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 194

post #5791 of 9539
Quote:
Originally Posted by markus767 View Post

What difference? The acoustically relevant areas look virtually the same.

^^ True, but the cosmetic change in the graph might lead one to a different conclusion if we're not careful (i.e. not thinking about the practical outcome).
Edited by sdrucker - 10/27/13 at 11:03am

Gear mentioned in this thread:

post #5792 of 9539
^
Not sure I understand. What you're looking at?
post #5793 of 9539
Quote:
Originally Posted by kbarnes701 View Post

Yes, thanks for the encouragement. I am not unhappy with what has been achieved and at least I tried it, which is always worthwhile, even if only so one can say "well I have learned something".
 
Overall, the sound is remarkably good I think, for what is essentially a very difficult space. I never imagined I would be able to achieve such good SQ in this room and the system gives me considerable pleasure.
Many in this hobby, upon being faced with a square room, either give up on good sound or forever use the shape of their room as an excuse for mediocre results. By comparison, you've put in the time, energy and resources to get measurable improvements and excellent results.

So, you're welcome for the encouragement. Just wanted you to understand where it's coming from.
post #5794 of 9539
Quote:
Originally Posted by sdurani View Post
 
Quote:
Originally Posted by kbarnes701 View Post

Yes, thanks for the encouragement. I am not unhappy with what has been achieved and at least I tried it, which is always worthwhile, even if only so one can say "well I have learned something".
 
Overall, the sound is remarkably good I think, for what is essentially a very difficult space. I never imagined I would be able to achieve such good SQ in this room and the system gives me considerable pleasure.
Many in this hobby, upon being faced with a square room, either give up on good sound or forever use the shape of their room as an excuse for mediocre results. By comparison, you've put in the time, energy and resources to get measurable improvements and excellent results.

So, you're welcome for the encouragement. Just wanted you to understand where it's coming from.

 

Thanks, Sanjay - appreciated. And, of course, for your help and advice. That goes for all the other guys who have helped and encouraged me along the way too of course.

post #5795 of 9539
Quote:
Originally Posted by artur9 View Post

There are 3 subs. One is at the end of the sofa, and the other two are at either end of the entertainment center. I want those two close to the mains for when the sub amp gets replaced/repaired.

Placing a sub anywhere along the wall behind the sofa causes a horrible thrumming about as bad as the Incredible Mr. Limpet.
Now there's a reference from my childhood (Mr. Limpet, not 3 subs).

With 2 of the subs at fixed locations near the mains, the only option I can think of is to creep your remaining subwoofer along the walls of your room and/or around the stairwell to see where the interaction of all 3 subs yields the best results. It's a lot of trial & error, combined with meticulous record keeping of which measurement goes with which location, but at least you'll know for sure that you've placed the sub at its best possible location.

BTW, the dip around 40Hz is modal. Divide the speed of sound by that frequency and you get your room dimension (28 ft). While moving your subwoofer around, if it ends up landing at a location of one of the nulls for that mode, the dip will flatten out.

Finally, you might want to consider building as many DIY bass traps as you can accomodate (decor and cost) to see how much it smoothens out the response. It will do something, just don't know how much.
post #5796 of 9539
I am having a slight issue with REW. When I go to open up previous saved measurements they all load on the left vertically like normal but when I click on the the main part of the screen just to the right where you view your graph, change parameters, or generate a waterfall is blank. Even when I click on different ones none of them shoe up in the main part you normally view in. I am on a Mac and just did a java update that was needed for REW but it has worked after the update and this just as of last night when I tried to view previous measurements. mad.gif
post #5797 of 9539
Quote:
Originally Posted by jlpowell84 View Post

I am having a slight issue with REW. When I go to open up previous saved measurements they all load on the left vertically like normal but when I click on the the main part of the screen just to the right where you view your graph, change parameters, or generate a waterfall is blank. Even when I click on different ones none of them shoe up in the main part you normally view in. I am on a Mac and just did a java update that was needed for REW but it has worked after the update and this just as of last night when I tried to view previous measurements. mad.gif

 

I use a Mac (for viewing and messing with the data files - Windows for taking measurements (don;t ask). Can you do a screen cap (CMD-Shift-4) and upload the image in your reply. I am not sure what you mean. I am on Mavericks OS-X. Also did java update yesterday.

 

This is what I am seeing:

 

post #5798 of 9539
Sdurani, or anyone is welcomed, I am looking for advice on speaker placement. I have a near exact equilateral triangle between LR and MLP. I I'll provide a graph in a few. Big dip at 120hz. Although the 80hz crossover looks great. The JBL towers were set to full by audyssey so the 8" woofers are really capable as I wasn't sure the Submersives were on. Well they were not as "stereo" listening preset had the towers set to full and subs off. Subs on obviously do give it a deeper better impact but they can do full range. Very decent speakers overall. So now I am trying get speaker placement honed in.









Those last two are directly from the MLP. I have 48 sheets of 2x24x48 Roxul AFB, black micro suede fabric and I will buy wood soon and make a plethora of treatments. I will experiment with just setting raw peices all over before to get thicknesses right as I don't want to waste or use too little. Yes Sdurani I will make custom panels for the ends of the kitchen cabinet and and counter smile.gif
post #5799 of 9539
Quote:
Originally Posted by kbarnes701 View Post

I use a Mac (for viewing and messing with the data files - Windows for taking measurements (don;t ask). Can you do a screen cap (CMD-Shift-4) and upload the image in your reply. I am not sure what you mean. I am on Mavericks OS-X. Also did java update yesterday.

This is what I am seeing:



I get this when I click on any graph on the left. I too just updated...
post #5800 of 9539
Can you clarify--are the JBL's still set to large, or are they set to small with an 80Hz crossover? You say you have a problem, but you didn't post any REW graphs. Are you going to?

Edit: Oops, if you are having issues with the MAC, then obviously you can't post graphs, sorry.
post #5801 of 9539
Keith it just worked for some reason. I closed REW then reopened. It wasn't until the third time though that they would show. Thanks for responding:)
post #5802 of 9539
Quote:
Originally Posted by AustinJerry View Post

Can you clarify--are the JBL's still set to large, or are they set to small with an 80Hz crossover? You say you have a problem, but you didn't post any REW graphs. Are you going to?

Edit: Oops, if you are having issues with the MAC, then obviously you can't post graphs, sorry.

Yes Jerry I am getting to that. I just wanted to get physical pics of visual placement first. I have to remeasure quickly as the files I didn't save unknowingly a couple nights ago. Graphs in 30...
post #5803 of 9539
Quote:
Originally Posted by jlpowell84 View Post

Yes Jerry I am getting to that. I just wanted to get physical pics of visual placement first. I have to remeasure quickly as the files I didn't save unknowingly a couple nights ago. Graphs in 30...

The JBLs are not set to large, i just mentioned that Audyssey did so we would know their capabilities. They are now at small with 80hz crossover...
Only 2 channel for awhile for me. I could if I wanted use one of my Definitive Technology pro monitor 1000's as a center. I was using them as LR awaiting the jbl rebuilt drivers.
post #5804 of 9539
Quote:
Originally Posted by jlpowell84 View Post

I get this when I click on any graph on the left. I too just updated...

Happens when you double-click a .mdat to open REW. First open REW then open the .mdat.
post #5805 of 9539
Quote:
Originally Posted by sdurani View Post

With 2 of the subs at fixed locations near the mains,

They are fixed for now. The TBIs, being passive, are only 8" high so there are some options for placing them under things. Once the sub amp comes back I can look into moving them if necessary. Just need to buy hundreds of feet of speaker wire smile.gif

The crossover I ended up at, after a bunch of trial and measure, is 90Hz. Unfortunately, just above being locatable sonically.
Quote:
the only option I can think of is to creep your remaining subwoofer along the walls of your room and/or around the stairwell to see where the interaction of all 3 subs yields the best results. It's a lot of trial & error, combined with meticulous record keeping of which measurement goes with which location, but at least you'll know for sure that you've placed the sub at its best possible location.

Yeah. One of the best places I've found previously was next to the stove. But with the sympatico between it and the sub I moved it to its current location.
Quote:
BTW, the dip around 40Hz is modal. Divide the speed of sound by that frequency and you get your room dimension (28 ft). While moving your subwoofer around, if it ends up landing at a location of one of the nulls for that mode, the dip will flatten out.

That's good to know. I need to find that null to put the sub into.

What do you think about putting in the closet? Alternatively, above the Outlaw there is a built-in shelf about 56" high that is big enough to hold the sub. Is that a viable location?

Quote:
Finally, you might want to consider building as many DIY bass traps as you can accomodate (decor and cost) to see how much it smoothens out the response. It will do something, just don't know how much.

Yeah, looks like the next area of research for me. How to make them and where to put them.
post #5806 of 9539
Quote:
Originally Posted by jlpowell84 View Post

I have 48 sheets of 2x24x48 Roxul AFB, black micro suede fabric and I will buy wood soon and make a plethora of treatments. I will experiment with just setting raw peices all over before to get thicknesses right as I don't want to waste or use too little.
Good to hear. But will you remember to make custom panels for the ends of the kitchen cabinet and and counter?
Quote:
Originally Posted by jlpowell84 View Post

Yes Sdurani I will make custom panels for the ends of the kitchen cabinet and and counter smile.gif
How did you know I was going to ask that?
post #5807 of 9539
Quote:
Originally Posted by artur9 View Post

What do you think about putting in the closet? Alternatively, above the Outlaw there is a built-in shelf about 56" high that is big enough to hold the sub. Is that a viable location?
Putting the sub in an open closet or on a shelf is fine IF it improves the sound.
post #5808 of 9539
the path of progression.















One note is these speakers have a +-2db high end switch on the back...looks good for untreated i think!!!biggrin.gif

post #5809 of 9539
Quote:
Originally Posted by AustinJerry View Post

An interesting and challenging room!

Regarding the big dip at 7K, this could be caused by interaction between your left and right speakers, especially if the mic is not exactly centered between the two.  Centering the mic is difficult, and you can get big swings in the higher frequencies.  Two suggestions:

1.  Measure Left+subs, and then Right+subs.  If the dip does not appear in these two graphs, then the dip in the combined graph is probably due to interaction.
2.  Place the mic in what you think is the center position, run a measurement of Left+Right+subs, move the mic a little to the right (only about an inch!), repeat the measurement, then move it to the left a bit, re-measure, etc.  Observe whether the dip at 7K changes when the mic is moved.  This is another indication that the dip is caused by interaction.

Here are the results of my making adjustments and measuring. When I measured the speakers individually they both had dips as you can see below.
  • First, since it was the easiest, I tried the Cinema 11's limited EQ.
  • Then I tried moving the speaker away from the wall.
  • Then I took it down from its stand.
  • Then I tilted it towards the MLP.
  • I took the grill off.

Then I rememebered reading somewhere that some speakers sound better upside down so I tried that. That fixed it.

Here are the before and afters left only.

Before:


After


Both on the same graph.


Finally, I moved the mic an inch to the right.


So, score one for measurements! smile.gif
post #5810 of 9539
Man you guys are a tough crowd...I had a recommendation elsewhere my 120hz dip could be a ceiling bounce issue. You think?
Actually looking at the entire FR with 1/6 smoothing it looks like 150hz bulge rather. Another important side note is the 80hz area. It didn't matter if I cut 8db or 16db there it didn't budge down any at all. So I must have a natural room boost in that area. And of course I left that 50hz area alone rather than trying to boost and pump energy into it. I watched how to train your dragon demo scene and the bass is very very nice. I can tell where a little tightening up will help with treatments. Overall the sound is very nice. Running only 2 channel currently though leaves a definitive sweet spot right in the middle
Edited by jlpowell84 - 10/27/13 at 9:36pm
post #5811 of 9539
Quote:
Originally Posted by markus767 View Post

^
Not sure I understand. What you're looking at?

I was thinking of the absolute level of decline on Keith's two "after" waterfalls. One chart, the purple one, hadn't finished decaying at the graph floor of 45 db in the 50-60 Hz range, while the second one, the one in green at a slightly lower db SPL, has decayed by the 50 db mark in that 50-60 Hz region. The assumption in this context is that 45 db is the noise floor for this purpose. Could you clarify what you mean by it having the same acoustic interpretation so I know how to interpret the waterfall without drawing what you may call a meaningless conclusion, even if it's driven by the absolute SPL level measured, that a decay to 50 db and a decay to 45 db represent a different acoustic result? If I'm interpreting the graphs incorrectly that way, please explain for us caveman laypeople so we know not to be led astray..smile.gif...I mean, the SPL level on the lower graph in green isn't exactly too soft...

Quote:
Originally Posted by kbarnes701 View Post

This is the waterfall after adding the additional absorption:




There is some improvement but not where I expected it to be! I was trying to get rid of the problem at about 50-60Hz yet that seems pretty much unchanged. Comments welcome.

Showing the importance of the measuring SPL, here is an 'after' waterfall made from a measurement taken at slightly lower level:




Big 'graphed' difference.

Edited by sdrucker - 10/27/13 at 11:33pm
post #5812 of 9539
^
The areas of interest in such a waterfall are a) where does the ringing begin in relation to the direct sound and b) how fast does it decay. Looking at the lower end of the graph is misleading because it's most likely already showing artefacts from the noise floor. Each slice of a waterfall is based on data that spans over several hundreds of milliseconds (the window length).
post #5813 of 9539
Here's an example:




The only difference between the measurements is that the second measurement was taken 20 seconds after the first. Nothing else was changed.
post #5814 of 9539
Sorry Markus but your posts tend to be somewhat cryptic to the lay man. Can you explain the reason for the difference? 20 seconds seems like a long time for one measurement to effect the other but is that what you are implying?
post #5815 of 9539
I think what is being pointed out is that the measurement apparatus (software, mic, ect...) will vary in minor ways even when the setup is exactly the same where waterfalls are concerned. In this case, minor variations after or -30db shouldn't be taken too seriously.
post #5816 of 9539
Quote:
Originally Posted by markus767 View Post

^
The areas of interest in such a waterfall are a) where does the ringing begin in relation to the direct sound and b) how fast does it decay. Looking at the lower end of the graph is misleading because it's most likely already showing artefacts from the noise floor. Each slice of a waterfall is based on data that spans over several hundreds of milliseconds (the window length).

Markus,
I had thought that the two graphs I copied from Keith's OP were different in their implication at 45-50 db because of the following. On page 61 of Jerry's REW 3.0 guide, he notes that for his sample waterfall:
"Note where the graph extends out and touches the bottom scale (58Hz and 77Hz in the example above). This means that bass notes at those frequencies have not fully decayed below 50dB within the specified time window (300ms in the example). This is undesirable “bass ringing”.

This is the sample waterfall I'm referring to (from Jerry's v3.0 guide):


On Keith's graph that I cited in my post #5811, one chart shows an area at 50-60 Hz that hasn't finished decaying to 45 db at 450 ms, and the other shows that it decayed by 50 db at the same 450 ms level. Literally that was the way I was indicating the graphs were different, and that they could draw different conclusions as a result, even the difference in measuring technique is subtle. That assumes that 45 db rather than 50 db was the standard for the bottom of the noise floor, and that the charts showed the "bottom" of the reliably measurable noise floor.

But if you're indicating that "looking at the lower end of the graph is misleading", are you saying that the conclusion I drew from the Guide, and the advice there, is incorrect, or that there's a more valid way to interpret the decay on waterfall than Jerry's excerpt describes? Or is there an esoteric point I'm missing?

Sorry to beat this to death, but if I'm interpreting waterfalls incorrectly based on reading the REW guide (which I thought was quite clear on this point), we should clarify things so that none of us is prone to making the same mistake. Or we need to change the guide LOL....
Edited by sdrucker - 10/28/13 at 10:47am
post #5817 of 9539
Quote:
Originally Posted by sdrucker View Post



Are you saying that the conclusion in the Guide is incorrect, or that there's a more valid way to interpret the decay on waterfall than this excerpt describes?

 

Well, my understanding of waterfall graphs is tenuous at best, so I would welcome a collective effort to improve our overall understanding on how to generate and interpret these graphs.

post #5818 of 9539
Quote:
Originally Posted by AustinJerry View Post

Well, my understanding of waterfall graphs is tenuous at best, so I would welcome a collective effort to improve our overall understanding on how to generate and interpret these graphs.

DK about you, but what I thought was "clear" is a heck of a lot less so now....

OTOH, what I think I/we may be missing is that it's less about not reaching the bottom of the noise floor, as the delta, i.e. having at least a 40-45 db gap between the TOP of the waterfall in a particular region and where the bottom of the waterfall is. You have to measure loud enough, as per Jason (85db bare minimum, preferably 90-95db) to accurately capture bass decay in the 350/450/600 ms timeframe.

So from that POV, Keith's two charts are showing the same information, but you have to look at the charts VERY carefully to know that. On the purple chart with the higher SPL, in the 50-55ish Hz range, his bass decay hasn't fallen at 40 db (or more) from the top of the waterfall - a gap of slightly more than 90 db (92 db?) to below 50 db. OTOH, on the green chart with the slightly lower SPL, in the 50-60 Hz range, his bass decay either just makes or just misses a 40 db rate of decay: a gap of around 87-88 db to about 50, where you just "miss" at about 50-55 Hz and just "make it" at about 55-60 Hz. All of this @ 450 ms, which is IIRC the level of rigor we should aim at if we can't go for the "gold standard" of 300 ms.

Basically it's a similar conclusion, but how we got there was two different ways: one that you have as written in the Guide, the other an interpretation of what I think Markus and Jason said. Markus' point, possibly, is that we don't look at the touching the bottom of the graph with the graph's bottom being the theoretical noise floor as the range from the top to bottom of the waterfall, because we can't measure reliability without incurring noise at the bottom of the noise floor, given the 50-55 db observed noise floor levels most of us have. In that context, a 45 db for the bottom of the chart is as relevant as having an top of the db scale at 105 db, giving us "headroom" in how we read the graph. Nothing more.

The graph's bottom origin is besides the point as long as we capture at least 40-45 db from the top of the waterfall. Strictly a matter of "zoom", as we shouldn't talk about measuring to the bottom of the "noise floor" so much as capturing the decay in that 40-45 db from a waterfall top in the 20-300 Hz range (15-300 Hz for UK-based Seaton Submersive owners).

So basically we're drawing the right conclusion, but maybe we should change the Guide language to reflect how we got there better than saying we touch the bottom of the scale to avoid confusion, or making possibly erroneous conclusions based on a graph origin.

Does that make any sense or is this another few pegs down the rabbit hole?
Edited by sdrucker - 10/28/13 at 11:40am
post #5819 of 9539

OK, here are four waterfalls, all the same underlying measurement, with the vertical axis lower limit varied by 5dB, from 60db down to 45dB.  Focus on the 40-50Hz range.

 

 

 

 

 

 

 

As the lower limit is varied, the 40-50Hz area changes considerably, from a non-issue to a serious issue.

 

So, what are the waterfalls telling us?  AFAIK, the key is the noise floor.  If the noise floor were 60dB (unrealistically high), then the first waterfall tells us that we won't hear the ringing in the 40-50Hz range beyond ~400ms.  If the noise floor were 50dB (more reasonable), then the ringing in the 40-50Hz range is still ~10dB above the noise floor at 450ms, i.e. still clearly audible.  If the noise floor is 45dB, then the ringing in the 40-50Hz range is still ~15dB above the noise floor, very audible (and pretty bad).

 

Someone please tell me if I am on track, or totally off course.

post #5820 of 9539

BTW, and update on my ceiling fan, and the reflections issue it has been causing.

 

I replaced the fan with a Big Ass Fan.  Yes, you read that correctly.  Here is the web site:  http://www.bigassfans.com/

 

The model I purchased is the "Haiku".  It differs from the previous fan in that it has three blades (the old fan had five blades).  This means that there is a stationary position for the fan blades that causes no reflection at the MLP.  And, as I demonstrated earlier, when the fan blades are rotating, REW ETC measures no reflections.

 

Not a perfect solution, but a step in the right direction.  And I can say now that I have a Big Ass Fan (priceless!), and that when the fan blades are rotating, they make a sound like "Twerk, Twerk, Twerk...".   :D

 

 

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