The No.1 Website for Pro Audio
Lets talk graph(s)
Old 17th November 2009
  #1
Lets talk graph(s)

!!!UPDATE, I posted new graphs on 27th of February. Please scroll down if you want to see the current state. You can avoid the older discussion

Hey guys,

Ive been lurking around this board for quite a while and I know most of the basics Ethan and Glen (among others, of course) have been spreading here. I used this knowledge to build a little mix room for myself and I started the first measurements today.

I did not have the time to start the usual photo diary (this will follow, be assured) so here are the facts you need to know:

Room:
Length: 4,1m
Width: 3,1m
Height; 2,5m

I did not use rockwool but something similar made entirely out of wood (it looks like rockwool and has similar "tech specs"), its called GUTEX Thermoflex for those interested.

I built quite a lot of absorbers:

The corners behind the speakers (FOCAL TWINs) are filled with the stuff, 20cm thick and from floor to ceiling.

One rear corner has a corner trap (40cm deep !!), also from floor to ceiling. The other rear corner is untreated since there is the entrance. I might add a door mounted absorber later.

The rear wall has 2 absorbers each 120cmx60cm (10cm thick). These are mounted in the middle of the wall. The ceiling has a cloud consisting of 3 absorbers (2x 180cmx60cm and 1x 120cmx60cm each 10cm thick). The cloud is positioned above the calculated position mix position (38% rule etc etc) and covers the area around that position to eliminate the first reflections. The gap between the ceilind and the cloud is about 10cm.

I also have 2 panels on the sides to kill the 1st reflection points (2x 180cmx60cm, again 10 cm thick).

And last but not least I did add 2 traps for the corners between the floor and the front/rear wall, each of them 120cmx60cm but 20cm thick. I positioned them angular between the floor and the wall(s).

I hope you were able to follow me up to this point

The speakers have been setup 90cm from the front wall and around 80cm from each side wall. That seemed the best position after listening to music and moving the speakers.

During these listening tests I was not really satisfied. It was very obvious that the bass response is not very consistent and there are positions in which almost no bass is audible. Thats kind of the usual problem in a small room like this but to be honest I did expect a better result.

I also thought the room is a little dull, but to be honest I dont know. During the time this was built I listened to music almost exclusivly on headphones and I think I just adjusted to the sound, which always seemed brighter than on speakers (I never thought about it, but is that true?). The spacial image of the music and the sound of the reverb, however, was very satisfying.

Attached are the first measurements of the room. (Red= left, Green=Right)

As you can see, there are 2 major problems. The dip slightly below 100hz and down to about 68Hz and the dip around 2Khz.
Im really wondering why the dip around 100 is more severe on the left speaker...
I also think the decay of highs above 10k seems very steep, any comments on that?

Apart from that, how does that look to you? As mentioned earlier, I did expect better results. I did a similar build a year ago with the same material but in a slightly bigger room (1m more in length and width) and the results were way better.

I must add that I didnt fiddle around with the positions too much yet. The floor-corner traps as well as the 1st reflection panels on the sides are mobile the rest is mounted and thus fixed in its position.

Any recommendations are welcome. Discuss at will

Thank you
chris

ps: In advance, kudos to anyone happily sharing their knowledge on such a complex topic like acoustics. Without this board this room would have never happend!
Attached Thumbnails
Lets talk graph(s)-picture-1.png   Lets talk graph(s)-picture-2.png  
Old 17th November 2009
  #2
Lives for gear
 
Weasel9992's Avatar
 

Hey Chris. You mentioned that one corner had a trap and the other didn't. That might explain the disparity between the left and right speaker plots. I'd also say that the reason the room sounds "dull" is that 2.2Khz valley. I'd take a hard look around your listening position for that one...is there a console, desk or video monitor between you and your speakers? Racks maybe?

Aside from that, it looks like a regular room that needs more treatment. You've done a good bit, but it's not at all unusual to need more than that. My room is close to those dimensions and I have five 6" bass traps, fourteen 4" broad band panels and five 2" panels.



Frank
Old 17th November 2009
  #3
Hi Frank,

right, the one untreated corner. Funny thing is the side which is worse in the measurement is the one with the treated corner

About the 2.2khz dip: I forgot to mention that this room is empty at the moment. No desk, no computer, no screen, just the speakers and the measuring mic. So I am really wondering why this dip is there.

Im still concerned about the loss of highs above 10k as it seems more than usual. That might be due to the carpet floor...

I will post pictures tomorrow in which you can see that most of the space is already covered with bass traps. Right now the only possible option seems to be on the side walls (next to the 1st reflection traps) and the corner door.

Just checked out other graphs on this board. Kind of showed me that Im not on the bad end with my room, phew

I hope that positioning and adjusting the listening position might solve some of my problems.
Old 17th November 2009
  #4
Lives for gear
 
Weasel9992's Avatar
 

Hmmm...I've just been chatting with somebody else about your room, and he brought up the point that the measurement mic may be a bit off center. That could explain the high frequency weirdness to some degree. Also, you should probably be playing the test tone out of both speakers rather than one at a time, at least to get a good reading on low end.

So, check the mic positioning and your playback method. Also, what mic and preamp are you using to conduct the tests? As I've thought about this a bit, are you sure that GUTEX Thermoflex really has the same density properties as 703 or mineral wool? Do you have any documentation that you could post? The only thing I could find was in German.

Frank
Old 17th November 2009
  #5
Lives for gear
 

Any position other than where you listen doesn't matter. Symmetry left to right in front of you is crictical in any case.

Anything above 300Hz is likely either speaker issue or serious reflection issues rather than treatments or seating position.

Bryan
Old 17th November 2009
  #6
Just to clarify the specs of "Thermoflex":

Density is 45kg/mΒ³ but the gas flow resistance is 5000rayls/m. I calculated the effectiveness using the "Porous Absorber Calculator" and it looks just fine. Plus its 100% wood which I really like (it smells like wood for a couple of months).

About the measuring method:

I am using Fuzzmeasure and a Checkmate CM-140 SPL meter (intentionally bought for REW which didnt work with my mac but did work on a pc a year ago). No micpre involved, Lineout into LineIn. The soundcard compensation of Fuzzmeasure is on.

I will double check the mic position tomorrow and I will have to make a cable for feeding both speakers at a time, but thats a valid point too.

And I rather doubt that the 2khz dip is a speaker problem since the speakers are new and have hardly been used(except from slowly burning them in).

Thanks for your help!
Old 18th November 2009
  #7
Ok a little update:

I re-measured using both speakers at once. The results were... quite different.

I also attached pictures this morning (yes, the height of both speakers is not identical in this picture, it is now, dont worry thumbsup )

Im a little confused right now. The bottom does look similar to the old measurements, however the problems are now located higher than I thought and the loss of highs is even more dramatic, which I cannot explain.

Oh wait, maybe I can. Its a C-weigthend SPL meter which results in a frequency range from about 31.5hz to 8khz. Correct me if im wrong?

Nonetheless, the 2khz Bump still remains. I will have to adjust the listening position...

I attached 2 measurements. Picture 3 was done with the bass trap between the speakers moved to the door (out of curiosity and to check if I need to add a mobiel trap at the door) and Picture 4 is the same setup as the first measurements but with both speakers. As you can see the difference of moving the trap is... none.

Suggestions?
Attached Thumbnails
Lets talk graph(s)-studio_web001.jpg   Lets talk graph(s)-studio_web002.jpg   Lets talk graph(s)-studio_web003.jpg   Lets talk graph(s)-picture-3.png   Lets talk graph(s)-picture-4.png  

Old 18th November 2009
  #8
Gear Guru
 
Glenn Kuras's Avatar
I think at this point I would keep moving the mix spot and get that as good as possible. After that then either live with it or add some additional treatment.
Old 18th November 2009
  #9
Hey

Glenn: I will work on that tomorrow.

Victor: As mentioned earlier I used GUTEX in another studio and its in there for over a year now. The smell is gone and we havent had any problems... Maybe a matter of taste or maybe the humidity was too high in your case?

I understand your concerns. I was confused as well but cannot find the difference. The only thing I changed was that the second measurements were done using both speakers simultaniously.

The drop above 8khz is due to the SPL meter.

About the position of the SPL meter: Since its an omni pattern it shouldnt really matter where its pointed, I suppose.

Any more suggestions on the graphs or the possible differences between the measurements?!

thanks
Old 18th November 2009
  #10
Gear Guru
Measurement

Only a purpose built room, much bigger than yours, can have even bass response. Your only available goal should be the mix position only.
Headphones can be a great tool, but you need to know their response. Almost all closed ones are hyped and weird at HF in my experience. There are some wonderful natural sounding ones by Sennheiser. You need a healthy known reference.
I wouldn't sweat the measurement thing. It has very specific uses but overall it's not easily useful or intuitive. There is no agreed way to measure a room. There are various methods recommended by those of us studying the subject but you will get opposing opinions. e.g. Many recommend driving both speakers simultaneously. However I totally disagree with that, FuzzMeasure doesn't even allow it. Unless the single mono mic is exactly between the two tweeters and the paths are identical there will be disastrous HF cancellations at the single point. With the exception of the Final Front Ear, our ears are at two points in space and the path to each is shielded from the other ear by the head. I suggest that unless you get into Binaural measurement, use should use one speaker for Frequency response. Use both if you are investigating modes. In fact it may be best to place the speakers in corners, and the mic in a corner when investigating modes. Another confusion is which way to point the mic? The up thing has to do with diffuse fields, US practice vs EU practice and so on. I say point the mic at the single tweeter you are measuring. Your falling HF is likely due to the mic in the SPL meter, which doesn't need much response above 8K. A falling HF is no bad thing. See Understanding RTA at studiotips.com
I like to see +3 around 100Hz falling to -3 around 10K. This used to be common knowledge, I can't reconcile this with the current common ' flat with pink noise' assumption.
A clear run to the finish line here:-
Don't measure absolutes. Thats for Labs, experts, and calibrated Lab grade equipment.
I would use the measurements only to experiment with speaker and listening positions, including speaker height. Optimise each one separately. Do check Bass at the mix position with both driven, but don't take any notice of the crazy HF readings.
It looks like you have done a good job treating the room. I have only one question, are the speakers firing into the long dimension?
DD

Last edited by DanDan; 18th November 2009 at 06:22 PM.. Reason: Details
Old 18th November 2009
  #11
I didnt see it that way, when positioning the SPL meter. I will try to point it to the ceiling (lets see if the tripod can do that).

Did you mean 10khz?
I really dont know how that happend and why its so different than yesterday. Since the room doenst sound that dull I dont think buying the ECM + preamp will be necessary.

@Dan:

Of course they are firing into the long dimension. :-)

Funny thing is, you might just have explained the HF problem in the later measurements

Quote:
Many recommend driving both speakers simultaneously. However I totally disagree with that, FuzzMeasure doesn't even allow it. Unless the single mono mic is exactly between the two tweeters and the paths are identical there will be disastrous HF cancellations at the single point.
Thanks. Very valid input!
Old 18th November 2009
  #12
Gear Guru
Not Bad

Jeez I was looking at the wrong pictures, i.e. Frank's room. I type quicker than I think :-)
You might find this interesting. The dimensions seem to be quite good viewed in this way.
Lets talk graph(s)-picture-4.gif

Have you considered SBIR? e.g. Try moving the speakers close to the wall and lets see how the bottom end reacts. Also try different speaker heights. Careful position tweaking will very likely improve your Bass response a lot. The 1.5K dip looks like some sort of fight or cancellation.
DD

Last edited by DanDan; 18th November 2009 at 06:54 PM.. Reason: More
Old 18th November 2009
  #13
Gear Guru
 
Glenn Kuras's Avatar
Quote:
Have you considered SBIR? e.g. Try moving the speakers close to the wall and lets see how the bottom end reacts. Also try different speaker heights. Careful position tweaking will very likely improve your Bass response a lot. The 1.5K dip looks like some sort of fight or cancellation.
Once again Dan Dan brings up a good point. Just in case you don't understand SBIR.
Learn what is SBIR (Speaker Boundary Interface Response).
Old 18th November 2009
  #14
Thanks again. I knew about SBIR but didnt think of it before.

I will try all of this tomorrow and will post an update immediately. I guess Ill be moving speakers for the next few days
Old 18th November 2009
  #15
Lives for gear
 
Weasel9992's Avatar
 

Okay, hang on a sec. Those plots don't look right. I mean, I can see the low end response looking similar to that down to 40Hz, but there's absolutely no way on earth you're flat below that. Your speakers won't even produce anything below 30Hz (or not much). Secondly, your high end looks really weird. You should be able to hear that...it should be screamingly obvious. Does the room sound wooly and dark, like there's a blanket thrown over the speakers? According to those plots, you're producing hardly anything above 10Khz, which is hard to believe.

Frank
Old 18th November 2009
  #16
The second plots seem to be flawed, obviously. The longer I look at it, the less realistic does it seem to be DanDan had a shot at guessing the cause (a few posts above, I also quoted it). I will start measurement all over again tomorrow, using the ideas DanDan shared. I will start measuring for each speaker individually and I will try if SBIR placement will improve the bass response.

So far, thanks for all your input!
Old 18th November 2009
  #17
Gear Guru
Crazy plots

Those LF straight lines are probably DC leakage. This is a cheap Chinese made SPL meter from Thomann. Clairvoyant? No, I have one that looks like that, I can measure 20-30mV DC on the AC output. I think there are a couple of things up for grabs.
1 Speaker and listener positions. Tweak to optimise.
2 When that is done, if there is a particular troublesome frequency area, I suggest firing a sinewave at that frequency into the room. Move around, including in particular the ceiling corners which are not trapped. If that frequency is very loud up there, you might consider some ceiling corner traps. Ones with a membrane towards the ceiling. It is not unusual to have a bit of a fight in a room with lots of fully rigid surfaces. I am assuming most of what we see is concrete. For interest, try a measurement with that door open. You may have to do that when doing final mixes.
DD
Old 18th November 2009
  #18
Lives for gear
 
Weasel9992's Avatar
 

I think Dan's on the right track. I'd still like to know how the room *sounds*. Are any of these issues sonically obvious? They should be.

Frank
Old 18th November 2009
  #19
Thanks again Dan.

About the sound: I did the listening tests using sine waves around the 100hz area. And yes this was obvious. As the dip in the first plots showed, there are spaces in which the waves cancled out quite severely.


The dramatic loss of highs we did see in the second set of plots was not audible. As mentioned, something must have went wrong during these tests, I will correct this tomorrow and hopefully show you some more relevant plots. I could have seen that this is nonesense, so Im sorry for any time you lost staring at them.

Which was audible, in a way, was the dip around 2khz. When listening to music I always felt that my favorite vocals were missing some of the action and that was probably it.

I will do all the stuff Dan has mentioned. I will spend tomorrow just moving the speaker and the mic.

The room can be better than it is now, and I want that
Old 19th November 2009
  #20
Okay an update:

I did solve the problems with the graphs, there was something wrong with the SPL meter and some settings (my fault) and now everything should be correct, or at least closer to the truth than yesterday.

I spent the last hours measuring and moving each speaker individually.

Pic5 and 6 show the left and right speaker before moving.

Picture 7 and 8 show the left and right speaker after trying SBIR like Dan suggested. Now thats a big difference. Not just in the measurement but also when listening to music. It improved ALOT!

Picture 9 and 10 show both speakers at their current (maybe final) position. I will tweak the listening position further since right now im sitting 2m away from the speakers with them being only 1,6m apart. During the first listening test I felt that sitting 10-15cm closer to the speakers would improve the overall quality, I might measure this later.

So far Im very pleased with the results and can only say thanks for the idea of SBIR
Attached Thumbnails
Lets talk graph(s)-picture-5.png   Lets talk graph(s)-picture-6.png   Lets talk graph(s)-picture-7.png   Lets talk graph(s)-picture-8.png   Lets talk graph(s)-picture-9.png  

Lets talk graph(s)-picture-10.png  
Old 19th November 2009
  #21
Lives for gear
 
Weasel9992's Avatar
 

Yeah...that's a whole lot better!

Frank
Old 19th November 2009
  #22
Gear Guru
Cool

Good result. Apart from SBIR, moving a speaker closer to the wall causes an upward tilt in overall bass response. Like a shelving tone control. Seeing this dramatic improvement prompts me to do exactly the same in my own room. i.e. I will check that my positions are optimum as hisblu has done here. i.e. I will now act on the advice myself LOL.
All good, DD
Old 19th November 2009
  #23
Lives for gear
 
macc's Avatar
Wow, great stuff!

So where did the speakers actually physically end up to get those measurements? How far from the wall?
Old 19th November 2009
  #24
Quote:
Originally Posted by macc View Post
Wow, great stuff!

So where did the speakers actually physically end up to get those measurements? How far from the wall?
Thanks. I just checked the distances:

Each speaker is 25cm from the front wall with the tweeter being around 55cm from the wall (i had them closer before but this seemed to be the best distance). The width of the room is 3,1m, each speaker is 60cm from the side wall. The height of the room is 2,5m and the tweeter is around 1,23m.

Hope that helps

@Dan: funny I hope your improvement will be just as fine.
Old 20th November 2009
  #25
Lives for gear
 
macc's Avatar
Thanks man, interesting stuff

Enjoy making music in there!
Old 27th February 2010
  #26
!!! UPDATE - New graphs, new absorbers

Id thought I bump this again since I did some further improvements:

As you can see in the previous graphs and discussions, I still had some problems in the bass region and there was a major dip around the 100hz region (almost 15db). Naturally this was very audible and resulted in mixes that had a over exaggerated bass region (who whould have guessed? ).

As I wasnt satisfied I made further adjustments:

1) I removed all DIY absorbtion from the backwall and replaced it with a massive 155x200x20cm absorber which was professionally made and which also includes a resonating plate that is tuned to 80-100hz. I added a smaller version of this one to the bottom corner of the front wall as well (155x90x20cm). The placement of these panels was chosen by listening to the problem frequencies and searching for the most powerful regions within the room. I placed them directly onto the wall.

The old absorber panels from the backwall are now available to be placed on the sidewalls and the door to eliminate any comb filtering in the mid-high frequency regions.

2) I then further adjusted the speaker and listening positions. I spend 2 days measuring and it was very obvious that there were two general positions:

One was similar to the previous position (SBIR, see graphs from above) which also had a similar frequency response, however, the dip improved by about 6-8db.

The second position was further from the front wall (~75cm) and quite a bit fruther from the side walls as well (77cm). This position resulted in the frequency response I attached to this post. I think, overall, its a better, more even response than before and the problem in the bass region is still there, however, less extreme. Please note, that I HAVENT positioned the additional mid-high absorbtion panels so Im really just talking about the bass region.

Since the sweet spot is about 145cm from the speaker point, with the speakers being only 140cm apart (measured tweeter to tweeter) Im a little concerned about my stereo image. I tried pulling them further apart which resulted in way worse frequency response but better imaging.

Why Im posting this? I want feedback about the improvement I might have acchieved and I need tips how to find the best positions for the mid-high absorption panels to elminate the problems at ~400hz, ~700hz and ~2khz. Is there any easy way to identify them (like the mirror trick for the 1st reflection points, etc)?

Once again, Im extremely thankful for any feedback and advice you might contribute and which I already received. Without this forum I wouldnt have come that far

By the way: Im sorry that these measurements have a different scale but I had to change from Fuzzmeasure to RoomEQWizard due to soundcard issues.

Thanks
chris
Attached Thumbnails
Lets talk graph(s)-left.png   Lets talk graph(s)-right.png  
Old 3rd March 2010
  #27
Noone? I need feedback
Old 3rd March 2010
  #28
Gear Guru
Finish

It would be difficult to compare the earlier responses, 9/10, to the new ones. Also would there be any reason to do that? I would install your RFZ side panels and cloud, then tweak your final positions. Optimise what you have, there is no going back! IMHO the 'mirror trick' is absolutely the best way to locate your reflection control panels and to determine how large they need to be. Move around within your target listening area. Have your mirror assistant move it around the wall. Mark the wall at the extremes of where you can see the tweeter. In this way you can see how much of the side wall you need to cover.
DD
πŸ“ Reply
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
πŸ–¨οΈ Show Printable Version
βœ‰οΈ Email this Page
πŸ” Search thread
♾️ Similar Threads
πŸŽ™οΈ View mentioned gear