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 95

post #2821 of 9564
Quote:
Originally Posted by Pres2play View Post

I haven't given 1/4 placement much consideration. I have that darn radiator blocking one corner of the front wall and on the other side there's a worm hole - no sub works there. That's where I had the twin ULS and the reason I decided to send it back. The equipment rack is there now. I'll try placing the sub in front of the radiator temporarily and see what kind of response I get.

Placing the sub higher is worth trying.

I have a question regarding DEQ. In the Onkyo 818, it seems to be on during calibration. If I turn it off, the new measurements show a drop in the low end. I'll put this question in the 818 thread.


Just tried 1/4 distance and FR is not better. In fact, the dips get deeper the closer I get to the wall. I'm going to try moving my entire setup and centering the sub on the front wall. Get back to you tomorrow.

This graph shows the sub at 1/4 distance (red), and two spots closer.

Gear mentioned in this thread:

post #2822 of 9564
Quote:
Originally Posted by Pres2play View Post

Just tried 1/4 distance and FR is not better.
Both subs or a single sub?
post #2823 of 9564
Two subs. I didn't move the sub on the side wall. It's at the half way mark, about 20 inches off the floor.

BTW, I did a factory reset on my Onkyo 818, so did not have a chance to run Audyssey.
post #2824 of 9564
Quote:
Originally Posted by Pres2play View Post

I didn't move the sub on the side wall. It's at the half way mark, about 20 inches off the floor.
I should have clarified what I meant by placement at quarter points. The middle of each subwoofer cone needs to be 55.5 inches from the side walls, which are the 1/4 and 3/4 divisions of room width. This can be done on the front wall or back wall (if the radiator prevents you from doing it on the front wall).
post #2825 of 9564
Quote:
Originally Posted by giomania View Post


Jerry,

Have you had a look at Nyal's article (first post below) and the full paper it is drawn from (second post below)?

It seems to me his methodology will help with your analysis.

Mark
 

 

Yes, Mark, I have read both articles.  Time to re-read them, as I'm sure my comprehension of the principles will be better the second time around.

post #2826 of 9564
I just purchased a UMIC-1 microphone from the CSL link. I am looking for the most accurate way to measure response, do I need hdmi output or would analog rca out or spdif output from my soundcard work just as well?

I have a delta audiophile 24/96 soundcard with dual rca analog output and input jacks. It also can output coax spdif. Since, I have read that there is no way to calibrate soundcards when using usb mics I was leaning towrads purchasing a hdmi video card. So should I juist purchase a hdmi video card and use hdmi out or is it just as good to use the spdif or rca analog coax output on the audiophile? Also, The microphone cable will be connected with a 32ft active usb extension cable from the pc.
post #2827 of 9564
Quote:
Originally Posted by proudx View Post

I just purchased a UMIC-1 microphone from the CSL link. I am looking for the most accurate way to measure response, do I need hdmi output or would analog rca out or spdif output from my soundcard work just as well?

I have a delta audiophile 24/96 soundcard with dual rca analog output and input jacks. It also can output coax spdif. Since, I have read that there is no way to calibrate soundcards when using usb mics I was leaning towrads purchasing a hdmi video card. So should I juist purchase a hdmi video card and use hdmi out or is it just as good to use the spdif or rca analog coax output on the audiophile? Also, The microphone cable will be connected with a 32ft active usb extension cable from the pc.

 

I take it you are using a desktop PC in another room, rather than a laptop?

 

I would try using the headphone output connection from your soundcard first.  In my opinion, this is a perfectly acceptable method of connecting.  Read the section in the Guide titled "Cabling Basics...".

post #2828 of 9564
Quote:
Originally Posted by AustinJerry View Post

I take it you are using a desktop PC in another room, rather than a laptop?

I would try using the headphone output connection from your soundcard first.  In my opinion, this is a perfectly acceptable method of connecting.  Read the section in the Guide titled "Cabling Basics...".

yeah its a desktop in another room. When I add up the cost of buying the USB extension cable, getting a hdmi video card, etc I started questioning if I should have just bought the galaxy CM-140 Verified+ meter from CSL instead of the umik-1.and run a 32ft coax rca cable back to the analog input on my soundcard
post #2829 of 9564
Quote:
Originally Posted by sdurani View Post

I should have clarified what I meant by placement at quarter points. The middle of each subwoofer cone needs to be 55.5 inches from the side walls, which are the 1/4 and 3/4 divisions of room width. This can be done on the front wall or back wall (if the radiator prevents you from doing it on the front wall).

I placed the second sub beside the front right speaker, a 1/4 distance from the sidewall. The response is better. Just some narrow ringing at 60 Hz. I will continue moving things until it's all dialed in. Thanks for your help!

Green= FR with sub relocated (Audyssey and Antimode enabled)
Red= FR with sub on the side wall (no correction)



Ringing at 60 Hz with 300 ms window



Ringing still there at 600 ms

post #2830 of 9564
Quote:
Originally Posted by Pres2play View Post

Ringing at 60 Hz with 300 ms window

Probably hum from the power line. Are you living in the US?

By the way, 1/6 smoothing is way too coarse for proper low frequency analysis. Switch to 1/48 instead.
post #2831 of 9564
Quote:
Originally Posted by Pres2play View Post


I placed the second sub beside the front right speaker, a 1/4 distance from the sidewall. The response is better.



 

 

First of all, the response represented by the green line is significantly better than the one represented by the red line--congratulations for that!  And the waterfall graphs at 300ms look pretty good.  Having said that, I would like to make several recommendations with regards to the presentation of the measurements, and the results:

 

- Always present both a "before" and "after" graph that isolates only the change you are measuring.  For example, you show Red as the "Before" measurement, with no correction, and the Green "after" graph with correction.  To assess the benefit of moving the sub into the 1/4 position, your "after" graph should be without correction as well.  Otherwise, how can we as observers tell whether the improvement is the new location, or simply the correction? 

 

- It is a good idea to enter a description for each measurement so you, and we as the interested observers, can tell what was measured.  "May 13 at 14:47" won't allow you to remember what was measured if you return to these measurements later.

 

- I agree with Markus' comment--low frequency measurements should always be presented with no smoothing.  I recommend presenting two graphs, 15-300Hz with no smoothing, and 15-20,000Hz with 1/6 smoothing (same measurement, just different presentation).

 

- I also agree with Markus regarding the narrow ringing at 60Hz.  This is likely to be caused by electrical noise somewhere in your system.  You should experiment with grounding your components. 

 

So, returning to the Green measurement, where are the two subs positioned for this measurement?  Are they both on the front wall at 1/4 and 3/4 points, or is the second sub somewhere else?  Sorry for all the suggestions, but you are the only thread participant who is making and posting measurements right now.

post #2832 of 9564
Quote:
Originally Posted by AustinJerry View Post

First of all, the response represented by the green line is significantly better than the one represented by the red line--congratulations for that!  And the waterfall graphs at 300ms look pretty good.  Having said that, I would like to make several recommendations with regards to the presentation of the measurements, and the results:

- Always present both a "before" and "after" graph that isolates only the change you are measuring.  For example, you show Red as the "Before" measurement, with no correction, and the Green "after" graph with correction.  To assess the benefit of moving the sub into the 1/4 position, your "after" graph should be without correction as well.  Otherwise, how can we as observers tell whether the improvement is the new location, or simply the correction? 

- It is a good idea to enter a description for each measurement so you, and we as the interested observers, can tell what was measured.  "May 13 at 14:47" won't allow you to remember what was measured if you return to these measurements later.

- I agree with Markus' comment--low frequency measurements should always be presented with no smoothing.  I recommend presenting two graphs, 15-300Hz with no smoothing, and 15-20,000Hz with 1/6 smoothing (same measurement, just different presentation).

- I also agree with Markus regarding the narrow ringing at 60Hz.  This is likely to be caused by electrical noise somewhere in your system.  You should experiment with grounding your components. 

So, returning to the Green measurement, where are the two subs positioned for this measurement?  Are they both on the front wall at 1/4 and 3/4 points, or is the second sub somewhere else?  Sorry for all the suggestions, but you are the only thread participant who is making and posting measurements right now.

Thanks, Jerry. Your right, of course. I should've posted both measurements without correction in order to properly compare the subs in the two separate locations. I will make new measurements soon, probably Monday at the earliest, when the apartment is empty.

I will include a description with each measurement. Thanks for reminding me tongue.gif

Now that you mention the possibility of electrical noise as the cause of the 60 Hz ringing, I have had issues before with ground-loops. I traced it to the cable box and fixed it with an isolator. What I also noticed, however, is that interconnect cables can also be affected by their proximity to power cords. Just moving the power cords away can help a lot here. I will check and sort out the cables so that this is not happening. I didn't know REW can pick this up.

Both subs are on the front wall now at 1/4 and 3/4 distance. Having a sub on the sidewall provides added boost at the very low end, but also causes spikes in the upper bass, so I moved it to the front wall. I'm learning wink.gif
post #2833 of 9564

I ran across some measurements that demonstrate my journey seeking good bass response in my listening room, and which also show the value of being able to use measurements to help the journey along.  Note also that I followed Keith's advice:  you can never have too much subwoofage....

 

First, two subs on the front wall, showing a poor placement that exacerbates a room mode at 58Hz (as determined by the Room Mode Calculator tool).

 

 

First improvement:  move the subs to the 1/4 and 3/4 spots on the front wall.  Notice how this movement tamed the peak at 58Hz (new location in blue).

 

 

Next improvement:  add two additional subs, placed on the back wall (front subs in blue, back subs in red).  Notice the difference in responses below 35Hz, and above 70Hz.

 

 

Looking at the combined response of all four subs, and how the combined responses below 35Hz and above 70Hz are now improved.

 

 

Add Audyssey room correction:

 

 

Now tweak the sub distances (only marginal improvement on my system):

 

 

So, here is a summary of the journey, showing the starting and ending response:

 

 

 

 

 

 

 

post #2834 of 9564
Jerry,

A fantastic voyage! I'm curious. What changes gave the improvement from #3 to #4?
post #2835 of 9564
Quote:
Originally Posted by Roger Dressler View Post

Jerry,

A fantastic voyage! I'm curious. What changes gave the improvement from #3 to #4?

 

Thanks, Roger.  I looked at my notes, and the only difference between #3 and #4 is that #4 represents the combined response with all four subs being driven, with Audyssey off.

post #2836 of 9564
Quote:
Originally Posted by AustinJerry View Post

Thanks, Roger.  I looked at my notes, and the only difference between #3 and #4 is that #4 represents the combined response with all four subs being driven, with Audyssey off.
Ok. Understood. And you did the delay adjustments after the EQ was applied? Or did you also try delays (and level?) adjustments prior to adding EQ? Not nitpicking, but another fellow on another forum is struggling with aligning a sub and your insights are helping me understand. Thanks!
post #2837 of 9564
There really is no simple way to approach this other than measure, tweak, measure, tweak, ...
Level, delay, location, frequency response, all of these parameters will affect the response at the listening location.

REW b17 offers a sub simulator for up to 4 subs. Playing around with that tool can give insight how some of these factors interact but in real rooms only trial and error will produce really good results.

People also need to look at locations around the listening position. Low frequency response can vary dramatically within just a few inches.

Even if one has managed to get a smooth frequency response, there's another important factor for good low frequency reproduction: modal decay. This can be thought of as "reverberation time" at low frequencies (although there really is no "reverberation time" in acoustically small rooms). If it's too long, low frequency reproduction will suffer even if the frequency response is flat.
Modal decay can be visualized with REWs "Waterfall" and "Spectrogram" tools.
Controlling modal decay is a bit harder. One needs to apply either passive or active absorption. Another simple approach is to put a sub very close to the listening position (<2'). This way the direct sound will dominate over room effects.
post #2838 of 9564
Quote:
Originally Posted by Roger Dressler View Post


Ok. Understood. And you did the delay adjustments after the EQ was applied? Or did you also try delays (and level?) adjustments prior to adding EQ? Not nitpicking, but another fellow on another forum is struggling with aligning a sub and your insights are helping me understand. Thanks!

 

I conduct the bass distance adjustments playing a 15-300Hz tone from the center speaker with all four subs on (REW mono tone fed to both channels in PLII Cinema mode), after running room EQ, with Audyssey engaged.  I then check and adjust the results with the tone coming from the left and right speakers (stereo mode).  There is a detailed write-up on the process, reflecting the efforts by Mark Seaton and Craig John, in the Audyssey FAQ.

post #2839 of 9564
Quote:
Originally Posted by markus767 View Post

There really is no simple way to approach this other than measure, tweak, measure, tweak, ...
Level, delay, location, frequency response, all of these parameters will affect the response at the listening location.

REW b17 offers a sub simulator for up to 4 subs. Playing around with that tool can give insight how some of these factors interact but in real rooms only trial and error will produce really good results.

People also need to look at locations around the listening position. Low frequency response can vary dramatically within just a few inches.

Even if one has managed to get a smooth frequency response, there's another important factor for good low frequency reproduction: modal decay. This can be thought of as "reverberation time" at low frequencies (although there really is no "reverberation time" in acoustically small rooms). If it's too long, low frequency reproduction will suffer even if the frequency response is flat.
Modal decay can be visualized with REWs "Waterfall" and "Spectrogram" tools.
Controlling modal decay is a bit harder. One needs to apply either passive or active absorption. Another simple approach is to put a sub very close to the listening position (<2'). This way the direct sound will dominate over room effects.

 

Of course, Markus, the decay is critical.  Here are a few graphs showing the modal ringing.  I currently have eleven 4'x2' velocity bass traps installed in my 16'x20'x10' room (a combination of RealTraps MiniTraps, and GIK Acoustics 244's).

 

 

 

 

Edit:  I should also note that you have been very helpful with providing advice over the last year or so, for which I am very grateful!

post #2840 of 9564
Quote:
Originally Posted by AustinJerry View Post

I conduct the bass distance adjustments playing a 15-300Hz tone from the center speaker with all four subs on (REW mono tone fed to both channels in PLII Cinema mode), after running room EQ, with Audyssey engaged.  I then check and adjust the results with the tone coming from the left and right speakers (stereo mode).  There is a detailed write-up on the process, reflecting the efforts by Mark Seaton and Craig John, in the Audyssey FAQ.

Can you post the URL? I looked here and didn't find it: http://www.audyssey.com/audio-technology/faq
post #2841 of 9564
Quote:
Originally Posted by mark_anderson_u View Post

Quote:
Originally Posted by AustinJerry View Post

I conduct the bass distance adjustments playing a 15-300Hz tone from the center speaker with all four subs on (REW mono tone fed to both channels in PLII Cinema mode), after running room EQ, with Audyssey engaged.  I then check and adjust the results with the tone coming from the left and right speakers (stereo mode).  There is a detailed write-up on the process, reflecting the efforts by Mark Seaton and Craig John, in the Audyssey FAQ.

Can you post the URL? I looked here and didn't find it: http://www.audyssey.com/audio-technology/faq

 

Jerry is referring to our own AVS Audyssey FAQ.

 

This FAQ answer has details on gain-matching, which started this thread and, in the Further Reading link, it enables you to download a comprehensive guide to the sub distance tweak procedure.

 

f)8.    How does Audyssey handle complex multiple subwoofer setups?

post #2842 of 9564
Thank you Jerry. I'm always happy to help when people actually listen wink.gif

What's the window in your graphs above?

By the way, here's the near field sub I was talking about in my last post:

post #2843 of 9564
Quote:
Originally Posted by markus767 View Post

Thank you Jerry. I'm always happy to help when people actually listen wink.gif

What's the window in your graphs above?

By the way, here's the near field sub I was talking about in my last post:

 

Wow! 

post #2844 of 9564




Jerry, here are mine done roughly in the same scales as yours. Thought you might be interested in comparative data.
post #2845 of 9564
Quote:
Originally Posted by kbarnes701 View Post

Jerry is referring to our own AVS Audyssey FAQ.

This FAQ answer has details on gain-matching, which started this thread and, in the Further Reading link, it enables you to download a comprehensive guide to the sub distance tweak procedure.

f)8.    How does Audyssey handle complex multiple subwoofer setups?

Thanks!
post #2846 of 9564
Quote:
Originally Posted by jim19611961 View Post


Jerry, here are mine done roughly in the same scales as yours. Thought you might be interested in comparative data.

 

Jim, the lower limit of the vertical axis should be the noise floor of your listening room,  The typical listening room has a noise floor of ~40dB.  The purpose of the Waterfall is to measure the speed at which bass frequency resonances decay to below the noise floor, and the "target" is to have all frequencies (at least those >40Hz) decay below the noise floor by 300ms.  Here is what my Waterfall looks like with a 75dB lower limit--it doesn't show any valuable information.

 

 

 

 

 

For the Spectral Decay Graph, the "target" is for all resonances shown by the 160ms line to be -20dB or more.  Your graph looks pretty good, actually.

 

Try re-doing the Waterfall and post your new graph.

post #2847 of 9564


Not sure this is more useful, but here you are. I never been able to get clean waterfalls below -35db or so. Must be my setup or something.
post #2848 of 9564
Something doesn't look right. Try one more time: set the lower limit to 50dB, and increase the time window to at least 600ms. We are looking for the modal ringing to decay below 50dB. Keep increasing the time window until the waterfalls end at the bottom of the graph. Unless there is something wrong with the measurement, there is a serious problem.
post #2849 of 9564
Quote:
Originally Posted by markus767 View Post

There really is no simple way to approach this other than measure, tweak, measure, tweak, ...
Level, delay, location, frequency response, all of these parameters will affect the response at the listening location.
Quote:
Originally Posted by AustinJerry View Post

I conduct the bass distance adjustments playing a 15-300Hz tone...

Thank you, Jerry and Markus, for the great info.
post #2850 of 9564
Quote:
Originally Posted by jim19611961 View Post

I never been able to get clean waterfalls below -35db or so. Must be my setup or something.

Typically only the first 40dB are useful in a waterfall graph, the rest is most often just noise. Always use REW's RTA to measure the noise floor. Helps a lot in interpreting the data correctly.
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