is of limited benefit without similar data on other hoods for comparison.

We can only guess at the actual CFM rates of other hoods. I know from conversations with energy auditors who've measured them in homes that better than 50% of rated is extremely rare and many are closer to 20-25% of rated with some below 10%.

Even if we had hyper accurate CFM measurements for other hoods that would still be of somewhat limited beneift as what we really want to know is extraction efficacy.

As I've told you before, if you want to offer something constructive then I'm all ears.

@kaseki: "Perfect is the enemy of good enough, ..."

Yes.

This can all get much too complicated. I'm trying to find a balance between enough information and not too much.

There is also a question of how accurate the base information is. As you pointed out we have little to no clue what the static pressures are on the supply and exhaust sides of the blowers nor what the actual CFM rates are. Extra decimal places aren't very useful if the accuracy of the measurements can't support it :-)

When I get time, hopefully this weekend, I'll see about sticking either the table of raw data in the FAQ or adding the data to the first chart.

@opaone said, "@bry911, that I have seen you like to swoop in and offer a lot of critique, some useful/accurate, most not, but you never or rarely offer anything of positive constructive value. You waste a lot of my and others time replying to your innane posts."

Let's tackle this little gem first. We started this discussion when you posted this nugget of wisdom on a thread someone started to sell a hood that had minor damage that they found unacceptable. "VAH is kind of the loud and noisy dog of hoods that most people don't want."

Let's be clear. Vent-A-Hood is one of 10 companies with greater than 5% share of the residential ventilation hood market, which is a $4.45 billion market. To say that people don't want Vent-A-Hood is a ridiculous assertion, as they still sell a lot of them. All you were doing is trying to jump in there and act superior and add no real value.

@opaone - With all due respect, you consistently misinterpret data when trying to make it more impactful. The last time I pointed this out, you proved unwilling to listen regardless of the effort I put into explaining.

In that case you represented an increase in the risk of illness as an increase in illness. Those differences may seem subtle, but they are tremendously important. Buying a lottery ticket increases your chance of getting rich, but you don't get richer every time you buy a lottery ticket.

----

I have no desire to go a bunch of rounds on this. I will endeavor to explain but I have no interest in turning this into a debate.

This is a fine chart:

There is nothing wrong with this chart as it gives the measurements. However, the chart above can't be converted to this chart below:

The problem has nothing to do with sound pressure being logarithmic, or decibels versus sones. The problem is that you don't know that the increase in sound to CFM is linear. I will endeavor to explain this in the following chart.

My line is just made up... but so is yours. Just because it produces about 62 dB(a) when running at 1000 CFM, doesn't mean it produces 31 dB(a) at 500 CFM. It doesn't matter that dB(a) is logarithmic, you have no data to establish the actual function of dB(a) to CFM (log of logs are quite common).

----

I think you are trying to demonstrate the noise efficiency of various hoods at max CFM, but that is not what your chart says. Nor would I, as someone who has to present data analysis in published research, attempt to visually represent that analysis. It is too likely to be misread as exactly what you presented even if you corrected the title.

Good luck.

ETA: I am sorry for the combative tone in the beginning, this post is certainly colored by our previous interactions. I am happy to help you improve this thread, and/or your blog. However, I am fairly busy right now and if your intention is simply to prove me wrong, I will happily concede defeat right now and move on.

"The problem is that you don't know that the increase in sound to CFM is linear."

It doesn't matter. These are single point measurements.

"My line is just made up... but so is yours. Just because it produces about 62 dB(a) when running at 1000 CFM, doesn't mean it produces 31 dB(a) at 500 CFM."

I have not made up a line nor presumed a line nor presupposed a line. These are single point measurements.

FWIW, if we produced a similar chart with as equal of CFM's as we can guesstimate (presuming that low CFM's is about 50% of high CFM's which is by engineering necessity the case w/ most baffle designs and we do have measurements for low for most of the hoods) it would make the VAH and Kobe look much worse and the others much better. VAH would not change at all but Accurex would be 0 ( dB(a) or sones, take your pick ), Modernaire, Wolf and Prizer would be lower, Kobe would remain about the same. If, as you stated I "misinterpret data when trying to make it more impactful." then I would certainly have done that but I didn't because I'm trying to keep it as simple as possible and simple conveyed the point.

----

@bry911, this is similar to previous times you've made such critiques. I think you have some surface level understanding but you don't understand the underlying principles nor science. Nor do you have a good understanding of statistics or data presentation.

I apologize for being so direct but I believe this is the third or fourth time you've made all kinds of critical comments without anything to back them up and for all of these you have yet to add anything constructive beyond possibly a typo.

I, @kaseki and the others who have been working on this and related topics greatly appreciate valuable input but your input has been of no positive value and has actuallly wasted a lot of people's time.

"However, I am fairly busy right now"

Good.

These charts only list the manufacturer name and no model numbers. Are these the most recent hoods put out by the manufacturer and what models are they? You installed an accurex with a silencer. Was a silencer used to come up with the noise rating for accurex in that chart? Were the other hoods tested using silencers? How many hoods were tested of each model to come up with the chart? Who did the testing? Did they use the same equipment to measure the sound? Were there any other variables not taken into account such as silencers, remote blowers, duct length, duct size etc? Was the testing done in a controlled environment?

@opaone said, "this is similar to previous times you've made such critiques. I think you have some surface level understanding but you don't understand the underlying principles nor science. Nor do you have a good understanding of statistics or data presentation."

Says the person who also says, "I have not made up a line nor presumed a line nor presupposed a line. These are single point measurements."

"dB(a) per 1,000 CFM" is not a single point measurement. It is quite literally a slope. Since we know that hoods make 0 noise when off, it is a formula for a line. You don't understand the difference between a point and a slope and are attacking my understanding of data analysis.

Let's see if we can demonstrate the difference with an unrelated example. Miles per hour (mph) tells you how far you will drive in an hour at that speed. So if we were driving 62 mph for 1/2 an hour we would drive 62*1/2 miles, which equals 31 miles. If you know the mph, we can plug in any number of hours and get the distance.

Your second chart gives dB(a) per CFM. In other words your chart says that total dB(a) = dB(a)/1,000 CFM * CFM. It is wrong. This is middle school level Cartesian plane stuff.

I believe @kaseki sees the mistake and agrees with my assessment but was trying to be polite.

ETA: Just to be clear. Slope is rise over run. Any time you see "per," you have a slope.

Despite the aspersions, I do understand statistical analysis. I have spent an unfortunate amount of time working in R, which is made worse by time spent learning Stata.

There are massive problems with the analysis in this and I normally wouldn't post it, but given the serious issues with the analysis presented in this thread...

https://m.youtube.com/watch?v=AYvzb3VoS-I

I think it is a well-meaning attempt at normalizing performance of examples for comparison that is fraught with hazard due to its potential for misinterpretation. And as we see, it is fraught with hazard due to its potential for argument. I have to admit that I didn't give it much thought before this argument blasted its way out of its crypt.

Getting time to measure in my kitchen is problematic at present, but my intent is to measure SPL vs. actual measured air flow rate (at one point under the hood) to reveal what the relationship is for my particular configuration.

My air speed sensor will be the one I described in Hood FAQ I.

My sound sensor will be an early Radio Shack 33-1028 meter that was sold to me by my then employer as scrap test equipment. It may have hidden modifications. I intend to assume that nothing was done to make its relative measurement defective. I don't have a 'smart' cellular telephone that I can put an SPL measuring app onto.

Note that this meter is unweighted, not "A" weighted, but historical frequency-response reports suggest that the effective weighting is "C" due to the microphone. "A" weighting fits the human loudness curve only for weak sounds. If my hood were as quiet as my Miele dishwasher, then "A" weighting would be a good filter function. "C" is probably better for noisy hoods, even if some clueless government factotum decreed that everything everywhere should be measured in "A" weighting, even the whomp-whomp of wind farms.

For dealing with the ambient, it is probably best to convert dB into pascals, rms, and assume the background adds in quadrature to the hood noise, assuming my ambient is high enough for this meter to read it.

"dB(a) per 1,000 CFM" is not a single point measurement. It is quite literally a slope."

You are confusing jr high school geometry with the real world.

We do not have a slope or a line. We do not have data to produce a slope or a line or a curve or S-curve or knee or whatever. We have a few individual single points of data (that are far less accurate than we'd like but appearing much more accurate than I'd have hoped).

-----

Getting a bunch of data points and fitting a line to them would be great (if of limited or zero real world value) if this were lab data. This IS NOT lab data. This is not a controlled environment.

OTOH, these are real world measurements of real world world range exhuast hoods in real world homes so in some sense are more valuable than lab data. By having people follow the same protocol as much as possible (using the same device (hopefully iphone 8 or newer), same app, standing in the same place facing the same direction, ...) we can hopefully reduce what variances/errors we can so that we can have fairly accurate comparisons.

Going beyond this without better base data collected in a lab environment will only introduce potential for greater error and greater misunderstanding.

You labeled your table as dB(a) / 100 CFM. That is a slope! Here is a hint: "/" means "divided by."

A point has an X and Y coordinate. Here is an example of a point... a Prizer hood produces 72 dB(a) at 100 CFM. The Y coordinate would be 72 dB(a) and the X coordinate would be 100 CFM.

A slope has rise over run. Here is an example of a slope... a Prizer hood produces 72 dB(a) / 100 CFM. There is no given X or Y coordinate. There is the change in Y per change in X.

Here is a middle school math problem using your chart. If a Prizer hood produces 72 dB(a) / 100 CFM, how many dB(a)'s would be produced at 800 CFM. The fact that anyone here can solve that problem means that you have a slope and not a point. If one did solve that problem they would arrive at the Prizer hood producing 576 decibels. Which is obviously wrong. (I realize that you made a typo in the chart, but that starts to become relevant as you denigrate my understanding of stats and data presentation). Just to be clear this is wrong in the real world, we all use slope in our lives.

You are the one who drug this conversation over here. I was completely fine pointing out your error on another thread and going about my day. You have good intentions but that doesn't make you correct. Since you are the one who is accusing me of not understanding math and stats, let's take this to some math or stats boards over on reddit and see what they think.

"these are real world measurements of real world world range exhaust hoods in real world homes so in some sense are more valuable than lab data."

Not really. You are putting a chart up that has no controls. If one hood is being measured with a silencer against another hood without a silencer then you are not comparing apples to apples. You have no models for the hoods just brands. Are these hoods 10 years old, 5 years old, the most recent models? Some are residential and some are commercial. It makes a difference.

The chart is deceiving.

these are real world measurements of real world world range exhaust hoods in real world homes so in some sense are more valuable than lab data.

As@vinmarks points out, this is not correct.

There is a false dichotomy here. There are really just practical limits on the money that can be spent studying exhaust hoods.

There are a lot of things that might cause a hood to be louder or quieter in a home. Researchers typically attempt to control for those things by removing their variability. A great "lab test," would be a series of actual hoods installed in the same way in an actual house. In that situation we don't need to worry about eliminating differences in installation because all the hoods were tested under the same conditions. Because it is impractical to install a series of hoods into the same home, these tests are often done on using some sort of quasi-installation that still hold the condition constant. Another option would be to take a lot of measurements from a lot of hoods installed a lot of different ways to arrive at an average for each and then compare.

It is a fair criticism to note that quasi-installed hoods are different than installed hoods and that the results achieved in that test are different than they would be once installed. However, it is unfair to say that a test with a lot of uncontrolled variables is better than a test with controlled variables. You really don't, and can't, know that.

This is not to say that examples of tests done in a few different installations doesn't have value. I am saying that you really want to be careful about manipulating that data. It is great to say that a particular hood achieved a particular result in a particular installation.

"Not really. You are putting a chart up that has no controls. If one hood is being measured with a silencer against another hood without a silencer then you are not comparing apples to apples. You have no models for the hoods just brands. Are these hoods 10 years old, 5 years old, the most recent models?"

Evaluation of range exhaust hoods/systems is, by necessity, a system v system comparison, not a component v component comparison. Some are fully self-contained all-in-one units while others are component systems of hood, silencer, blower and controls as 4 separate components that are combined on-site.

A silencer is not an option with an all-in-one unit, it is a common component on seperate component systems and is one of the advantages to choosing a component system with a remote blower. It would actually be deceiving to not include the silencer since that is one of the key benefits to choosing a component system.

A lab test would actually not change things that much. The biggest difference would be knowing accurate CFM's given some specific duct/static-pressure systems - which would be quite helpful. A lab could have similar or identical SP's for each portion of the system and could vary these to show how they affect noise and CFM's. Components though would be the same with the only difference being that a lab test might provide a with and without silencer measurement for component systems.

"Some are residential and some are commercial. It makes a difference."

Yes. And no. If you read it that is a key point of the FAQ - that 'commercial' hoods all have large containment volume (and better designed baffles) and consumer, very stupidly, do not. But that is really the only significant difference - a bunch of empty space. Keep in mind that the hood itself is just a simple sheet metal box. That's it. My 'commercial' box is just taller than typical 'consumer' boxes.

The only thing that makes mine 'commercial' and @kaseki's 'consumer' is who sells it. Wolf or Modernaire or Prizer or Prestige could easily sell a hood exactlly like mine in which case would that exact same box then be 'consumer'? I could just as well have paired my 'commercial' box with the same 'consumer' blower as @kaseki or any other. But @kaseki's blower is, I believe, actually the same as is used in commercial systems.

In the end though, 'consumer' hood mfr's focus on marketing and taking advantage of the gullibility of American consumers rather than actual performance. So, if you want the quietest and best performing system then today you must buy a 'commercial' system.

"The chart is deceiving."

How so? The chart does a quite good job of showing what real world actual results people can expect from different options (and actually downplays how much quieter the Accurex, Modernaire and Wolf systems are).

If someone installs the same Accurex system that I have in their home they can expect pretty much the same level of quietness depicted in the chart. If they install a VAH then they can expect the same noise as depicted in the chart. How is that deceiving?

How so? The chart does a quite good job of showing what real world actual results people can expect from different options

The first chart (Range Hood Noise @ Highest Setting) does a somewhat good job of showing what real world actual results people can expect from Vent-A-Hood, and notes what a single installation of all the others were. That is all.

The second chart is wrong, biased, and too likely to be misinterpreted even if it were correct.

• It is wrong. It is not dB(a) / 100 CFM. If that were correct your quietest hood (the Accurex) which looks to be about 38 d(B)a/100 CFM would be 532 decibels @ 1400 CFM. Just for comparison Krakatoa's eruption, which is purported to be the loudest sound ever heard by humans, was only thought to be 310 decibels.
• I realize it is a typo, but the fact that you posted this chart around for this long without realizing it was a typo goes a long way in demonstrating just how unreliable your assertions are. It was only after I did the math that you realized the mistake and then proceeded to claim I have "some surface level understanding" but I supposedly don't "have a good understanding of statistics or data presentation." You make a mistake that no data scientist and relatively few high schoolers would make and tell me I don't understand statistics.
• It is also wrong because it is a slope and it assumes a linear slope. A fact that you already admitted in another thread but somehow forgot once you drug the thread over here. Just to be clear, I calculated the sones to decibels using it as a slope and got a ridiculous number which you admitted was too high. If I can use it as a slope in the conversion of sones to decibels in the calculation that discovered your mistake... it is a slope.
• It is biased and likely to be misinterpreted because it assumes that sound efficiency of marginal CFM is always good. It really isn't. Here is the ELI5 version. In your first chart we see the third loudest hood is Wolf, but in the d(B)a / 100 chart we see it as the third quietest. That is because it produces more CFM at full power, but if you don't need 1,500 CFM then there is no benefit to a hood that is louder in daily use even if it is much more efficient. You are just adding useless capacity for a louder hood. You didn't provide measurements at lower CFM's and since you admit that it is not linear, you can't arrive at the sound level of Wolf at lower CFM's.
• Where are other metrics. Where is the $ / dba chart?

I honestly believe there is value in what you are doing. However, you consistently attempt to add impact to your analysis by manipulating the data and you consistently show why you shouldn't be doing that.

My main blower is a Wolf sourced (but Broan manufactured) down-roof assembly not very dissimilar from the counterpart available from Abbaka. It may have imposed requirements different than the model Broan sells. My auxiliary blower is a baby down-blast blower sold originally by NuTone. Its style is commercial but flow rate is too low for most commercial purposes.

In my mind, the physical difference between residential and commercial hoods is their setting height 2.5 - 3.0 ft vs. 4 ft above the cooktop, and their volume under the baffles for averaging transient plumes. More volume means greater hood height, although some of this can be achieved by larger hood entry area.

@opaone how tall is your hood assembly? Residential hoods have to accommodate ceilings at 7 to 8 feet tall. This precludes Wolf etc. selling tall reservoir commercial style hoods into this market. (As one might imagine, a hood at 7 ft under an 8-ft ceiling won't have room for reservoir, baffles, and above-baffle volume for gathering the flow into the duct while keeping the flow even over the baffle space.

Another factor that may be in play is that residential stainless steel hoods are expected to remain shiny and unstained. For this purpose, the higher-end manufacturers use 300 series CRES,* with its higher difficulty of fabrication and higher cost (includes chromium and nickel). Commercial hoods in many cases are stuck in kitchens hidden from the public, and they are fine using 400 series or similar CRES, which will stain over time but not corrode.

Fancy residential kitchen hoods need perfect welds that won't discolor over time.

(Is your hood magnetic? 400 series is, 300 series isn't.)

________

*Corrosion resisting steel

How so? The chart does a quite good job of showing what real world actual results people can expect from different options

What options? All you do is list a manufacturer name. You do not say what options each hood was installed with. You do not list model numbers. You do not list how old the hoods are. You can't compare a hood installed 15 years ago to one installed now because changes may have been made.

If they install a VAH then they can expect the same noise as depicted in the chart.

Really? Well which Ventahood models are depicted in the chart? Ductless, remote blower, baffle system etc?

Kaseki has a Wolf hood that he says is quiet. I had a Wolf hood in my previous house which sounded like a freight train. Why the difference? They are both Wolf hoods correct?

The charts lack a lot of information to be helpful. There are too many factors that go into how loud a hood is and it is not just based on brand. But according to your chart that is all the information we are given. A brand name.

Maybe include all of what went into your accurex installation. Commercial custom made hood, silencer in the duct, proper isolation of blower and other components from the structure, makeup air, 18 inch duct size. Include model numbers for the hoods tested. Include how old they are, if silencers were used, are they remote blowers and if so where was the remote blower located, was proper duct size used. What other factors may have influenced how loud or quiet the hood was.

"Kaseki has a Wolf hood that he says is quiet." Umm, better would be to say quiet enough to converse under with a normal speaking voice; it is not as quiet as a Miele dishwasher, next to which one can converse at a whisper.

@kaseki yes quiet enough. My Wolf was so loud. Forget about talking anywhere in the kitchen with it on.

@vinmarks Did you have an external blower with a silencer in between?

@kaseki I had an external blower but no silencer. The blower was right on the outside wall which I am sure contributed to some of the noise.

Thanks. Barely different than internal, in that case, unless an overpowered, under-operated blower was deliberately chosen to minimize noise.

"What options? All you do is list a manufacturer name. You do not say what options each hood was installed with. You do not list model numbers. You do not list how old the hoods are. You can't compare a hood installed 15 years ago to one installed now because changes may have been made."

Agree. I thought I'd included this in the discussion but did not. I add what info I have. FWIW, the Accurex, Modernaire and Wolf are all external blowers w/ silencers. The VAH are all MagicLung.

"The charts lack a lot of information to be helpful. There are too many factors that go into how loud a hood is and it is not just based on brand. But according to your chart that is all the information we are given. A brand name."

Yes. In some cases though this is simply the best info that we have. I wish that we had better info but we don't.

"Maybe include all of what went into your accurex installation. Commercial custom made hood, silencer..."

This is a post that's been in the works for some time and I've simply not had the time to finish it. I think it is all covered in the thread but that's a gob of reading to find it. Hopefully soon.

"The blower was right on the outside wall which I am sure contributed to some of the noise."

How much duct was there between your hood and blower? How many elbows of what degrees?

I added a details chart with what information I have.

Here are my measurements, better late than never I suppose.

Hood system characteristics are given above and in other threads. Blower control is continuous type, so "medium" speed is not perfectly repeatable, whereas "slow" minimum and "fast" (maximum) are.

Measurement is by Triplett SoniChek PRO 3550 with windscreen on. This device has a specified error of +/- 1.5dB. Operation was in "slow" mode (2 sec) and "A" weighting. Slow is not slow enough to average out whatever was slightly modulating the sound level, the the rough limits from eyeball observation are listed below. Mid value may be good enough given the base error.

There are two sets of measurements, denoted "cook" and "tripod." "Cook" is hand held at my approximate ear level (~65 inches) standing at the cooking counter edge (roughly under the light bar of my Pro Island Wolf hood). "Tripod" was intended to meet @opaone's designated distances of this thread, but kitchen layout precluded achieving that exactly. Distance from sound sensor to inside of the light bar (edge of capture aperture) was 32 inches; distance from outside of hood depends on definition; height above floor was 5 ft, 9 inches. Note that this hood is wide enough that one is in the near sound field over over a range of positions along the hood length. This also means that 1/R^2 does not exactly apply if one were to try to scale to other perpendicular distances.

Ambient varied apparently randomly between 30 and 37 dBA over numerous sequential sensor displayed readings. I couldn't hear obvious noises that were varying, so I can't explain it.

The "Tripod" "slow" motor speed result is likely being modulated by whatever modulates the ambient, but it is possible to hear that the blower is on in this condition.

Your hood is very quiet!

@opaone I would like to know if you have any general thoughts about the actual noise level of a range hood compared to its published noise level . I have recently bought two range hoods and been frustrated that the actual noose levels were MUCH higher than was claimed in their web sites and manuals. For one of them, the installer said ”thats’s just how it is.” Is that true? Do range hood manufacturers commonly publish misleading noise levels . My current range hood is a GE UVW9631 . They say the noise level on low fan speed is 1 sone which is about 28dB . The actual level for my range hood is 51dB . What is going on?

Unlike the EU that requires specific measurements, the U.S. does not as far as I know. So manufacturers are free to measure however they want and obviously choose to show their product in the best light. My guess is that they measure in an anechoic chamber and measure with no ducting, short ducting and longer ducting and then publish whatever provides the lowest measurement.

So yeah, pretty much worthless.

Which is the point of this thread - to collect real data that provide real benefit to people.

BTW, 1 sone is not necessarily 28dB. That sounds like marketing poop.

@kaseki, I'm looking at the 42" Wolf chimney-style hood with an 1100cfm inline blower. The duct will go up to the attic and turn once to go out. There will be additional soundproofing in the ceiling above the kitchen. I asked about a silencer, and my architect said he's never needed to spec a silencer with this kind of setup. What do you think? Is the inline blower generally quiet enough?

@opaone Thanks for the reply . My recent google search says that there is a direct formula to convert sones to decibels . Maybe this will help some people:

dB = 33.2×log10(Sones) + 28

But for those a little less math oriented, there are conversion tables online. Of course, this formula does not help to tell you how loud a range hood will be based on published information since the published information is garbage . But at least it will help people prove that their range hood manufacturer is a liar . I did exactly that with a previous range hood and got a refund .

So here are my data:

Range hood make and model: GE UVW9631 professionally installed. Duct is about 1.5 meters long , right angle, externally exhausted , a reducer from 8” to 6” was required

Ambient: 35 dB

Low: 51 dB

Medium: 60 dB

High: 64 dB

Boost: 67 dB

Ambient: 35dB

This is the published noise rating from GE (cfm/sones):

100/1.0 (Working Speed)

290/2.5 (Medium Speed)

380/4.0 (High Speed)

610/- (Boost Speed)

For reference, the actual noise level of 51dB on low fan speed is about 5 sones , compared to the advertised 1 sone, 28dB. Way off.

@2rickies: Generally, compact high CFM blowers (versus, say, a Casablanca ceiling fan) have significant blade tip turbulence generated noise. This is best attenuated by having a silencer between the blower and the hood such as the silencers Fantech makes. Of course, you have to fit it into the space available, and they are 4-inches larger in diameter than the ducting. An external blower instead of an in-line blower may allow additional length if needed, particularly if angling the duct path towards a particular roof location.

Suggest purchasing the connection clamps from Fantech to make installation easier.

Note that noise goes as some power of the blade tip velocity, so another approach to quiet is to use an oversized blower and be able to operate at less than full speed when searing or wok cooking.

https://shop.fantech.net/en-US/silencers/c115516

https://shop.fantech.net/en-US/mounting--equipment/c115517

Thank you @kaseki, this is very helpful. I think/hope the blower I'm planning is larger than what I'll need, so I'm hoping I can use it on lower settings. I'll have a 36" induction cooktop, vs the pro gas range I have right now, so at least I won't have to clear burning propane.

For 'quiet' cooking, induction will result in a lower plume velocity as well as plume volumetric flow rate, but that isn't the condition that needs to be dealt with. Searing meat, wok cooking, etc., will lead to large transient plumes, and for this one needs high CFM and/or a large volume under the baffles to average out the plume across them. This latter capability is typically only available from commercial hoods.

Some here have gone quasi commercial (see https://bamasotan.us/range-exhaust-hood-faq/) with varying volume under the baffles for just this reason.

@kaseki, I see. The KD is bent on finding a modern (flat rather than pyramid) hood, and I'm insisting on angled baffles...They priced hoods from modernaire that met both of those requirements, but they were out of my budget range. I've been focused on the chimney style with angled baffles that seems readily available for residential customers and in my price range. The bulky appearance of a commercial hood is also an issue, because it will be against a wall with no cabinets above it and it needs a chimney extension for a 13-ft ceiling. It's at one end of a great room, so very visible from everywhere.

I'm wondering about something you said about silencers--how is it they're too large for the ducts? I see fantechs that are made for 10" ducts, which would be the size i need. I must be misunderstanding something. And would a silencer cut down on the potential cfms? Or does it have no effect on that?

The silencers are not too large for the ducts, but they may be too large for the space available. I have a 10-inch duct from my hood to my roof blower. If you were in my attic and could see the amount of bulk that represents, including the silencer section that is 14 inches in diameter, the hangers, and the dampening material I used on the duct sheet metal, you might wonder how someone could try to stick the duct and silencer in a wall cavity, or some other compact area. (I think @opaone built an internal chase in his house for his ducts.)

I also have an 8-inch duct with silencer going to the roof where a down-blast NuTone blower is mounted. This adds a bit more volume fill.

There is very low pressure loss in the Fantech silencers, probably nearly the same as the equivalent length of duct by itself. I once got a number from Fantech but would have to hunt for a while to find it. The interior of the silencer is the diameter of the duct, with little holes entering the outer volume (that is filled with some dampening material, I suspect).

@kaseki thank you for clarifying. I should have plenty of room in the attic where the duct will run, and I'm going to add a Fantech to my plan.

I read @opaone's guide--very helpful. I noticed a recommendation there for a Prestige canopy-style hood as being better than most noncommercial hoods in terms of containment area, etc. I looked at their classic chimney hood (unfortunately I can't use a canopy hood), and it also seems better designed than the one I was considering. I like that the lights are in the center, and not in the front. It should cover the front burners better as a result.

I think sones to dB is a bit more complicated. Partially because a simple calc like that doesn't account for frequencies which is why I was hesitiant to include a sones chart.

In the end the best vehicle for accurate comparison is I think using dB(a) and taken 1m in front of the hood at a height of 1.75m.

Oh well, looks like I can't use an inline blower with that Prestige pyramid classic hood, so i'm back to the Wolf or something else... It would be simple if I could just go with a canopy hood.

I am not sure whether this works. I suggest: just test range hood at each speed, attach mfg data of CFM would be better. No need to test on different freq. you need to refer to US HIV/Hers, I think it asks for 200CFM <= 3sone. I think 200/400CFM would be enough. noise at 100CFM would be too low to consider, 100 CFM also too low to be useful much.

why 1.75m (5'9") from the floor? the ear height when people stand?

Data provided by manufacturers in the U.S. is largely useless because there are no standards for how that data must be gathered and presented. This applies to both CFM and noise. Many consumer range hoods sold in the U.S. actually produce about 1/2 to 1/3 of the CFM's advertised.

400 (actual) CFM is not enough airflow to remove harmful effluents in many and likely most cases. And as pointed out in the FAQ, there is more to effective effluent removal than just CFM's.

Noise has little to do with CFM. There are many consumer range hoods sold in the U.S. that produce much greater noise @ 100 CFM than a commercial Accurex does at 1200 CFM. Noise is a function of turbulence, not CFM.

Yes, 1.75m approximates the height of human ears.

"Noise is a function of turbulence, not CFM."

Hmmm. Noise is a direct function of turbulence, but turbulence in a given system is a function of CFM -- perhaps square law - ish -- similar to pressure loss because it will track air velocity, whether in a housing, a duct, or at the blower blade tips. So, @opaone, I would not word it the way you did because non-technical readers might be misled. Your earlier point that design can influence noise in spite of CFM is valid and important.

I don't think it's even square lawish... I can move 12,000 cfm of air through a duct system in near silence with appropriate design to maintain laminar flow and suppress vibration transmission. You can stand directly next to a 12,000 cfm duct and the only way you'd know there's air moving through is when there's enough dust that particles are hitting the sides or vanes.

Turbulence is only a function of CFM when a system is not designed appropriately.

That said, turbulence, and some noise with it, is sometimes a necessary element. In the case of a range hood it's turbulence that seperates grease from the airstream and leaves that grease in the baffles rather than spewing it outside and on to adjacent surfaces.

"Data provided by manufacturers in the U.S. is largely useless because there are no standards for how that data must be gathered and presented. This applies to both CFM and noise." The Home Ventilation Institute (HVI) has established testing protocols for range hoods that includes both air flow and sound levels. HVI certification requires equipment be tested by independent/third-party testing laboratories. While manufacturers are not obligated to certify the performance of their equipment, the performance of different makes/models which are HVI certified can be directly compared.

Baffles work by centrifugal extraction, which would function even with laminar flow. (Yes, physics students, in modern terminology "centrifugal force" is depreciated -- mainly because it doesn't exist in these scenarios; the baffles force centripetal air flow and the grease tangential momentum causes it to impact the baffle surfaces.)

Thanks. Totally forgot about HVI. It'd be nice if testing like this were mandated and aligned w/ EU/CE standards.

"Baffles work by centrifugal extraction"

Yes. But if there is noise then I think it's no longer laminar flow? It may be very near laminar but noise indicates turbulence.

Very little if any air flow in a hood system is laminar. I used the term to note that even turbulence-free flow can still impinge grease particles on surfaces when the flow has to curve (change direction). Now, the act of impingement might lead to some turbulence, but the point is that it is not turbulence that causes grease extraction, but flow bending.

And using my Wolf Pro Island hood as an example, the baffle hiss tells us that there is turbulence there, likely enhanced by the baffle edges. The shape is also not smoothly curved, but \_/ shaped, so I would never assert laminar flow was present.