or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs
New Posts  All Forums:Forum Nav:

Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How... - Page 183

post #5461 of 9539
Quote:
Originally Posted by Sean Spamilton View Post

I measured about 3 inches away from the driver, and with the mic on top of the port ( on a stand with the actual speaker about 2 inches away from the top).

That's too far away. The mic needs to be closer than 10% of the diaphragm radius. You also need to make sure the distance is always the same. In the near field of a driver the level changes rapidly with distance. That's probably why we see that 20dB difference.
Other than that the two subs seems to behave the same.

Gear mentioned in this thread:

post #5462 of 9539
Quote:
Originally Posted by jlpowell84 View Post

So is my 15hz drop a cancellation or something else? Any way to bring that up?

It's probably modal. How does the free field response of the sub look like?
post #5463 of 9539
Quote:
Originally Posted by AustinJerry View Post

@Sean:

To be honest, when I have moved subs around looking for the best position, I have never adjusted the distances in the AVR during the process. I am not sure it is that important. If you are obsessing about sub distances and changing them every time you move a sub, you may be doing more harm than good.

I repeat, I have never gone to this trouble when moving my subs around. Perhaps someone else can share an opinion as to whether they think this is necessary or not.
Yes, it makes a difference.

It has to do with phase cancellation. When the subs are not equidistant to the MLP (especially if the difference is significant), changing the individual delays so the direct wave arrives at the MLP simultaneously ensures that it is in phase. If you don't change the distances and have one sub closer than the other (or move one sub more than the other after running a brief calibration for Audyssey to set the distances), the direct wave could be out of phase, producing what looks like a null due to the cancellation.

If you move the subs around looking for the best positions in the room but don't change the distances/delays, you could end up ruling out, and missing out on placement positions that could produce great results, because not changing the distances produced this phase cancellation. I saw this when I was trying to find the best placements for dual subs. It occurred to me due to the distance tweak, i.e. if inaccurate delay settings for the subs could produce a dip in the bass region that could be fixed post-calibration simply by altering the delays, this must hold true for the unEQ'd response as well and measurements confirmed it.

That's why my recommendation was to measure the inherent delay in the subs by comparing the measured physical distance vs the Audyssey detected sub distance while measuring each sub separately. Then, when moving the subs, you simply adjust the distance settings + electronic delay in the avr after each move, and see which placements result in the best pre-EQ results. run the calibration in that position, then if necessary, adjust the avr distance settings to the measured distance + electronic delay already determined, to see if that removes any dips.


Max
post #5464 of 9539
^
To expand on that a bit, there is no direct wave at subwoofer frequencies. It's all swamped by the room. 40Hz is 8.6m. Before your ears had the chance to hear a single period of 40Hz, reflections from the room superimpose the direct sound. Our hearing isn't capable of separating direct sound and room response anymore. It's all the same. The room acts as a filter. What makes matters worse is that each point in space represents a different filter. By the way, that's why single point measurements don't tell the whole story and why single point equalization doesn't work well in most cases.
Each frequency arrives at different times, hence no absolute timing reference exists. If we pick 30Hz as a reference, 40Hz might be delayed. If we pick 40Hz, 30Hz would arrive too early. All we can do is find the best compomise within the crossover region sub/satellites. Time-aligning subs is NOT the best compromise.
Edited by markus767 - 10/17/13 at 4:05am
post #5465 of 9539
Quote:
Originally Posted by markus767 View Post

^
. Each frequency arrives at different times. So there's no timing reference. If we pick 30Hz as a reference, 40Hz might be delayed. If we pick 40Hz, 30Hz would arrive too early. All we can do is find the best compomise within the crossover region sub/satellites. Time-aligning subs is NOT the best compromise.
Are you saying that the speed of sound is frequency dependent?


Max
post #5466 of 9539
^
No, I'm saying that not every frequency from a speaker arrives at the same time. Here's a more detailed explanation: http://bobmccarthy.wordpress.com/2010/02/08/phase-alignment-of-subs-why-i-dont-use-the-impulse-response/
post #5467 of 9539
Quote:
Originally Posted by Sean Spamilton View Post

Ok something funny is going on here. For some reason moving these two subs around the room, I wasn't able to recreate the graphs I got before. I discovered that the subs ( both PC12NSD) that I've got within a month of eachother are giving different readings in the same spot. For some reason the one I just got has a different looking decal on it - and seems to play louder. I looked in the port at the top, and they clearly look designed differently - or have a different driver all together. In fact the first one I got ( I was assuming last years model PC12 as the new one just shipped here to Canada ) - when you look down the port tube it looks like the driver has scratches all over it. For instance:

First PC12 I got last month:

New Sub:



What the hell?

 

I would call or email SVS. I know from personal experience that their CEO, Gary Yacoubian, will move heaven and earth to address customer complaints or issues. They are probably the most customer-focused company I have ever dealt with and I am sure they will address your concerns to your satisfaction.  

 

EDIT: Now I've caught up with all the posts, I see you already did and that you have made good progress. I'll leave the post here in case anyone else with SVS subs might not know just how amazing their customer service is.  Gary Yacoubian actually intervened personally when one of my subs appeared to have got lost between Norway and here (SVS subs in the EU come from Norway) - and this was over the Christmas/New year holiday period too. Not only that, they offered me a substantial discount due to the delay (which turned out to be the shipping company anyway).


Edited by kbarnes701 - 10/17/13 at 4:52am
post #5468 of 9539
Quote:
Originally Posted by EndersShadow View Post

Ok guys, I am a noob here so please bear with me.

I have REW installed on my Windows 7 laptop and am using a oSoundblaster Live! 24-bit USB External soundcard as well as a digital radio shack SPL meter mounted on a tripod in my main listening position.

I thought I had everything installed correctly cord wise, however when I was going through the calibration portion of the setup, I could get the right channel meter to bounce up and down, but was unable to get the left meter to move, even when I changed the dropdown to "left" instead of "right".....

I was basically stuck right at this page of the online guide as it looked different in Windows 7 for me at least: http://www.hometheatershack.com/roomeq/wizardhelpv5/help_en-GB/html/calsoundcard.html#top

Any suggestions on what to do?

I setup my stuff exactly like the diagram below:



I am aware that there is a USB mic out there that allows you to connect it directly to your laptop to get measurements, but I dont want to spend the 90 bucks to get it as I am simply wanting to mess with toe in of my speakers as well as calibrate the phase and level of my sub to integrate it better with my mains.

 

You're in the wrong thread. This thread is specifically for REW with USB mics and HDMI connections.

post #5469 of 9539
Quote:
Originally Posted by AustinJerry View Post
 
Quote:
Originally Posted by kbarnes701 View Post
 

 

Thanks J. Would it help to stuff the mineral wool in the gap that is currently between the corner and the GIK bass trap?  They are currently across the corners at a 45 degree angle, so there is a nice, rwady-made space lurking behind every corner panel. I could easily do that, and it wouldn't be seen either. Worthwhile?

 

Interesting question.  I was under the impression that for a velocity trap, the air space behind the panel was important, which is why we straddle a corner and not place the panel flat on the wall.  Filling the air space would seem to defeat this purpose, IMO.

 

Yes, this was my own thinking too, but then I thought that the purpose of the air gap was to allow the sound to reflect back to the trap and get a 'double' absorption effect. Filling it with actual absorber would surely be better than filling it with air?  I'm only guessing though, which is why I asked J.

 

EDIT: And now answered - thanks J.


Edited by kbarnes701 - 10/17/13 at 4:53am
post #5470 of 9539
Quote:
Originally Posted by AustinJerry View Post
 
Quote:
Originally Posted by jkasanic View Post
 

 

I was discussing what to use in my corners (e.g. rockwool straddling the corner vs. superchunk style corner bass traps) and Sanjay referred me to this post.  Sounds like members over at Gearslutz are using pink fluffy in their superchunk style traps with better results than rockwool or OC703 which previously was considered to be on par with an absorber straddling the corner (creating the air pocket).

 

I have seen this post.  Interesting that there is a door in the right rear corner of the room.  Evidently, he moves two stacked traps into the corner covering the door when he is watching movies, and then moves it back out of the way.  :eek:

 

:)  I used to do that, before Mrs K gave me permission to permanently block  the door, which has now been 'disappeared' by having fabric stretched across the architrave.

post #5471 of 9539
Quote:
Originally Posted by jevansoh View Post
 
Quote:
Originally Posted by kbarnes701 View Post
 
Quote:
Originally Posted by jevansoh View Post

 
 
PS Keith, I wouldn't spend a large amount of money on pre-fab "bass traps" at this point. You have all the panels you need. What I'd do is stuff as much (don't compress it) Safe N Sound equivalent (Rockwool/Mineral Wool) in any corner, weird space, and possibly even build soffits with a fabric covering - it sure would help in making suggestions if I could SEE your room smile.gif ) in open/available areas because that frequency is not too difficult to fix. Anything much below would be, and if you had a larger room my suggestion would be use Pink Fluffy but I'm confident you can both fix this problem and that it will cause audible benefit that is worth the time, trouble, and expense.

Thanks J. Would it help to stuff the mineral wool in the gap that is currently between the corner and the GIK bass trap?  They are currently across the corners at a 45 degree angle, so there is a nice, rwady-made space lurking behind every corner panel. I could easily do that, and it wouldn't be seen either. Worthwhile?

ABSOLUTELY YES!!! Always better than just air gap. Don't compress it though and use medium density (not Rockboard 80 for instance) for that amount of space.

If you do this in ALL corners it may just be enough on its own to fix your issue.

If I could only see pics of your room, I could make more specific recommendations. wink.gif

--J

 

Thanks J. I will have a go at that as it is easy and cheap and won't change the appearance of the room. I am struggling to take pics in the room since it has been painted black and gray. I have some old ones before the painting which I used to help Sanjay help me with sub placement. I will see if I still have them and PM them to you if so.

post #5472 of 9539
Quote:
Originally Posted by jim19611961 View Post
 
Quote:
Originally Posted by Sean Spamilton View Post

Having gain matched the subs again, pretty closely together as evidenced by the graphs above - I've gone ahead and rerun the tests - Taking measurements in the same location but varying the Old and New subs position .... And got the same thing:


The magenta colored plot looks pretty good if they are crossed over before that +8db peak around 75hz.

 

And Audyssey will be effective in bringing that peak down. He's made good progress.

post #5473 of 9539


should i use audyssey or no audyssey according to me graph?
post #5474 of 9539
^
Is the lower curve with XT32? What's the crossover frequency? Could you please post sub+l and sub+r separately?
post #5475 of 9539
Quote:
Originally Posted by markus767 View Post

^
Is the lower curve with XT32? What's the crossover frequency? Could you please post sub+l and sub+r separately?

the bottom is with multi XT (I dont have xt32) the crossover is 100hz.(was the best at getting rid of a hump at the xover frequency) I dont have the l=SUB AND r+SUB with me at work...it will have to wait tioll tonight...I was looking for a quick tip wink.gif
post #5476 of 9539
Quote:
Originally Posted by markus767 View Post

It's probably modal. How does the free field response of the sub look like?

Free field? Is that like the outdoor ground plane measurement? Or a sub measurement free of room interaction?
post #5477 of 9539
^
Free of room interaction. A near field measurement will do.
post #5478 of 9539
Quote:
Originally Posted by djbluemax1 View Post

Yes, it makes a difference.

It has to do with phase cancellation. When the subs are not equidistant to the MLP (especially if the difference is significant), changing the individual delays so the direct wave arrives at the MLP simultaneously ensures that it is in phase. If you don't change the distances and have one sub closer than the other (or move one sub more than the other after running a brief calibration for Audyssey to set the distances), the direct wave could be out of phase, producing what looks like a null due to the cancellation.

If you move the subs around looking for the best positions in the room but don't change the distances/delays, you could end up ruling out, and missing out on placement positions that could produce great results, because not changing the distances produced this phase cancellation. I saw this when I was trying to find the best placements for dual subs. It occurred to me due to the distance tweak, i.e. if inaccurate delay settings for the subs could produce a dip in the bass region that could be fixed post-calibration simply by altering the delays, this must hold true for the unEQ'd response as well and measurements confirmed it.

That's why my recommendation was to measure the inherent delay in the subs by comparing the measured physical distance vs the Audyssey detected sub distance while measuring each sub separately. Then, when moving the subs, you simply adjust the distance settings + electronic delay in the avr after each move, and see which placements result in the best pre-EQ results. run the calibration in that position, then if necessary, adjust the avr distance settings to the measured distance + electronic delay already determined, to see if that removes any dips.


Max

Thanks for your reply, Max. It makes good sense. Perhaps the distance adjustment has not come into play in my situation because of how I configure my four subs. I have two pairs of subs, one pair on each AVR sub channel, and the two subs within a pair are always placed so they are equidistant from the MLP (e.g. the two fronts are at the 1/4 and 3/4 points).

If Sean has read your response, I think the important take-away is to make sure the internal delays of the subs are added to the manually-calculated distances as well. Perhaps that would account for some of the anomalies we see with his measurements.
post #5479 of 9539
Quote:
Originally Posted by jlpowell84 View Post


And the magic! Look at 20-80hz!!!





Much better. It is nice too see the hard work produce a satisfying result.
post #5480 of 9539
Quote:
Originally Posted by markus767 View Post

^
Free of room interaction. A near field measurement will do.

I will do one
post #5481 of 9539
Quote:
Originally Posted by AustinJerry View Post

Much better. It is nice too see the hard work produce a satisfying result.

Thanks! It was surely satisfying, educating, and cool to see science in action. Just the beginning and a fun one.

I'm not really disappointed with the 15hz dip. Surely a lot better there than 25 or 35. I tested out the Olympus has fallen Washington monument scene witch has a thick hot line of bass up 20hz. It was cool but gives a good barometer of heading into the in audible.
post #5482 of 9539
Quote:
Originally Posted by markus767 View Post

That's too far away. The mic needs to be closer than 10% of the diaphragm radius. You also need to make sure the distance is always the same. In the near field of a driver the level changes rapidly with distance. That's probably why we see that 20dB difference.
Other than that the two subs seems to behave the same.

Markus - Those initial measurements were done before I had done a proper gain match, or after I had hit the volume of one of the subs inadvertently during one of my many moves. The actual graphs I got were:



Unfortunately - I spent the rest of the night chasing my own tail. Without doing the " sub tweak " for distance, I'd find a decent response but it would immediately change as soon as Auddessy re measured and re-trimmed during setup. So what was good before became not so good after. Will have to re try today, but as I spent all 12 hours of my day off yesterday attempting to get it right - I don't have as much time or patience available to me at the moment.
post #5483 of 9539
Quote:
Originally Posted by Sean Spamilton View Post

Markus - Those initial measurements were done before I had done a proper gain match, or after I had hit the volume of one of the subs inadvertently during one of my many moves. The actual graphs I got were:

Could you please zoom the dB-axis so it shows only about 50dB and not 240dB?
Quote:
Originally Posted by Sean Spamilton View Post

Unfortunately - I spent the rest of the night chasing my own tail. Without doing the " sub tweak " for distance, I'd find a decent response but it would immediately change as soon as Auddessy re measured and re-trimmed during setup. So what was good before became not so good after. Will have to re try today, but as I spent all 12 hours of my day off yesterday attempting to get it right - I don't have as much time or patience available to me at the moment.

What exactly are you trying to do?

You first have to find sub locations that minimize dips and point-to-point variance. Then EQ the subs as one with MultEQ. When that's done adjust delay of the subs as one to optimize the crossover region to the satellites.
post #5484 of 9539
Quote:
Originally Posted by markus767 View Post

Could you please zoom the dB-axis so it shows only about 50dB and not 240dB?
What exactly are you trying to do?

I will re do the near-field drivers with a smaller ( as you say 50db ) window as soon as I get the chance ( the originals were lost in the near 100 tests I did yesterday ) - perhaps tomorrow before work. Unfortunately I won't have another day off until Sunday, and only have a bit of time to play around before work ( i.e. now ). I did manage to have some success in the 15 - 80 hz region at the MLP today by adjusting the phase on my rear sub to 180. It actually made quite a difference.

Quote:
Originally Posted by markus767 View Post

You first have to find sub locations that minimize dips and point-to-point variance. Then EQ the subs as one with MultEQ. When that's done adjust delay of the subs as one to optimize the crossover region to the satellites.

Right now I'm ( and this might address you're question what am I trying to do ) only focusing on evening out the frequency response for the subs from 15 - 80 hz. I haven't been running tests to 300hz yet incorporating the Front L and Front R - Though I do understand that I can measure the sub response at the same time doing this. Once I get the sub response smoothed out as much as I can by placement and / or phase adjustment, I'll move on to the higher frequencies. At least that's the plan smile.gif I might also just get the best response I can from the subs and call it a day.
Edited by Sean Spamilton - 10/17/13 at 11:43am
post #5485 of 9539
Update: Ok - With the phase switched on the rear sub to get me a better pre - setup graph, I re-ran Aud Setup and finally found my sweet spot - I don't think I'll get better then this through positioning - and given all my trials and tribulations I think this is pretty good to call it a day with respect to the subs positioning no?

post #5486 of 9539
Quote:
Originally Posted by jlpowell84 View Post

This is an eyeball of centered front to back of room, just to see what is there

Just wanted to use your newer measurements to show you (and everyone) how predictions coincide with what's actually happening in the room.


Remember this graphic?



It predicted that your 72Hz mode would be nulling at the quarter points and peaking in the middle of the room.

When you took your earlier measurement 3 feet from the wall, you were pretty close to the null:



Close enough to the null that the sound at 72Hz should have been pretty low.

Which is exactly what your measurements showed:



For this latest round of measurements, you moved your mic to the midpoint of the room:



From the diagram, you would predict that the same frequency would be peaking.

Check out what your measurement showed:



Even in a room that is not a sealed rectangle, merely having a pair of parallel walls can make modal behaviour reasonably predictable.
post #5487 of 9539
I did measure the entire new length front to back Sdurani. It it 23ft 8 inches. I guessed it was 24 smile.gif. And my head at the MLP sits about 9 ft 6 inches from the front wall.
post #5488 of 9539
So with all of your expertise around here...
Would any of you say tweaking the placement of my two corner loaded subs could improve response? Like pulling them out from the wall a little more or up the side wall slightly. I know that graphs will be the proof but...

Right now they are both in the corners 7 inches from each wall...
post #5489 of 9539
Ok, so I took some time off work to play with REW today. Here are a couple things I need some help with.

I cannot get the calibration of the levels (input and output) the same using the Soundblaster live external USB card. I can get it within about 6 db's but thats it. I got a calibration file that seemed decent, but still not like the examples I see on the forum. I then attempted to run a calibration and I kept getting a "level is too low" on the measurements. This was with my AVR at 75 (its MAX) and the soundcard output at max volume.

I dont have any more time today to play with this.

I will have all day on Monday to play with REW as I am home alone with the baby, so whenever she doenst require my attention I will post and hopefully I can get things configured.

I am assuming that once the initial setup is done its fairly easy to take measurements in different rooms?

My plan is to take measurements in my main HT to get my mains and sub integrated right. Then just for giggles I want to look at how bad my 2 channel room is. Its a 11x10 room so I know room modes suck, but I am just curious about things, and if there are any placement things I can do to get it to sound better.

I plan to hopefully get a SMS-1 for my sub, but I have debated if I can get REW to work right getting a mini-dsp to play with integrating my mains and sub together better, since IIRC you can use REW to help you set EQ filters in the mini-dsp....

Anyway I wont ramble further, I am on some heavy pain meds...
Edited by EndersShadow - 10/17/13 at 4:32pm
post #5490 of 9539
ok here are the measurements...with audyssey on and off...i am leaning towards no audyssey...thoughts?



New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Audio theory, Setup and Chat

Gear mentioned in this thread:

AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs