Official DataSat RS20i thread. (Setup Tips, Questions,General Info, etc) - Page 41 - AVS Forum
Forum Jump: 
 8Likes
Reply
 
Thread Tools
post #1201 of 1866 Old 01-15-2014, 07:23 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by Curt_Trinnov View Post

Are there any other remapping technologies out there besides Trinnov?? 
Dirac Dimensions seems to be their version of remapping. DTS had remapping built into their HD audio codec, which resulted in some problems in the early days of Blu-ray (the famous "DTS bomb" discussed here at AVS). http://www.avsforum.com/t/1123436/dts-remapping
Quote:
Originally Posted by Curt_Trinnov View Post

The opposite is also is true- a wider sweet spot is possible- where a Trinnov Remapped loudspeaker improves a broader listening area because the added speaker is used to fill in a spatial gap.
You're right and I should have been clearer: IF you're remapping to more playback speakers than source channels, then you replace phantom images with actual speakers and end up with more stable imaging. In my defense, I was discussing remapping in the context of Wookii's comment that "the vast majority of home cinema systems cannot achieve reference layout speaker positions".

Sanjay
sdurani is online now  
Sponsored Links
Advertisement
 
post #1202 of 1866 Old 01-15-2014, 07:50 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by Wookii View Post

So what are you saying, their description is incorrect, or Dirac Live doesn't work as they say it does?
My experience has been that most room correction systems aren't "designed to correct higher frequencies where the reflections are strongest, most consistent and have the greatest impact on audibility". In fact, it is usually the opposite, addressing low frequencies because those problems are more noticable and correcting them yields the most audible improvements. Many room correction systems (like Dirac) even allow the user to set a limit to how high (not low) the correction goes.

Keep in mind that 10kHz has a wavelength of around 35mm, so using a filter to pull down a peak will only be helpful at a small location in space. Which of your two ears will be at that spot?

Also, there are less reflections as the frequency goes higher and loudspeaker dispersion narrows. Lower frequencies are more omni-directional, resulting in stronger and more consistent reflections off multiple boundries.
Quote:
Originally Posted by Wookii View Post

Did they say whether it could be applied to multi-channel playback, or just stereo tracks?
The demo used 2 channel material, played back over the front L/R speakers. However all 7 speakers in the set-up were active, playing back "anti room" signals for the L/R speakers.

With 7.1-channel sources, there would be 49 signals (not counting the subs). Each speaker would be playing back its respective source channel (corrected using Dirac Live) and also playing back the anti-room signals for the other 6 speakers.

Sanjay
sdurani is online now  
post #1203 of 1866 Old 01-16-2014, 01:44 AM
Senior Member
 
Wookii's Avatar
 
Join Date: Aug 2008
Posts: 243
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 24
Quote:
Originally Posted by sdurani View Post

My experience has been that most room correction systems aren't "designed to correct higher frequencies where the reflections are strongest, most consistent and have the greatest impact on audibility". In fact, it is usually the opposite, addressing low frequencies because those problems are more noticable and correcting them yields the most audible improvements. Many room correction systems (like Dirac) even allow the user to set a limit to how high (not low) the correction goes.

Keep in mind that 10kHz has a wavelength of around 35mm, so using a filter to pull down a peak will only be helpful at a small location in space. Which of your two ears will be at that spot?

I think we're talking at crossed purposes here. I understand what you are saying, but you seem to be describing, particularly with your latter statement, the effect of frequency response (unless I am misunderstanding your post), not impulse response. I understand that for pure frequency response adjustment, there is little value in attempting to correct for anything other than bass frequencies, other than to generally shape the average frequency response to achieve a target/house curve.

In my previous posts I was referring specifically to the way I understand Dirac to be designed to combat direct early reflections in the time domain, i.e. impulse response. If my understanding is correct (which I am happy to be corrected on if it is not) Dirac will attempt to identify first order reflections reaching the listening position with a time delay to the main signal (presumably within the usual 300ms within which our brains cannot differentiate from the direct sound). It will then generate a signal through the appropriate speaker, that is exactly of inverse phase to the reflected sound with an appropriate timing such that the reflection is essentially cancelled (at least in part) at the listening position.

This treatment would be effective at a wider range of frequencies than straight frequency response correction and indeed I assume the software would more easily detect such direct reflections where they become less omni-directional - i.e. outside of the bass region. This is particularly relevant to mid-range frequencies where the majority of the speech in a movie soundtracks exists, as the smearing the first order reflection can create, can significantly affect the intelligibility in this area, not to mention altering the spatial cues in the soundtrack.

Anyway, what we are both doing is speculating without the full knowledge of what Dirac is actually doing, so perhaps, Neil, Carl or other with direct experience of implementation of the software can comment on exactly what Dirac is doing to correct impulse response errors, within what range it will successfully correct these, and indeed how effective it is in practice on the RS20i (to bring us back on topic)?
Wookii is online now  
post #1204 of 1866 Old 01-16-2014, 08:46 AM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by Wookii View Post

If my understanding is correct (which I am happy to be corrected on if it is not) Dirac will attempt to identify first order reflections reaching the listening position with a time delay to the main signal (presumably within the usual 300ms within which our brains cannot differentiate from the direct sound). It will then generate a signal through the appropriate speaker, that is exactly of inverse phase to the reflected sound with an appropriate timing such that the reflection is essentially cancelled (at least in part) at the listening position.
As I said earlier, if you believe that Dirac's equalization is cancelling physical reflections, as opposed to just the effects of those reflections (peaks & dips), then I won't try to tell you otherwise. IF you start from that premise, then it is understandable to have the assumption you previously stated about Dirac Live "already correcting" things that Dirac Unity is supposed to correct. From that perspective, Dirac Unity is either duplicating or refining what Dirac Live was already doing, rather than the "new radical approach" that Dirac is claiming.

Sanjay
sdurani is online now  
post #1205 of 1866 Old 01-16-2014, 08:47 AM
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 297
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 24
Quote:
Originally Posted by sdurani View Post


...Keep in mind that 10kHz has a wavelength of around 35mm, so using a filter to pull down a peak will only be helpful at a small location in space. Which of your two ears will be at that spot?
 

 

Sorry, I really don't understand this comment.  If a filter is used to drop 10k some amount, that change would be apparent anywhere there is 10k propagation in the room from said loudspeaker.   Could you be referring to the often assumed theory that the intended correction of 10k only works to improve the overall response (by adjusting 10k) "at a small location in space," due to the small wavelength?  

 

If so, I don't agree with this principal in such general terms.  The improvement is dependent on the measurement and the variables used to come up with the filter.  High frequency correction can be shown to be very significant, particularly when one throws the added value of personal voicing into the mix.

 

Cheers,


___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is offline  
post #1206 of 1866 Old 01-16-2014, 08:55 AM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by Curt_Trinnov View Post

Could you be referring to the often assumed theory that the intended correction of 10k only works to improve the overall response (by adjusting 10k) "at a small location in space," due to the small wavelength?
I was talking about a peak at the mic location caused by a reflection, and attempting to reduce that peak by timing the cancellation signal to arrive at that location at the same time as the reflection does.

Sanjay
sdurani is online now  
post #1207 of 1866 Old 01-16-2014, 09:29 AM
Senior Member
 
Wookii's Avatar
 
Join Date: Aug 2008
Posts: 243
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 24
Quote:
Originally Posted by sdurani View Post

As I said earlier, if you believe that Dirac's equalization is cancelling physical reflections, as opposed to just the effects of those reflections (peaks & dips), then I won't try to tell you otherwise.

Its not a matter of what I 'believe', I am simply seeking clarification/confirmation of what is actually happening and how this impulse correction works. I have simply outlined my understanding of the explanation Dirac themselves provide for how Dirac Live works, for further discussion. I have no 'belief' either way as to whether what they state is actually correct or not, and I would like to hear actual experiences for people who have used the system in the field or been involved in its implementation at the device level.

I presume you have used Dirac yourself and/or have a good understanding of how their software works? In which case I would welcome you sharing your thoughts on how their impulse response correction works (or more importantly, doesn't work!)
Wookii is online now  
post #1208 of 1866 Old 01-16-2014, 10:37 AM
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 297
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 24
Quote:
Originally Posted by sdurani View Post


I was talking about a peak at the mic location caused by a reflection, and attempting to reduce that peak by timing the cancellation signal to arrive at that location at the same time as the reflection does.

What about the earliest reflections:  let's not overlook those, as they are equally important.  Early reflection correction at mid to very high frequencies can be achieved over a broad listening area, providing improved clarity.

 

Cheers,


___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is offline  
post #1209 of 1866 Old 01-16-2014, 12:27 PM
AVS Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 20,499
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 153
As I remember it the Dirac white paper gives a credible explanation of how it deals with early reflections.

The oft-mentioned issue of it only being valid at one position is not always true, or at least not 100% true.

As an example, if there's a back wall reflection from the center speaker, and the listening area subtends not too big of an angle, the arrival of the reflection at all seats will not vary much in time, and a less than full-strength inverse of the reflection could partially cancel it at all listening positions.

Kurt, does Trinnov do something like that?

Noah
noah katz is offline  
post #1210 of 1866 Old 01-16-2014, 04:36 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by Curt_Trinnov View Post

Early reflection correction at mid to very high frequencies can be achieved over a broad listening area, providing improved clarity.
Depends on how broad a listening area you're describing for correction of very high frequencies.

A speaker launches a sound towards the listener. The direct sound reaches the listener first. A reflection will arrive a few milliseconds later. So the speaker launches a second sound a few milliseconds later, which is an inverse of the reflection, timed to arrive at the listener right when the reflection does.

http://www.trinnov.com/technologies/loudspeaker-room-optimization/3d-simulations/

If the timing of the arrival of the cancellation signal is slightly off for large wavelengths, that's not as much of a problem as it is for small wavelengths (very high frequencies).


Sanjay
sdurani is online now  
post #1211 of 1866 Old 01-17-2014, 04:26 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 13
Or in the case of MultEQ, correct that reflection with a strong "marketing" campaign over a wide listening area.wink.gif
StevenLansing is offline  
post #1212 of 1866 Old 01-17-2014, 07:04 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by StevenLansing View Post

Or in the case of MultEQ, correct that reflection with a strong "marketing" campaign over a wide listening area.wink.gif
Haw, haw, haw.

Actually, what Audyssey did is improve their room correction. The measurements below were taken from the pre-out of a receiver, in order to see what the filters were doing.



The blue line is their older version (XT), which isn't doing much in the important bass range but is attempting to address individual peaks & dips in the high frequencies (hair, grass, whatever it's called).

By comparison, the green & red lines show their latest version (XT32), which is concentrating most if its correction below 1kHz and doing more of a timbre shaping to the overall sound above that range.

An actual improvement, not mere marketing, in the right direction.

Sanjay
sdurani is online now  
post #1213 of 1866 Old 01-18-2014, 01:47 AM
AVS Addicted Member
 
CINERAMAX's Avatar
 
Join Date: Sep 2000
Location: everywhere
Posts: 12,610
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 162 Post(s)
Liked: 57
Interesting cable manufacturer of Tascam type cables. These seem to be the highest end option I have seen for Datasat.

These people seem to take it seriously, they offer unshielded first option and if you do not want to gamble then shielded.

VOVOX® link direct S

Balanced sound conductor built without compromise and therefore without screen. Perfectly suited for all balanced signals, unaffected whether it's on line- or microphone-level.

Since the sound conductor is not shielded, a small risk of interference must be taken in consideration. We suggest to use 'protect' sound conductors when interference will necessarily lead to an uncontrollable situation, e.g. for live concerts or live recordings. For all other applications, VOVOX® link direct S is the perfect choice.
•single solid core conductor for signal and ground
•silver-plated, oxygen-free copper conductors
•pigment and plasticizer-free high purity polymers
•non-shielded

VOVOX® link protect S

Shielded, balanced sound conductor with optimal noise protection. Offers optimal sound quality for situations in which absolutely no risk of interference can be taken.
•single solid core conductor for signal and ground
•silver-plated, oxygen-free copper conductors
•pigment and plasticizer-free high purity polymers
•shielded
•optimally protected against microphonics

New: VOVOX® tubelink protect S

Shielded, balanced sound conductor for the perfect connection of vacuum tube microphones with the PSU. Due to the famous, VOVOX-specific design of the cable, tube microphones do not only benefit from the optimal transmission of the audio signals, but also from the optimal supply of current.
•single solid core conductor for signal, ground and voltage
•silver-plated, oxygen-free copper conductors
•pigment and plasticizer-free high purity polymers
•shielded
•individually hand-made, customized for many microphone types

Important note: The manufacturers of tube microphones assemble diverse types of connectors. The number of contacts but also the pinout diagram is different from brand to brand. For that reason VOVOX® tubelink cables are individually made for their respective microphone model. The customer also defines the length of the cable. What you will finally get is a handmade single piece of outstanding quality – exactly what the treasures of your studio deserve.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
CINERAMAX is online now  
post #1214 of 1866 Old 01-18-2014, 07:18 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 13
Quote:
Originally Posted by sdurani View Post

Haw, haw, haw.

Actually, what Audyssey did is improve their room correction. The measurements below were taken from the pre-out of a receiver, in order to see what the filters were doing.



The blue line is their older version (XT), which isn't doing much in the important bass range but is attempting to address individual peaks & dips in the high frequencies (hair, grass, whatever it's called).

By comparison, the green & red lines show their latest version (XT32), which is concentrating most if its correction below 1kHz and doing more of a timbre shaping to the overall sound above that range.

An actual improvement, not mere marketing, in the right direction.

Are those pre-out measurements taking into account a crossover or subwoofer correction?

Did you look at the impulse response.Is there an improvement in the impulse response that supports the marketed claims and graphs of reflections being reduced?
StevenLansing is offline  
post #1215 of 1866 Old 01-18-2014, 07:24 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by StevenLansing View Post

Is there an improvement in the impulse response that supports the marketed graphs of reflections being reduced?
Sure, in the minimum-phase low frequencies, just as happens with other room correction systems (and properly dialed-in PEQ).

Sanjay
sdurani is online now  
post #1216 of 1866 Old 01-18-2014, 07:28 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 13
Quote:
Originally Posted by sdurani View Post

Sure, in the minimum-phase low frequencies, just as happens with other room correction systems (and properly dialed-in PEQ).

Strange? I haven't seen any graphs yet that show that with MultEQ. Do you have any examples? I've seen graphs of Dirac and Trinnov that show that.
StevenLansing is offline  
post #1217 of 1866 Old 01-18-2014, 07:48 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by StevenLansing View Post

I've seen graphs of Dirac and Trinnov that show that.
Can you post some before/after examples of in-room impulse response measurements?
rblnr likes this.

Sanjay
sdurani is online now  
post #1218 of 1866 Old 01-18-2014, 07:58 PM
AVS Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 2,154
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 110 Post(s)
Liked: 136
I think Steven's talking about amt's impulse response graphs on the R-972 with Trinnov thread last December, and that center channel pre-post one (% FS) I sent him from my own R-972 Trinnov version, and posted to the REW thread last week. But I didn't do any frequency filtering for that IR overlay chart, and it was in-room, not pre-out measurement. I can do a pre-out, but I've got a few things ahead of it on the list.

DK about any empirical measuring of IR for Dirac, though.

Stuart

 

Denon 4311 with XT32 and Audyssey Pro

Oppo 93 and 103

Panasonic VT50

Sherwood R-972 with its version of the Trinnov Optimizer

MiniDSP 10x10 HD

PSB Imagine T2, Center, and Surrounds (as of 5/2014); HSU ULS-15 subs (2)

 

The Audyssey FAQ Guide can be found here:


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

sdrucker is online now  
post #1219 of 1866 Old 01-18-2014, 11:04 PM
Member
 
Interleave's Avatar
 
Join Date: Apr 2007
Posts: 57
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Hi all just wanted to see how you have found Dirac impacts the SPL before and after calibration. I have found an 8db difference is this normal?
Interleave is offline  
post #1220 of 1866 Old 01-18-2014, 11:31 PM - Thread Starter
AVS Club Gold
 
JapanDave's Avatar
 
Join Date: Jul 2009
Posts: 2,353
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 99
Quote:
Originally Posted by Interleave View Post

Hi all just wanted to see how you have found Dirac impacts the SPL before and after calibration. I have found an 8db difference is this normal?
I think it will depend entirely on your original in room frequency response. Mine was not affected that much, though I did notice it go down a few dB's.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

Couldn't pour water out of a boot with instructions on the heel.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

JapanDave is offline  
post #1221 of 1866 Old 01-19-2014, 02:21 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,564
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 464 Post(s)
Liked: 263
Quote:
Originally Posted by Interleave View Post

Hi all just wanted to see how you have found Dirac impacts the SPL before and after calibration. I have found an 8db difference is this normal?
Where does the 8 dB appear?

If by "uncalibrated" you mean the system was originally calibrated for a given SPL with a known reference (like noise at -30 dBFS), then after Dirac was performed, then any gain offsets resulting from the EQ would be absorbed by the gain trims. The master volume should operate the same as before. Even so, the impression of loudness may be different (regardless of the SPL meter) if the Dirac is correcting for rather significant errors.

Roger

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
Roger Dressler is offline  
post #1222 of 1866 Old 01-19-2014, 07:28 PM
Member
 
Interleave's Avatar
 
Join Date: Apr 2007
Posts: 57
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by Roger Dressler View Post

Where does the 8 dB appear?

If by "uncalibrated" you mean the system was originally calibrated for a given SPL with a known reference (like noise at -30 dBFS), then after Dirac was performed, then any gain offsets resulting from the EQ would be absorbed by the gain trims. The master volume should operate the same as before. Even so, the impression of loudness may be different (regardless of the SPL meter) if the Dirac is correcting for rather significant errors.

Quote:
Originally Posted by JapanDave View Post

I think it will depend entirely on your original in room frequency response. Mine was not affected that much, though I did notice it go down a few dB's.

OK guys here is what I have discovered. By playing the pink noise through the RS20 I can get all channels level matched to within .5db of each other brilliant. So I thought........................

I have the HAA guys out to calibrate and tune the room at the moment and when they put their pink noise generator onto the system playing back through SPDIF my left channel is out by 4-6db lower than the rest of the system. Has anyone else tried an external calibrated pink noise generator on the system to see what levels you are getting on the speakers? Is it possible that the internal pink noise generator could be that far off? If this is the case, this could explain why my post Dirac is lower than pre Dirac filters.

The equipment the guys are using is a Quantum Data 780..................

I thought I read someone else on here posting their left channel playing lower than the rest but I need to go through to see if I was mistaken or not.
Interleave is offline  
post #1223 of 1866 Old 01-19-2014, 08:06 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 13
Quote:
Originally Posted by sdurani View Post

Can you post some before/after examples of in-room impulse response measurements?

Actually I did find this not too long ago.

http://emotivalounge.proboards.com/thread/33047/thoughts-on-audyssey-multeq-dirac?page=1&scrollTo=566498


They are interesting results. MultEQ XT doesn't look like it's very effective on the impulse response by comparison. The funny thing is, all the REW graphs I've seen on AVS don't show any improvement with MultEQ XT or XT32 compared to the base measurements.The before and after impulse responses look identical.Even the ones sdrucker took so far of the ETC show no improvement.
StevenLansing is offline  
post #1224 of 1866 Old 01-19-2014, 08:41 PM
Senior Member
 
ceenhad's Avatar
 
Join Date: Dec 2005
Location: Glasgow UK
Posts: 406
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 24
Interleave. Check if your haa guy is using the DTS noise tracks off the 780. If so I suspect that may be the issue. Using the built in rta on the RS20 you can quickly confirm the input and output channel levels are accurate.

Neil Davidson - Visit my
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
ceenhad is online now  
post #1225 of 1866 Old 01-19-2014, 08:43 PM
AVS Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 2,154
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 110 Post(s)
Liked: 136
Quote:
Originally Posted by StevenLansing View Post

Actually I did find this not too long ago.

http://emotivalounge.proboards.com/thread/33047/thoughts-on-audyssey-multeq-dirac?page=1&scrollTo=566498


They are interesting results. MultEQ XT doesn't look like it's very effective on the impulse response by comparison. The funny thing is, all the REW graphs I've seen on AVS don't show any improvement with MultEQ XT or XT32 compared to the base measurements.The before and after impulse responses look identical.Even the ones sdrucker took so far of the ETC show no improvement.

My charts are still tentative, Steve. The last round of charts I did with IR and % FS are the most appropriate methodology for comparing alternate EQ, as I was informed by an expert on IR plots on the REW thread, in order to capture the raw impulse response. They show a reduction for at least some of the IR spikes with Trinnov compared to no Trinnov, given the same measuring position, the db level measured on the SPL meter of REW +/- 0.5 db, and the same speaker. But that's one chart, not a conclusive round of charts for multiple speakers. And there's the issue of whether unfiltered IR plots are a proper way to compare results, or we should look at specific frequencies (and if so, which ones with Filtered IR). Eventually I'll do a pre-out measurement of the speakers to see what the correction actually is. Testing Audyssey vs. a baseline is something I still need to do with the same % FS methodology.

The only thing I'll point out is that the Dirac charts on that Emotiva thread aren't up to the HTS-based standards that we've used on the REW thread for curve evaluation. YMMV, but 1/1 smoothing has a lot less resolution than the 1/6 or 1/12th smoothing for the 15 Hz to 20 kHz plots that are the HTS and AVS REW standards. And note that the % FS charts have a much broader graph scaling than the ones I've done. My % FS charts were 0 to 40 ms (based on the range for ETC plots of Filtered IR we have in our thread's REW guide), and these charts are 0 to 500 ms.

Thus, I think we have to regard the Dirac plots on the Emotiva thread as inconclusive unfortunately, although on the surface they look like Dirac is doing IR correction that Audyssey isn't.

I can't comment about whether the 100% to -100% range is too coarse or not. Maybe Sanjay has some thoughts about the proper axes for % FS measurements, and whether it's appropriate to do unfiltered plots or filter on specific frequencies.

Stuart

 

Denon 4311 with XT32 and Audyssey Pro

Oppo 93 and 103

Panasonic VT50

Sherwood R-972 with its version of the Trinnov Optimizer

MiniDSP 10x10 HD

PSB Imagine T2, Center, and Surrounds (as of 5/2014); HSU ULS-15 subs (2)

 

The Audyssey FAQ Guide can be found here:


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

sdrucker is online now  
post #1226 of 1866 Old 01-19-2014, 08:55 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 634
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 13
Quote:
Originally Posted by sdrucker View Post

My charts are still tentative, Steve. The last round of charts I did with IR and % FS are the most appropriate methodology for comparing alternate EQ, as I was informed by an expert on IR plots on the REW thread.

The only thing I'll point out is that the Dirac charts on that Emotiva thread aren't up to the HTS-based standards that we've used on the REW thread for curve evaluation. YMMV, but 1/1 smoothing has a lot less resolution than the 1/6 or 1/12th smoothing for the 15 Hz to 20 kHz plots that are the HTS and AVS REW standards. And note that the % FS charts have a much broader graph scaling than the ones I've done. My % FS charts were 0 to 40 ms (based on the range for ETC plots of Filtered IR we have in the REW guide), and these charts are 0 to 500 ms.

Thus, I think we have to regard the Dirac plots on the Emotiva thread as inconclusive unfortunately, although on the surface they look like Dirac is doing IR correction that Audyssey isn't.

I can't comment about whether the -100 to 100% range is too coarse or not. Maybe Sanjay has some thoughts about the proper axes for % FS measurements, and whether it's appropriate to do unfiltered plots or filter on specific frequencies.

I realize your graphs are still tentative. This example still shows something different happening between the 2 EQ systems.I don't think any measurements taken with REW will ever be completely conclusive.
StevenLansing is offline  
post #1227 of 1866 Old 01-19-2014, 09:05 PM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,369
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1008 Post(s)
Liked: 865
Quote:
Originally Posted by StevenLansing View Post

Actually I did find this not too long ago.

http://emotivalounge.proboards.com/thread/33047/thoughts-on-audyssey-multeq-dirac?page=1&scrollTo=566498


They are interesting results.
Interesting only because it is an earlier iteration of Audyssey room correction, not the one they released about 3 years ago.
Quote:
Originally Posted by StevenLansing View Post

MultEQ XT doesn't look like it's very effective on the impulse response by comparison.
In the XT vs XT32 graph I posted, I pointed out that XT wasn't doing anything in the bass range, so there shouldn't be much change to the impulse response. Hence Audyssey changing to XT32.
Quote:
Originally Posted by StevenLansing View Post

The funny thing is, all the REW graphs I've seen on AVS don't show any improvement with MultEQ XT or XT32 compared to the base measurements.
That is funny, because the most recent XT vs XT32 comparison I've seen on AVS does show an improvement in the impulse response:



Sanjay
sdurani is online now  
post #1228 of 1866 Old 01-19-2014, 09:12 PM
AVS Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 2,154
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 110 Post(s)
Liked: 136
Quote:
Originally Posted by StevenLansing View Post

I realize your graphs are still tentative. This example still shows something different happening between the 2 EQ systems.I don't think any measurements taken with REW will ever be completely conclusive.

No, and I agree that something is happening with Trinnov because I've seen the same effects on two different types of charts: the center/subs with an ETC plot, although my "pre" measurement was taken from my Denon 4311 vs. my R-972, and the center channel chart one using % FS, where my measurements were made from the R-972 in both cases, with the only change being turning Trinnov on or off and verifying the same speaker crossover Hz and db level.

Still, the only way I know to be "conclusive" is a full round of charts for Trinnov and Audyssey for individual speakers, comparing no correction to each REQ individually (i.e. a paired comparison). We can do this for both actual room measurement and for the bottom line, which is pre-outs. But even so, the real question is over how stable an area the degree of correction is. My "ad-hoc" off-axis plot (where I took a measurement about a foot from MLP; I had no way to measure a specific off-axis angle) shows at least some stability, but the question is how generalizable those results are. It may only hold for MLP and right positioning around MLP. Or for certain types of speakers/rooms for all we know.

Stuart

 

Denon 4311 with XT32 and Audyssey Pro

Oppo 93 and 103

Panasonic VT50

Sherwood R-972 with its version of the Trinnov Optimizer

MiniDSP 10x10 HD

PSB Imagine T2, Center, and Surrounds (as of 5/2014); HSU ULS-15 subs (2)

 

The Audyssey FAQ Guide can be found here:


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

sdrucker is online now  
post #1229 of 1866 Old 01-19-2014, 09:27 PM
Senior Member
 
stephenbr's Avatar
 
Join Date: Jan 2004
Location: Melbourne, Australia
Posts: 432
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 32 Post(s)
Liked: 20
Maybe some of this EQ discussion needs it own thread? It appears it is moving beyond aspects of the RS20i and even Dirac.
stephenbr is online now  
post #1230 of 1866 Old 01-19-2014, 11:12 PM
Member
 
Interleave's Avatar
 
Join Date: Apr 2007
Posts: 57
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
OK I have determined and figured out the reason for the large db change between the Dirac filters on/off. What the system does is if one channel is lower than the others then it level matches to the lowest output. In my case the left channel is off by 4db with the internal pink noise generator. Once I raise this level via the trims to +4db and rerun Dirac the output is then virtually on par SPL wise +-1db not the -8db I was getting in the past.

Now why would my left channel be -4db than the rest when using the internal pink noise, yet it is fine with the external pink noise generator? Is there a way to test the RS20 for faults via a multimeter on the XLR? If I send pink noise down the line should I be able to see the voltage difference on the line, is it this easy to check?

Otherwise is there any other way I can generate Pink Noise on the system and verify that potentially the RS20 is functioning correctly yet the external device is the fault. The interesting point here is I used the internal pink noise to level set the speakers months ago and all was good and nothing changed on the setup. Yet today rerunning the same pink noise to level set we are down with the internal noise but fine on the external noise.

Its all good and well to figure the problem out but the issue actually prevents me from running Dirac correctly as it believes Lchannel is -4db worse off as I dont have the ability to generate the external pink noise during Dirac optimisation.
Interleave is offline  
Reply Ultra Hi-End HT Gear ($20,000+)

User Tag List

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