Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs - Page 787 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 2287Likes
Reply
 
Thread Tools
post #23581 of 23608 Old 10-05-2017, 09:27 AM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Quote:
Originally Posted by jevansoh View Post
While I'm here, I thought I'd ask...

Does anyone else have problems getting HDMI working with ASIO4ALL and Nvidia cards?

I've built two different systems recently with Nvidia cards and cannot get REW working with HDMI.

AMD cards and Intel on board video/audio work fine every time.

Is this a known problem or am I just unlucky?

Thanks,

--J
Hello,
I am way off an expert. I remember that I read somewhere (unfortunatly do not remember where) that not all Nvidia cards works with hdmi and Rew. Nvidia Quadro K600 works fine for me.

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall
FargateOne is offline  
Sponsored Links
Advertisement
 
post #23582 of 23608 Old 10-06-2017, 08:22 PM
AVS Forum Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 3,291
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 888 Post(s)
Liked: 600
Hi guys,
Been a long time....but I still lurk here. I've been happily enjoying my 11.4.6 Trinnov Altitude system that Curt Hoyt and I calibrated, until I decided to do some measuring to try to tweak the subs for solving some niggling issues outside of Trinnov goodness. One thing led to another, so....

Just downloaded the latest REW build (version 5.19 beta 7, running under Windows 10, ASIO4ALL, etc.) and I have a few questions:
a) For comparing full-range speaker frequency responses, is the norm still to do 1/6th smoothing, or are the newer methods like variable smoothing, psychoacoustic smoothing or ERB getting some favor? Wondering because I did a search on the thread and I've read some comments suggesting that for determining EQ, you might want to use variable smoothing as it focuses correction better above Schroder, roughly speaking. Jerry's guide only mentions to "refer to REW Help". Has any closure come on this? And which method is "closest" to what we hear?

b) If you want to access the overall success of achieving a target curve, such as we'd draw in Dirac (or Trinnov, in my case), what smoothing is closest to determine the overall trend? Reason I'm wondering is that when I did some post-Trinnov full-range plots, with 1/6th or 1/3rd (ugh) smoothing, I have +/- 5 db or so variation in my speaker responses at 100 Hz to maybe 8 kHz, which used to fall within "acceptable" back when. However, that also washes out my implementation of a Harman curve, which is gently descending from about 100 Hz by about 4 db or so until we get to that 8 kHz. I'm wondering how I can best distinguish random variation due to the room from a trend in what I hear...

c) Waterfalls - is the 300 ms for "great" and 450 ms standard for "good" still the norm in practice, and has decay graphs been a more rigorous way to determine if there's ringing? Reason I'm wondering is I have <450 ms ringing at about 60 Hz on my L/R+Subs charts, representing about 15 ms difference between the plot and the decay at 160 ms. Not sure if this is an addressable problem or not, unless I find it's from my mains (meaning bass traps)? I have a pretty good speaker placement for Atmos with a 25-30 degrees placement, so I'm loathe to move the mains.

d) For more severe ringing - my sub response has a funeral pire at 600+ ms at about 40 Hz, which is also a null. Right now I have four corner sub placement, but I may look at switching out to midwall and see if it improves, as well as 1/4 and 3/4 length side positions. Anything I can do otherwise that’s practical there?

e) Distortion - anything I can do about 22% distortion at about 20 to 25 Hz short of giving up on low bass? My sub distortion is 5% or less everywhere else above 30 Hz.

f) Clarity - what the heck is anyone using it for? It sort of looks vaguely like what I'd expect a target curve to look like, but other than that I'm not sure if anyone has a standard response for "music" vs. "speech" clarity to look for

I'm eventually going to tackle SBIR with treatments and IR in general once I decide just how much of a Toolean I am about sidewall reflections, but for now the focus is the subs and getting my tonal balance down before I move onto Group Delay, IR and other goodies.

Audio Gear: Trinnov Altitude 32 (24 channel), NAD M27 amps (3)
Video: JVC RS600, Seymour 100" UF Screen, Lumagen Radiance Pro 4444 (coming soon)
Misc.: Oppo 103, Oppo 203 UHD player, JRiver
Speakers: PSB Imagine T3 LCR, Imagine T Wides and Side Surround 1, T2 Side Surrounds, Imagine XB rears, PSB CS1000 ceilings (6), HSU ULS-15 Mk 2 subs (4) - 11.4.6

Last edited by sdrucker; 10-06-2017 at 09:20 PM.
sdrucker is offline  
post #23583 of 23608 Old 10-07-2017, 07:37 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 2,919
Mentioned: 103 Post(s)
Tagged: 0 Thread(s)
Quoted: 1666 Post(s)
Liked: 646
Quote:
Originally Posted by sdrucker View Post
Hi guys,
a) For comparing full-range speaker frequency responses, is the norm still to do 1/6th smoothing, or are the newer methods like variable smoothing, psychoacoustic smoothing or ERB getting some favor? Wondering because I did a search on the thread and I've read some comments suggesting that for determining EQ, you might want to use variable smoothing as it focuses correction better above Schroder, roughly speaking.
It depends what you want to do with the data, REW recommends var for eq purposes though I wouldn't apply full range EQ on the basis of fractional octave smoothing. I don't find REW's frequency dependent window implementation alone v useful for in room measurements so IMV psy is the best option of the smoothing options for full range use if you want to just get a reasonable view of what is going on.

Quote:
Originally Posted by sdrucker View Post
Hi guys,
b) If you want to access the overall success of achieving a target curve, such as we'd draw in Dirac (or Trinnov, in my case), what smoothing is closest to determine the overall trend? Reason I'm wondering is that when I did some post-Trinnov full-range plots, with 1/6th or 1/3rd (ugh) smoothing, I have +/- 5 db or so variation in my speaker responses at 100 Hz to maybe 8 kHz, which used to fall within "acceptable" back when. However, that also washes out my implementation of a Harman curve, which is gently descending from about 100 Hz by about 4 db or so until we get to that 8 kHz. I'm wondering how I can best distinguish random variation due to the room from a trend in what I hear...
I'd use psy for this. REW does also now have an option for automated time aligning and vector averaging of multiple measurements which can be used in this situation too (i.e. take n measurements around the listening position and then use the vector averaging to get an overall view)

Quote:
Originally Posted by sdrucker View Post
c) Waterfalls - is the 300 ms for "great" and 450 ms standard for "good" still the norm in practice, and has decay graphs been a more rigorous way to determine if there's ringing? Reason I'm wondering is I have <450 ms ringing at about 60 Hz on my L/R+Subs charts, representing about 15 ms difference between the plot and the decay at 160 ms. Not sure if this is an addressable problem or not, unless I find it's from my mains (meaning bass traps)? I have a pretty good speaker placement for Atmos with a 25-30 degrees placement, so I'm loathe to move the mains.
waterfall for modal decay, wavelet spectrogram is useful for a full range view of the direct sound.

Quote:
Originally Posted by sdrucker View Post
d) For more severe ringing - my sub response has a funeral pire at 600+ ms at about 40 Hz, which is also a null. Right now I have four corner sub placement, but I may look at switching out to midwall and see if it improves, as well as 1/4 and 3/4 length side positions. Anything I can do otherwise that’s practical there?

e) Distortion - anything I can do about 22% distortion at about 20 to 25 Hz short of giving up on low bass? My sub distortion is 5% or less everywhere else above 30 Hz.
a modal resonance or null or will impact in room distortion data like this so that 22% reading, assuming it is 2nd order, is likely to simply be a function of that 40Hz issue. If you see further spikes at ~15Hz and ~11Hz then you can be quite sure of this. What subs are they? are they really capable of ULF?

Quote:
Originally Posted by sdrucker View Post
f) Clarity - what the heck is anyone using it for? It sort of looks vaguely like what I'd expect a target curve to look like, but other than that I'm not sure if anyone has a standard response for "music" vs. "speech" clarity to look for
from what I've read, it's as useful as something like RT60 in a small room, i.e. not very.

Last edited by 3ll3d00d; 10-07-2017 at 07:41 AM.
3ll3d00d is online now  
 
post #23584 of 23608 Old 10-07-2017, 04:09 PM
AVS Forum Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 3,291
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 888 Post(s)
Liked: 600
Quote:
Originally Posted by 3ll3d00d View Post

waterfall for modal decay, wavelet spectrogram is useful for a full range view of the direct sound.
I've always found spectrograms easier to read than waterfalls, so my comments are taken more from there than the waterfall as such, but point taken. I'll have to train myself on Jerry's waterfall notes more carefully.

Quote:
a modal resonance or null or will impact in room distortion data like this so that 22% reading, assuming it is 2nd order, is likely to simply be a function of that 40Hz issue. If you see further spikes at ~15Hz and ~11Hz then you can be quite sure of this. What subs are they? are they really capable of ULF?
HSU ULS-15 Mk II, which is a sealed sub, and closely similar to what AustinJerry has in an older version. They're rated by HSU at 20 hz +/- 2 db anechoically in EQ1 mode, and I've read reviews in Audioholics with measurements that have them flat to at least 25 Hz more practically. I'm not a basshead (maybe 60/40 HT vs. music) so I can live with relatively smooth output to 20 Hz ideally.

They're not in the same league as Seatons for <16 Hz ULF, but since I'm using multiples, and I've seen plots from Jerry and Roger Dressler that are smooth < 20 Hz in their rooms with the ULS-15s with a quad system, I'd at least like to get to 20 Hz with clean output at my typical listening volume.

Looks like I'll have to work on that 40 Hz area at a minimum. I do see a slight spike at 15 Hz, but that's well beyond where the sub has useful output. If I can get the null smoothed out I might get the 2nd order benefit, possibly.

Audio Gear: Trinnov Altitude 32 (24 channel), NAD M27 amps (3)
Video: JVC RS600, Seymour 100" UF Screen, Lumagen Radiance Pro 4444 (coming soon)
Misc.: Oppo 103, Oppo 203 UHD player, JRiver
Speakers: PSB Imagine T3 LCR, Imagine T Wides and Side Surround 1, T2 Side Surrounds, Imagine XB rears, PSB CS1000 ceilings (6), HSU ULS-15 Mk 2 subs (4) - 11.4.6
sdrucker is offline  
post #23585 of 23608 Old 10-07-2017, 06:27 PM
AVS Forum Special Member
 
LastButNotLeast's Avatar
 
Join Date: Feb 2007
Location: 08077
Posts: 8,351
Mentioned: 35 Post(s)
Tagged: 1 Thread(s)
Quoted: 1804 Post(s)
Liked: 1416
Quote:
Originally Posted by sdrucker View Post
I've always found spectrograms easier to read than waterfalls, so my comments are taken more from there than the waterfall as such, but point taken. I'll have to train myself on Jerry's waterfall notes more carefully.
Think of the spectrogram as looking down on the waterfall; same information, different perspective.
Once you're caught up on FR, you may want to look at the progress being made on TR (tactile response).
The VibSensor Accelerometer Test Thread
Or you might want to quit while you're ahead.
Michael
bmcn and unretarded like this.

Did you really need to quote that entire post in your reply?
Downloadable FREE demo discs: Demonstration Blu-Ray Discs (Independently Authored)
Welcome to AVS - Get out while you still can!
Don't guess, measure: Getting Started With REW: A Step-by-Step Guide
LastButNotLeast is offline  
post #23586 of 23608 Old 10-08-2017, 01:31 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 2,919
Mentioned: 103 Post(s)
Tagged: 0 Thread(s)
Quoted: 1666 Post(s)
Liked: 646
Quote:
Originally Posted by sdrucker View Post
I've always found spectrograms easier to read than waterfalls, so my comments are taken more from there than the waterfall as such, but point taken. I'll have to train myself on Jerry's waterfall notes more carefully.
Quote:
Originally Posted by LastButNotLeast View Post
Think of the spectrogram as looking down on the waterfall; same information, different perspective.
you can use either but just bear in mind the default configuration for the spectrogram is quite different to the waterfall these days

Quote:
Originally Posted by sdrucker View Post
If I can get the null smoothed out I might get the 2nd order benefit, possibly.
here's an example, my room has a big modal peak at ~41Hz

with PEQ only to pull down the peak



using multiple subs (with MSO driven EQ which did not involve a massive PEQ cut)



there were other differences between these two setups but, in either case, the subs were loafing so distortion should not have been an issue.
Attached Thumbnails
Click image for larger version

Name:	after.png
Views:	334
Size:	219.7 KB
ID:	2294578   Click image for larger version

Name:	before.png
Views:	323
Size:	180.4 KB
ID:	2294580  
3ll3d00d is online now  
post #23587 of 23608 Old 10-08-2017, 08:15 PM
AVS Forum Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 3,291
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 888 Post(s)
Liked: 600
Quote:
Originally Posted by 3ll3d00d View Post
you can use either but just bear in mind the default configuration for the spectrogram is quite different to the waterfall these days
I just find the format of the spectrogram - looking for the height of the plume, and a two-dimensional plot with the ms as the y axis and a heat map for the db level - easier to read. The Fourier configuration is the one you want, correct? It looks a LOT more like the spectrograms I used to generate on older versions of REW (back in 2015) than Wavelet.

The other thing I'm going to do is download MSO...I have one seat of interest to me in my single three-seat row, but I already do a minimum of five measuring positions and often as many as nine on my 72" wide sofa with Trinnov, and have a weighting scheme, so I can work with that from what I've seen of MSO. The later versions of the Altitude support PEQ for the individual subs before arriving at the summated sub response as a step before the Optimizer, so I can run the optimization and get information I can feed into the Altitude. From what I know with the PC-based architecture, there's no limit in practical terms to the number of filters we can use, the degree of attenuation, and Q. Should be fun to test out...

Audio Gear: Trinnov Altitude 32 (24 channel), NAD M27 amps (3)
Video: JVC RS600, Seymour 100" UF Screen, Lumagen Radiance Pro 4444 (coming soon)
Misc.: Oppo 103, Oppo 203 UHD player, JRiver
Speakers: PSB Imagine T3 LCR, Imagine T Wides and Side Surround 1, T2 Side Surrounds, Imagine XB rears, PSB CS1000 ceilings (6), HSU ULS-15 Mk 2 subs (4) - 11.4.6
sdrucker is offline  
post #23588 of 23608 Old 10-09-2017, 12:22 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 2,919
Mentioned: 103 Post(s)
Tagged: 0 Thread(s)
Quoted: 1666 Post(s)
Liked: 646
Quote:
Originally Posted by sdrucker View Post
The Fourier configuration is the one you want, correct?
yes, actually I just noticed the default settings vary according to the spectrogram type so setting the fourier spectrogram to default will give you something v similar to the default waterfall.

Quote:
Originally Posted by sdrucker View Post
The other thing I'm going to do is download MSO...I have one seat of interest to me in my single three-seat row, but I already do a minimum of five measuring positions and often as many as nine on my 72" wide sofa with Trinnov, and have a weighting scheme, so I can work with that from what I've seen of MSO. The later versions of the Altitude support PEQ for the individual subs before arriving at the summated sub response as a step before the Optimizer, so I can run the optimization and get information I can feed into the Altitude. From what I know with the PC-based architecture, there's no limit in practical terms to the number of filters we can use, the degree of attenuation, and Q. Should be fun to test out...
I use MSO with a PC based (acourate) setup, it is somewhat time consuming (albeit mostly time the computer spends churning away) but certainly works well for me & is interesting to play with to see what it comes up with.
3ll3d00d is online now  
post #23589 of 23608 Old 10-10-2017, 06:48 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Hello,
should I be concerned with the following.
I utilise Rew and Umik-1(with cxal. file 90 degrees) to set all channels levels at 75db.
In the receiver, the front left gian is set to 0db and the main volume as requested by the SPL meter in Rew. To get the front righ to the setting I must rise the gain into rotel by 2db.
I expected to see the 2 curves to be together. But it is not the case espacially above 1k hz (see the mdat).
Is it a problem?
If I put 0db into Rotel for the 2 fronts, SPL with Rew gives 75db for left and 73db for right but, the measurement shows the 2 responses together.




https://www.dropbox.com/s/yr6z9happz...sure.mdat?dl=0
Attached Thumbnails
Click image for larger version

Name:	SPL and main level.jpg
Views:	38
Size:	126.1 KB
ID:	2295970  

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall
FargateOne is offline  
post #23590 of 23608 Old 10-11-2017, 10:27 AM
AVS Forum Special Member
 
Alan P's Avatar
 
Join Date: Jan 2003
Location: South Dakota
Posts: 9,993
Mentioned: 66 Post(s)
Tagged: 0 Thread(s)
Quoted: 5142 Post(s)
Liked: 3836
Quote:
Originally Posted by FargateOne View Post
Hello,
should I be concerned with the following.
I utilise Rew and Umik-1(with cxal. file 90 degrees) to set all channels levels at 75db.
In the receiver, the front left gian is set to 0db and the main volume as requested by the SPL meter in Rew. To get the front righ to the setting I must rise the gain into rotel by 2db.
I expected to see the 2 curves to be together. But it is not the case espacially above 1k hz (see the mdat).
Is it a problem?
If I put 0db into Rotel for the 2 fronts, SPL with Rew gives 75db for left and 73db for right but, the measurement shows the 2 responses together.




https://www.dropbox.com/s/yr6z9happz...sure.mdat?dl=0
It is perfectly normal to see different response characteristics from the FL and FR speakers due to placement and interaction with room boundaries. Unless your room is a perfect rectangle and both speakers are evenly spaced from boundaries, you will see results like this. Nothing to worry too much about.

EDIT: However, thinking more about this, it is strange that the frequencies above 1khz are so much lower from your FL. Unless something is obviously blocking that speakers tweeter, you may have a problem with the tweeter in that speaker.
Alan P is offline  
post #23591 of 23608 Old 10-11-2017, 12:39 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Quote:
Originally Posted by Alan P View Post
(...)
EDIT: However, thinking more about this, it is strange that the frequencies above 1khz are so much lower from your FL. Unless something is obviously blocking that speakers tweeter, you may have a problem with the tweeter in that speaker.
That's what I thought. I swap both speakers and I got the same results (differences between the left and the right channel). If I equalize the level to get 75 both sides with Umik and Rew, I must to raise the right channel level into the receiver by 2 db and then I got the graph that I sent. On the contrary, if I want to get both measurements equal above 1 k hz, I must let gain to 0 into the receiver but then I got 75 db on the left and 73 on the right.
So it does not seem to be the speaker. I know nothing about electronics, but to me it does not make sense to suspect the receiver either. Unless you've got an idea?

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall
FargateOne is offline  
post #23592 of 23608 Old 10-11-2017, 01:15 PM
AVS Forum Special Member
 
Alan P's Avatar
 
Join Date: Jan 2003
Location: South Dakota
Posts: 9,993
Mentioned: 66 Post(s)
Tagged: 0 Thread(s)
Quoted: 5142 Post(s)
Liked: 3836
Quote:
Originally Posted by FargateOne View Post
That's what I thought. I swap both speakers and I got the same results (differences between the left and the right channel). If I equalize the level to get 75 both sides with Umik and Rew, I must to raise the right channel level into the receiver by 2 db and then I got the graph that I sent. On the contrary, if I want to get both measurements equal above 1 k hz, I must let gain to 0 into the receiver but then I got 75 db on the left and 73 on the right.
So it does not seem to be the speaker. I know nothing about electronics, but to me it does not make sense to suspect the receiver either. Unless you've got an idea?
When you say you swapped speakers, did you swap the speaker outputs on the AVR or did you swap their physical locations, or...?
Alan P is offline  
post #23593 of 23608 Old 10-11-2017, 02:21 PM
AVS Forum Special Member
 
ahblaza's Avatar
 
Join Date: Nov 2007
Location: Pittsburgh Steeler Country
Posts: 3,257
Mentioned: 60 Post(s)
Tagged: 0 Thread(s)
Quoted: 1971 Post(s)
Liked: 2221
Quote:
Originally Posted by Alan P View Post
When you say you swapped speakers, did you swap the speaker outputs on the AVR or did you swap their physical locations, or...?
I bet locations..........
ahblaza is offline  
post #23594 of 23608 Old 10-11-2017, 02:58 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Quote:
Originally Posted by Alan P View Post
When you say you swapped speakers, did you swap the speaker outputs on the AVR or did you swap their physical locations, or...?
Yes, I take left speaker to the right and so on.Not the cables at the output of the receiver

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall

Last edited by FargateOne; 10-11-2017 at 03:24 PM.
FargateOne is offline  
post #23595 of 23608 Old 10-11-2017, 04:03 PM
AVS Forum Special Member
 
Alan P's Avatar
 
Join Date: Jan 2003
Location: South Dakota
Posts: 9,993
Mentioned: 66 Post(s)
Tagged: 0 Thread(s)
Quoted: 5142 Post(s)
Liked: 3836
Quote:
Originally Posted by FargateOne View Post
Yes, I take left speaker to the right and so on.Not the cables at the output of the receiver
And the problem follows the speaker, or the location??
Alan P is offline  
post #23596 of 23608 Old 10-11-2017, 04:19 PM
AVS Forum Special Member
 
ahblaza's Avatar
 
Join Date: Nov 2007
Location: Pittsburgh Steeler Country
Posts: 3,257
Mentioned: 60 Post(s)
Tagged: 0 Thread(s)
Quoted: 1971 Post(s)
Liked: 2221
ahblaza is offline  
post #23597 of 23608 Old 10-11-2017, 04:40 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Quote:
Originally Posted by Alan P View Post
And the problem follows the speaker, or the location??
Changing speakers of place did not change the problem. The 2 db difference remains on the right front. I will later tonight (when the house will be calm !) change the cables at the output of the receiver (left speaker cable connected to the right front output and so on) and report.
OK, it is the room. I switched the cables output and the gain in the receiver followed the signal: left speaker went 2 db higher than the right.

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall

Last edited by FargateOne; 10-11-2017 at 07:05 PM.
FargateOne is offline  
post #23598 of 23608 Old 10-12-2017, 08:12 AM
AVS Forum Special Member
 
Alan P's Avatar
 
Join Date: Jan 2003
Location: South Dakota
Posts: 9,993
Mentioned: 66 Post(s)
Tagged: 0 Thread(s)
Quoted: 5142 Post(s)
Liked: 3836
Quote:
Originally Posted by FargateOne View Post
Changing speakers of place did not change the problem. The 2 db difference remains on the right front. I will later tonight (when the house will be calm !) change the cables at the output of the receiver (left speaker cable connected to the right front output and so on) and report.
OK, it is the room. I switched the cables output and the gain in the receiver followed the signal: left speaker went 2 db higher than the right.
Your first test (physically moving the speakers but leaving the connections alone) suggest that it is the room, but...

In your second test, if you left the speakers placed as they were in your original graph (FR louder than FL above 1khz) and you swapped cables and now the left speaker is higher in SPL, it is the AVR not the room.

Maybe I am not interpreting your explanation correctly though....
Alan P is offline  
post #23599 of 23608 Old 10-12-2017, 09:52 AM
AVS Forum Special Member
 
ahblaza's Avatar
 
Join Date: Nov 2007
Location: Pittsburgh Steeler Country
Posts: 3,257
Mentioned: 60 Post(s)
Tagged: 0 Thread(s)
Quoted: 1971 Post(s)
Liked: 2221
Quote:
Originally Posted by FargateOne View Post
Changing speakers of place did not change the problem. The 2 db difference remains on the right front. I will later tonight (when the house will be calm !) change the cables at the output of the receiver (left speaker cable connected to the right front output and so on) and report.
OK, it is the room. I switched the cables output and the gain in the receiver followed the signal: left speaker went 2 db higher than the right.
This seems like an AVR problem, now the left speaker is 2dB higher and not the right, by reversing the speakers connection to the binding posts correct, you moved the 2dB louder to the opposite speaker now. I don't know why you just can't raise the the speaker's trim level 2dB higher so you have the same output from both mains. 2dB is not a really big discrepancy.
ahblaza is offline  
post #23600 of 23608 Old 10-12-2017, 01:25 PM
AVS Forum Special Member
 
jamin's Avatar
 
Join Date: May 2001
Location: moot point
Posts: 1,128
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 43 Post(s)
Liked: 41
If, effectively, with all trims zero the speaker on left is always 2 db hot for total SPL then it is room/positioning problem.

That can make sense and underscore why built in test tones for level setting are band limited around 1 kHz. Looking at the plot below 1 kHz there is a lot of area where the right speaker response has large dips. This could account for the 2 dB difference in total SPL.

Line up the traces above 1kHz and let listening decide if you want to make further adjustment. Or fix whatever is causing the response anomalies.

plug in to play
Acoustic Mafia - Hear No Evil
jamin is online now  
post #23601 of 23608 Old 10-12-2017, 04:58 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Quote:
Originally Posted by jamin View Post
If, effectively, with all trims zero the speaker on left is always 2 db hot for total SPL then it is room/positioning problem.

That can make sense and underscore why built in test tones for level setting are band limited around 1 kHz. Looking at the plot below 1 kHz there is a lot of area where the right speaker response has large dips. This could account for the 2 dB difference in total SPL.

Line up the traces above 1kHz and let listening decide if you want to make further adjustment. Or fix whatever is causing the response anomalies.
I made some changes this afternoon. I moved all the set-up by 20 cm to the right to get the tv and the front at the center of the long wall. For the moment, i had time only to check the levels with Rew and Umik. Front left and right are at the same level with no need to add gain in the receiver.
Il will measure later tonight and report.

AlanP, you understood me well .

OK, after measuring all channels the culprit seems to the room and placement. New place for the fronts, 0 gain in the channels of the receiver, spl meter reads 75db for the right and the left front speakers and the traces line up .

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall

Last edited by FargateOne; 10-12-2017 at 08:17 PM.
FargateOne is offline  
post #23602 of 23608 Old 10-15-2017, 08:05 AM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
OK, here a kind of follow up of a conversation I had here with Markus767 in May (post 22956 and 22971 among others).

Surprisingly enough, I discovered that I could move the MLP (my wife agreed!!??) and the fronts speakers by 20cm to the right in order to put it (and the TV) at the very center of the long wall of my room and also to respect a little more the proportion 1 x 1,618 suggested here for speakers placement (http://www.cardas.com/room_setup_hor...and_square.php)

So the mic, FL, FR and Center speakers moved to the right (20cm) but surrounds and sub remained at there position untouched. Front right is the timing reference. Unfortunatly I had not the time to do multi positions measurements (will come later this week). To integrate the sub, we must see the excess group delay, say the experts here. Now I have this results. Am I correct to think that it seems good (better than what I had before)?

Then, I checked the sub distance for the integration with the fronts (see the mdat file https://www.dropbox.com/s/zlgxjgy2o0...20cm.mdat?dl=0 ) . Please note that the left and right are xo at 60hz, the center at 70hz and the surrounds at 80hz.
Is the difference between 390cm delay for the sub or 420cm worth it? I don't think so, correct?

To my ears, it sounds better (SS&I in 5.1 much better) but we know how our brain can imagine things with sound ! (put a new cable and try!) Does measurements support this?
Attached Thumbnails
Click image for larger version

Name:	MLP 20cm to right sub untouch delay rotel 390cm.jpg
Views:	26
Size:	117.5 KB
ID:	2298690   Click image for larger version

Name:	time alignment sub 390cm or 420cm.jpg
Views:	24
Size:	233.5 KB
ID:	2298692  

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall

Last edited by FargateOne; 10-15-2017 at 08:23 AM.
FargateOne is offline  
post #23603 of 23608 Old Yesterday, 09:32 AM
Member
 
Join Date: Aug 2017
Posts: 27
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 3
Hey everyone, Just downloaded REW last night, read Jerry's guide and here are my first set of measurements. I have a minidsp 88a and these measurements are after Dirac Calibration (Front left speaker)

Still trying to figure out how to interpret these graphs. My aim is also to compare these results with Audyssey (after removing minidsp from the setup)
Attached Thumbnails
Click image for larger version

Name:	Decay.jpg
Views:	14
Size:	86.0 KB
ID:	2302204   Click image for larger version

Name:	Distortion.jpg
Views:	14
Size:	66.0 KB
ID:	2302206   Click image for larger version

Name:	RT60.jpg
Views:	14
Size:	51.2 KB
ID:	2302208   Click image for larger version

Name:	Spectogram.jpg
Views:	12
Size:	72.0 KB
ID:	2302210   Click image for larger version

Name:	SPL and Phase.jpg
Views:	13
Size:	79.4 KB
ID:	2302212  

Click image for larger version

Name:	Waterfall.jpg
Views:	14
Size:	103.6 KB
ID:	2302214  

Last edited by Meyy Arunachalam; Yesterday at 09:47 AM.
Meyy Arunachalam is offline  
post #23604 of 23608 Old Yesterday, 10:12 AM
** Man of Leisure **
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 16,442
Mentioned: 214 Post(s)
Tagged: 1 Thread(s)
Quoted: 7732 Post(s)
Liked: 4329
Quote:
Originally Posted by Meyy Arunachalam View Post
Hey everyone, Just downloaded REW last night, read Jerry's guide and here are my first set of measurements. I have a minidsp 88a and these measurements are after Dirac Calibration (Front left speaker)

Still trying to figure out how to interpret these graphs. My aim is also to compare these results with Audyssey (after removing minidsp from the setup)
I applaud your first effort! HST, here are several suggestions to improve the presentation of the measurements:

1. The Spectral Decay, Spectrogram and the Waterfall are all measurements intended to represent the resonance in the low frequencies. They are more meaningful if the horizontal range is 15-300Hz, rather than all the way to 20KHz. It is almost impossible to interpret them as they are presented.

2. Your measurement peak is 96dB (as shown on the right side of the Spectrogram). Using this information, your waterfall vertical scale should be 56-96dB (a 40dB range). The timing window should be 450ms, not 300ms. And as previously mentioned, the horizontal scale should extend only to 300Hz.

3. Rather than show "SPL & Phase", click on the "All SPL" button and present the measurement without the phase and mic cal measurements. Use 50-100dB as the vertical scale.

4. Forget the RT60 for now. Distortion looks OK, no need to repeat.


If you make these changes, then your measurements will be presented in a format that we are used to here. Most of this was discussed in the guide.
TheBucketOfTruth likes this.
AustinJerry is online now  
post #23605 of 23608 Old Yesterday, 10:48 AM
Member
 
Join Date: Aug 2017
Posts: 27
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 3
Quote:
Originally Posted by AustinJerry View Post
I applaud your first effort! HST, here are several suggestions to improve the presentation of the measurements:

1. The Spectral Decay, Spectrogram and the Waterfall are all measurements intended to represent the resonance in the low frequencies. They are more meaningful if the horizontal range is 15-300Hz, rather than all the way to 20KHz. It is almost impossible to interpret them as they are presented.

2. Your measurement peak is 96dB (as shown on the right side of the Spectrogram). Using this information, your waterfall vertical scale should be 56-96dB (a 40dB range). The timing window should be 450ms, not 300ms. And as previously mentioned, the horizontal scale should extend only to 300Hz.

3. Rather than show "SPL & Phase", click on the "All SPL" button and present the measurement without the phase and mic cal measurements. Use 50-100dB as the vertical scale.

4. Forget the RT60 for now. Distortion looks OK, no need to repeat.


If you make these changes, then your measurements will be presented in a format that we are used to here. Most of this was discussed in the guide.

Thanks Jerry! Yes you are right, I got a little excited as soon as I started to get measurement readings without error messages lol. After prematurely posting those, I read the rest of your guide and realized the points you mentioned above. I am going to generate a waterfall plot again, this time with both my front speakers and subwoofer. And I'll just the scales again and post.
FargateOne likes this.
Meyy Arunachalam is offline  
post #23606 of 23608 Old Yesterday, 01:20 PM
Member
 
Join Date: Aug 2017
Posts: 27
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 3
Quote:
Originally Posted by AustinJerry View Post
I applaud your first effort! HST, here are several suggestions to improve the presentation of the measurements:

1. The Spectral Decay, Spectrogram and the Waterfall are all measurements intended to represent the resonance in the low frequencies. They are more meaningful if the horizontal range is 15-300Hz, rather than all the way to 20KHz. It is almost impossible to interpret them as they are presented.

2. Your measurement peak is 96dB (as shown on the right side of the Spectrogram). Using this information, your waterfall vertical scale should be 56-96dB (a 40dB range). The timing window should be 450ms, not 300ms. And as previously mentioned, the horizontal scale should extend only to 300Hz.

3. Rather than show "SPL & Phase", click on the "All SPL" button and present the measurement without the phase and mic cal measurements. Use 50-100dB as the vertical scale.

4. Forget the RT60 for now. Distortion looks OK, no need to repeat.


If you make these changes, then your measurements will be presented in a format that we are used to here. Most of this was discussed in the guide.


Here is the graph for the front, left and right together.
Btw is there a way to add the sub to this measurement (measure all 3 speakers at the same time?)
Attached Thumbnails
Click image for larger version

Name:	Front L and R spl.jpg
Views:	12
Size:	81.3 KB
ID:	2302352   Click image for larger version

Name:	Front Left and RIght Waterfall.jpg
Views:	13
Size:	106.0 KB
ID:	2302354   Click image for larger version

Name:	With Jerry's mod.jpg
Views:	12
Size:	76.7 KB
ID:	2302374  

Last edited by Meyy Arunachalam; Yesterday at 01:30 PM.
Meyy Arunachalam is offline  
post #23607 of 23608 Old Yesterday, 02:30 PM
** Man of Leisure **
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 16,442
Mentioned: 214 Post(s)
Tagged: 1 Thread(s)
Quoted: 7732 Post(s)
Liked: 4329
Quote:
Originally Posted by Meyy Arunachalam View Post
Here is the graph for the front, left and right together.
Btw is there a way to add the sub to this measurement (measure all 3 speakers at the same time?)
When the left and right speakers are set to small with a crossover, then frequencies below the crossover are being sent to the subwoofer channel. So when you measure left and right channels from 15Hz-20KHz, the subs are automatically included in the measurement.

Couple of other comments: the “Average” measurement is meaningless. No need to show it. And the sub measurement was probably taken with HDMI4, correct? As you can see, the LFE channel has a 10dB boost. Nothing wrong with measuring it this way, I just wanted to make sure you understood why the sub measurement was so much higher than the others. And try to remember to add a description to each measurement, like “Front Left”, “Front Right”, etc. Otherwise, it is difficult to tell which measurement is which on the All SPL screen shot.

Something with the “All SPL - Jerry’s Mod” doesn’t look right. Besides not being able to tell which speaker belongs with which curve color, something is wrong in the 50-70 Hz range. Why is there a dip? Do these measurements include the sub(s)? See why it is so important to label the measurements?
AustinJerry is online now  
post #23608 of 23608 Old Yesterday, 03:49 PM
Member
 
FargateOne's Avatar
 
Join Date: Feb 2015
Posts: 136
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 6
Hello,
May I refer you to my post 23602
Do I interpret correctly my results when I think that moving the mains and MLP by 20 cm to the right give a good result with the excess group delay ? Or is it impossible to without multi positions measures ( which I will do soon).

My system 5.1 :ROTEL RSX-1562 /Fronts: B & W 804 D2; surround : B&W 704 and center B&W HTM4d2 /sub SVS PC 2000 / TV Samsung UN55ES8000
/BD player Cambridge CXU/ set-top (cable box) rented from cable co. PEQ by nanoAvr-DL Dirac Live between CXU and receiver
Bis Audio cables and power bar from the wall to the receiver
My room:15'4" x 11' 6" x 7'6" tv near the center of the long side wall
FargateOne is offline  
Sponsored Links
Advertisement
 
Reply Audio Theory, Setup, and Chat

Tags
Dayton , Dayton Audio , Room Equilizer Wizard Rew

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off