Home Theater Forum and Systems banner

1 - 18 of 18 Posts

·
Registered
Joined
·
37 Posts
Discussion Starter #1




Hi,

Above are screenshots from a Measurement I took of my control room. I can't tell if I just don't know how to take measurements or if I just need help interpreting those measurements as reflected in particular in some waterfall graphs. The graphs above include one SPL measurement for context and then two sets of waterfall graphs (each set contains one screen shot of graph and then screen shot of graph showing the settings used for that graph). One screen shot is for the frequency range: 15hz-400hz and the other is 15hz-22,000khz.

Below is a description of how I configured REW, how I took the measurements and what my Control Room is like. I would like help understanding the waterfall graphs above so that I can determine whether my room needs more treatment, is as good as I can get it, or needs less treatment; and, depending on the answer, whether EQ would be appropriate at this time before getting on with mixing and mastering somewhat soft, though occasionally hard, pop/rock.

Firstly, my control room is roughly 15' x 12'. I am oriented parallel with the 12' walls; perpendicular to the 15' walls. I sit basically in the center of the room facing a wall that is mainly windows which are completely covered with several sound absorption blankets. I am a little off center about 20 closer to the wall to my left. My speakers are obviously closer to the wall I am facing. The speakers are roughly 3-5 feet from the back wall and I have two sets of near-field speakers. One set, NS-10s (Frequency range: 90hz-18,000khz) make a a 10 foot -sided triangle with me in my listening position, while my other set of near-field speakers sit inside of the NS-10s. They are JBL 4410s (Freq. Range - 30hz-20,000khz) and they make a 6'-sided triangle with me at one end. There is no subwoofer and I took my measurements with the correct setting, accordingly.

Where I sit is basically in the center of the room towards a wall which is made up mostly of windows that are covered with two large sound absorption blankets. I am a bit off-center from the walls to my sides - so that I am roughly two feet closer to the wall to my right. That wall is actually a slightly slanted wall, so the room is not a true rectangle. That wall meets the corner of the wall I am facing at about s 10-15 degree, so at that end of the room the angled wall is about 12' from the opposite wall. If you follow the wall to the other end, it is about 14' from the opposite wall where it meets the back wall behind me (if you can picture this; it's hard to describe, but too difficult to capture in a photo).

As to room treatment, the 12' wide wall I am facing, as I said, is made up mostly of windows and they are covered with sound absorption blankets. The side walls and areas of the ceiling are treated with quite a bit of Auralex 2" squares in shapes; some on the ceiling right above the speakers, some right above my head where I sit. Three of the four corners have two 4'x2' 4" thick bass traps which go from floor to ceiling (they are at an angle so the sides touch the walls and each side is at roughly a 90 degree angle to the wall? I made these traps using 4" corning fiberglass with the aluminum facing into the room and covered with a burlap-like fabric (I don't remember the exact type of fiberglass, but it is the one recommended on most DIY bass trap sites). I should also mention that I have set up against the back wall several different pieces of furniture, a large bookcase filled with books at the center of the wall and a small chest of wicker drawers next to it. There is one corner of the room I could not put bass traps, or really even treat because there is a door 1 foot from the corner against the two adjacent walls.


I took the measurement from which the above screen shots of the graph I generated using a Cross-Spectrum calibrated EMM-6 and the the calibration file loaded with a 90 degree narrow band file, with the microphone facing the ceiling. My soundcard, Lynx L22, cal file was also loaded and I calibrated the SPL with an analog Radio Shack SP meter set to 80 "C" and "Slow." When calibrating, I entered, I believe about 156 dbs to calibrate.

When taking the measurement, the microphone was facing the ceiling roughly where my right ear is in listening position with both JBL 4410 speakers playing the output of the right channel from my soundcard. The preamp showed the level of the speakers was just below 0 dbs. The measurement was a 512k, 11.9s, single sweep from 15hz to 22,000khz.

As you can see I have one screen shot showing the SPL from 15hz to 22,000khz for reference and then two pairs of waterfall screen shots (each pair shows one of the waterfall, and then the same waterfall with the "Limits" and "Control" box on the screen for reference. The first and more important waterfall is from 15hz to 400hz with 5db increments on the y-axis, the second is from 15hz to 22,000khz with, if I remember 10db increments on the y-axis.

One thing that confuses me is that with all my measurements, my SPL and the top of my waterfall graph shows up at about 160 dbs. Is that correct. Also, with the waterfall graph, there is no way to show both the top of the graph and the sound floor in 5 db increments, but I can see them at 10db increments because the top is around 160dbs, while the sound floor starts to appear between 45db and 65db. Also, the reading on the 15hz-400hz waterfall that concerns me is that from 15hz to I believe 60hz, or 90hz (see photo, I can't remember) the sound at that end is still going strong whether I select 400ms as in the graphs you see, or even if I select 1000ms. I know the JBLs don't really start to put out much sound until they really hit 40hz, so why is there so much sound below 40hz?

Can you tell from what you see and what I've said whether I'm taking measurements properly? If I am, can you tell me what I should deduce from the waterfall graph? Can I learn anything from the SPL graph. What other graphs should I generate based on what you see to help me focus on whatever problems you see?

Any help would be greatly appreciated as I am new to this and I have spent countless hours reading through Theater Shack forum threads, the REW manual, RealTraps Info, and anything else I could learn about the subject.

Thank you if you even read through this let alone provided any response or guidance. Thank you again!
 

·
Super Moderator
Joined
·
9,252 Posts
Hey marsx,

Unfortunately you’re waterfalls aren’t valid. Check this post to see what a good waterfall should look like, as well as some basic information on how to interpret them.

A valid waterfall has to start with a valid frequency response measurement, and you don’t have one of those either. You might want to open a new thread on that topic to get help with that.

In addition, this post will show how to properly scale your graphs, although for a waterfall I’d drop the floor to 35 dB and extend the upper limit to about 400 Hz (waterfalls aren’t valid above that point). Also, the “Capture” button at the top left of the graph will let you save the graph as a .jpg so you don’t have to do screen shots.

I took the measurement from which the above screen shots of the graph I generated using a Cross-Spectrum calibrated EMM-6 and the the calibration file loaded with a 90 degree narrow band file, with the microphone facing the ceiling.
That’s fine for low frequency measurements, but for valid full range it’s best to use the 0-degree file and point the mic at the speaker.

Regards,
Wayne
 

·
Premium Member
Joined
·
6,711 Posts
My soundcard, Lynx L22, cal file was also loaded and I calibrated the SPL with an analog Radio Shack SP meter set to 80 "C" and "Slow." When calibrating, I entered, I believe about 156 dbs to calibrate.
Why enter 156 dB? You should enter the reading from the Radio Shack meter.

You may also have ended up with a room measurement loaded as a cal file. Best using the latest REW beta version (from the end of this thread) as it has a few additional checks for common mistakes and includes a range of new features and bug fixes.
 

·
Registered
Joined
·
37 Posts
Discussion Starter #4
Thanks, I misread the screen when I entered the number. I recalibrated and now my readings are in the appropriate range.
 

·
Registered
Joined
·
37 Posts
Discussion Starter #6
I want to post my new graphs. I have already loaded them into my gallery. But when I hit "My Photos" above the box where I am typing this, nothing happens. I don't know if you can look using the following address, but since the "My Photos" button didn't work, I went to my gallery and right-clicked on the image and selected "Copy Image Location." This is the location of the SPL graph: http://www.hometheatershack.com/gallery/file.php?n=25073&w=s
The location of the waterfall graph is: http://www.hometheatershack.com/gallery/file.php?n=25081&w=s

If you can view those, maybe you can give me advice on how to get "My Photos" button working again. I tried exiting Firefox and relogging into Home Theater Shack and returning to this thread, but the button still didn't work. Anyway, if you are able to view the graphs, read on:

I have posted two new graphs from a new measurement I took after making corrections based on suggestions from members above. The measurements were taken with the EMM-6 mic place where my right ear is when I am mixing or listening to my speakers. The speakers that were playing were both JBL 4410 near-field monitors, with a frequency range, according to JBL, of 33hz to 30,0000khz. I have no subwoofer (though I am considering one). The microphone with set at 0 degrees and a cross spectrum 0 degree narrow range cal file was loaded. I took the measurements using frequency sweep from 15hz to 22,000khz for 1M or 23.8 seconds. I have posted both the SPL graph for the whole measurement range with 1/6 smoothing and the SPLs are all over the place and I'm not sure what to make of that. I have also uploaded a waterfall for the frequency range of 15hz to 400hz to focus in on the bass response of the room, and I used the default controls and you can see from the graph, the settings. I am confused by the waterfall readings as well.

If anyone can help me understand these two graphs I would be greatly obliged. Also, please tell me that this time it looks like I took my measurement correctly. If not, more guidance would be helpful; but I followed all the advice in this thread and did some more research and I certainly had miscalibrated the SPL because I misread the calibration screen, but now I think it is calibrated correctly.

Thanks.
 

·
Registered
Joined
·
37 Posts
Discussion Starter #7

My "My Photos Button" works again; I had to reboot my computer. Here are the photos for my last post.

Thanks all!
 

·
Registered
Joined
·
199 Posts
It looks like your measurement technique is good now, although I normally measure at higher SPL levels. Since a typical noise level is 45-50dB, I recommend measuring at a level of 90dB in order to provide more headroom for the waterfall.

The frequency response shows that bass response is almost non-existent. I think you would significantly benefit from a subwoofer. The frequency response above 200Hz looks reasonable.
 

·
Super Moderator
Joined
·
9,252 Posts
I have also uploaded a waterfall for the frequency range of 15hz to 400hz to focus in on the bass response of the room...

I am confused by the waterfall readings as well.
Your measurement is valid now, but you don’t really have any bass response so there’s not much a waterfall can tell us. Waterfalls can show the effectiveness of low frequency treatments like bass traps (e.g. comparing before and after graphs). If you don’t plan on doing any treatments they really are not of much use beyond “FYI.”

Speaking of bass, are these your speakers? If so, I have a hard time believing that name-brand speakers with 10” woofers fall off a cliff below 200 Hz. You have a 15 dB/octave roll-out there. Are you sure you don’t have some kind of high pass filter engaged?




Frequency response above 200 Hz looks good though, except for the broad depression in the 600-1500 Hz range.

Regards,
Wayne
 

·
Registered
Joined
·
37 Posts
Discussion Starter #11
Hi,
Those are similar but I didn't correctly note that mine are the 4410As which supposedly have the same frequency response as the 4410s but there are no controls for EQ on the 4410As, or whatever controls are on the 4410s. However, I notice that the same thing is happening to a lesser degree when I measure my NS-10s (the originals; not the Studios). Though they don't kick in until about 60hz instead of 45hz like the 4410s, in my measurements they also seem to have about 10db less power from 60-70hz and then rise logarithmically up to about 150 in a similar fashion to the 4410s. I found that no matter where I place the mic or take a reading from my Radio Shack SPL set at 90db, I get a similar response virtually everywhere in the room (except about 1 foot or less from the side wall where the low end becomes louder, but still a bit weak), AND this even happens when I point the mic directly at the woofer on a 4410 from about a meter.

HOWEVER, from doing a lot of reading, it may be normal for my room (since both pairs of speakers and both the EMM-6 mic and the mic in my SPL meter can't all be broken). I have written to JBL with a description of my situation and am awaiting a response. But here is a quote from Wikipedia regarding measurement of loudspeakers in a listening room (rather than an anechoic chamber):

"While the very best modern speakers can produce a frequency response flat to ±1 dB from 40 Hz to 20 kHz in anechoic conditions, measurements at 2 m in a real listening room are generally considered good if they are within ±12 dB, and efforts to produce anything like a flat response below 100 Hz are likely to provide endless scope for experimentation, and exercise of patience." My low end response is 10dbs lower than the "baseline" between 1000hz and 2000hz and slowly improves as the frequency moves closer to 100hz and then from 100-150 it is almost within the +/- 2db range stated on JBL's specs.

So who knows? Comments would be appreciated.

Wade
 

·
Registered
Joined
·
37 Posts
Discussion Starter #12
Here is a quote from Sound on Sound that sort of addresses my problem from Sound on Sound magazine. I'd love to get your take on it and how it relates to my low SPL-low end problem:

"Small, square rooms (or worse still, cuboid ones) pose the biggest acoustical problems, especially if your listening position is close to the centre of the room: low frequencies tend to null out at that position, so you hear far less low end than your speakers are actually producing. The temptation is to EQ to compensate for the lack of bass and, consequently, you'll tend to produce very bass‑heavy mixes."

Any responses; agree, disagree, solutions???
 

·
Registered
Joined
·
99 Posts
Hi,
Those are similar but I didn't correctly note that mine are the 4410As which supposedly have the same frequency response as the 4410s but there are no controls for EQ on the 4410As, or whatever controls are on the 4410s. However, I notice that the same thing is happening to a lesser degree when I measure my NS-10s (the originals; not the Studios). Though they don't kick in until about 60hz instead of 45hz like the 4410s, in my measurements they also seem to have about 10db less power from 60-70hz and then rise logarithmically up to about 150 in a similar fashion to the 4410s. I found that no matter where I place the mic or take a reading from my Radio Shack SPL set at 90db, I get a similar response virtually everywhere in the room (except about 1 foot or less from the side wall where the low end becomes louder, but still a bit weak), AND this even happens when I point the mic directly at the woofer on a 4410 from about a meter.

HOWEVER, from doing a lot of reading, it may be normal for my room (since both pairs of speakers and both the EMM-6 mic and the mic in my SPL meter can't all be broken). I have written to JBL with a description of my situation and am awaiting a response. But here is a quote from Wikipedia regarding measurement of loudspeakers in a listening room (rather than an anechoic chamber):

"While the very best modern speakers can produce a frequency response flat to ±1 dB from 40 Hz to 20 kHz in anechoic conditions, measurements at 2 m in a real listening room are generally considered good if they are within ±12 dB, and efforts to produce anything like a flat response below 100 Hz are likely to provide endless scope for experimentation, and exercise of patience." My low end response is 10dbs lower than the "baseline" between 1000hz and 2000hz and slowly improves as the frequency moves closer to 100hz and then from 100-150 it is almost within the +/- 2db range stated on JBL's specs.

So who knows? Comments would be appreciated.

Wade
I wouldn't consider a +/-12db response at the listening position good. But I would say that it is typical for a untreated room.

I also doubt there is a speaker on the planet that is +/-1 40-20k.
 

·
Registered
Joined
·
37 Posts
Discussion Starter #14
Thanks for your repsponse. If anything, the room might be overtreated with floor carpet, some auralex 2" squares above the listening position and above the speakers. Auralex on the walls across from listening position and speakers; a full bookcase against the wall behind me and since the wall in from of me (behind the speakers is almost 3/4 windows, the windows are fully covered with very thick acoustic blankets. Also, since the ceiling is 8' high, I was able to stack two 2'x4' x 4" inch thick 705FRK Owens Corning fiberglass with the foil facing into the room covered with a burlap-like fabric from floor to ceiling, angled against each corner so that it hits each wall evenly, leaving an isoceles triangle of air behind standing at three of the four walls (the fourth wall has a door, so I can't put bass trap.

It is a small room control room, 15' by 12' by 8' high (with the wall of less length does not have parallel walls (one wall is at about a 20 degree angle to the opposite wall). I sit oriented towards the narrower wall off center by about 1 1/2 feet. Could it be that the -10db to -5db response in the low end from 45hz to 150hz is from overtreating such a small room with too much bass trapping or not enough absorption of the highs and mids?

I'm stumped. Any thoughts?

Wade

Could it be that in such a small roo
 

·
Registered
Joined
·
37 Posts
Discussion Starter #15
If I try removing bass traps and other things and can't improve the bass response, what do you think of add a third set of small-field speakers set up on my mixing desk with a subwoofer. I wouldn't want to add a subwoofer with such broad range three way speakers as the JBL 4410s (10" woofer), and I like using NS 10s as they were meant to be used, but I have a pair of Yamaha HM5s sitting around, which I could use and just buy the matched HM10 subwoofer and set the crossover at about 150hz (the end of the range where I seem to have weak bass response)? Just a thought.

Obviously, I think I need to take down the bass treatment and see if that improves the bass response. The walls in the room are inches thick (2" drywall covered with a 2" soundboard and then that covered with another 2" drywall, which is why the room may not be amplifying the low end as it normally would in a small room in a home; does that sound like a reasonable deduction?).

Any thoughts on how best to isolate what's causing the absent low end frequency response, no matter where I take measurements in the room, what the orientation of the mic is and, as I said, I duplicate the results on each speaker, when both speakers are playing, when the NS-10s are playing rather than the JBLs, and when I use my Radio Shack SPL meter instead of the mic, so I know I'm not making a procedural error in measurement. I even took a measurement using a sine wave, starting at 45hz and looking at the db on preamp for mic and at the meter on the SPL meter which coincided and did that from 45hz to 100hz, moving up 1 hz at a time and then from 100hz to 150 hz, every 5 hz and then some random measurements throughout the rest of the frequency range up to 20,000hz and, in general, excluding anomalies, most readings were in what I would call a general baseline range. Surprisingly, there weren't an extraordinary amount of anomalies; just like when I did a measurement in REW from 15hz to 20khz and found that, as I have been discussing, the low end from 45hz (low end of published Freq. Response for JBL 4410A) through 150hz after 1/3 octave smoothing, the response is -10db moving up to -5db in that range, but for the rest of the Frequency range from 150hz to 22khz, the db level, general is zero, except for many nulls and peaks which, for the most part, disappear with 1/3 or 1/6 octave averaging.

I don't know if averaging affects waterfall graphs, but I took it off before making the waterfall graphs and, as I said in my first post, the low end energy just as low as in the SPL graph and goes down virtually at the same rate as the mid and high frequencies; so, from that perspective the bass traps, other treatment, or the room itself seems to be working in the time domain for the low end through 500hz, it diminishes at the same rate as all frequencies above 500hz. SO THAT'S WHY I AM SO CONFUSED AND FRUSTRATED ABOUT THE LOW SP LEVEL IN THE RANGE FROM 45HZ TO 150HZ (or in the case of the NS10s, -15db for 60hz to 70hz and then -8db to moving up to -5db at 150hZ).


Again, has anyone ever had an experience like this? Does anyone have any thoughts about what might be causing this. Can anyone give me some steps to try to reduce the problem before going through the headache of removing most of the treatment and starting from scratch (Something I am loathe to do, yet afraid I may have to). I'm hoping someone may have an idea to narrow the problem down before starting from scratch (as I scratch my head in disbelief.)

BY THE WAY I'm using REW Version 5.0, because the beta version froze every time I tried to select the ASIO drivers for my LYNX L22 card. Question in that regard, which I guess I should start a new thread and post, which is, is there a way around using ASIO drivers for my LYNX Card in the new beta version?

As always, any help is greatly appreciated.

Wade
 

·
Super Moderator
Joined
·
9,252 Posts
Here is a quote from Sound on Sound that sort of addresses my problem from Sound on Sound magazine. I'd love to get your take on it and how it relates to my low SPL-low end problem:

"Small, square rooms (or worse still, cuboid ones) pose the biggest acoustical problems, especially if your listening position is close to the centre of the room: low frequencies tend to null out at that position, so you hear far less low end than your speakers are actually producing. The temptation is to EQ to compensate for the lack of bass and, consequently, you'll tend to produce very bass‑heavy mixes."

Any responses; agree, disagree, solutions???
It’s absolutely true, and if your listening position is in pretty much the dead center of the room (like you noted in your first post), that accounts for the poor bass response in your measurements. I don’t know of any solution for the problem other than just increasing bass output to compensate. Probably your best bet with that would be to add an outboard subwoofer.


I wouldn't want to add a subwoofer with such broad range three way speakers as the JBL 4410s (10" woofer)
Why not? They sure aren’t “broad range” from your listening position – a good speaker with a 4” woofer in most situations can get lower than the 200 Hz extension your measurements are showing. IMO going with smaller speakers may well make the problem worse.

Regards,
Wayne
 

·
Registered
Joined
·
37 Posts
Discussion Starter #18
Thanks for all your help. I am closing this thread, unless you want to go on; I have learned what I need from this. I have started a new thread just focusing on how to deal with a room that eats bass from 45hz to 150hz to make the question straight and to the point. Thanks for all your help and suggestions.

Sincerey,
Wade
 
1 - 18 of 18 Posts
Top