R972 Trinnov User Notes - Page 19 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #541 of 649 Old 05-23-2013, 06:27 AM
Advanced Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 676
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 39
In one of the Trinnov settings, Audiophlie1 (or is it 2), it says that it "voices the surrounds to match the LCR" or something to that effect.

What does that mean? Is it that it'll try to give the speakers the same frequency response or something like that?

As my surrounds and my LR are the same I don't think it'll make any difference to me.
artur9 is offline  
Sponsored Links
Advertisement
 
post #542 of 649 Old 05-23-2013, 10:04 AM
AVS Special Member
 
Stereojeff's Avatar
 
Join Date: Oct 2000
Location: Whittier, CA USA
Posts: 1,501
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 21
Artur9:

It actually attempts to voice all of the speakers to match the L/R channel. Let's assume you have B&W 901's for L/R. This setting uses the considerable EQ capability of the R-972 to make the frequency response of all of the other speakers match the 901's.

Jeff
Stereojeff is offline  
post #543 of 649 Old 05-23-2013, 10:41 AM
Advanced Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 676
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 39
Quote:
Originally Posted by Stereojeff View Post

Artur9:

It actually attempts to voice all of the speakers to match the L/R channel. Let's assume you have B&W 901's for L/R. This setting uses the considerable EQ capability of the R-972 to make the frequency response of all of the other speakers match the 901's.

Jeff

That doesn't sound like it''ll work unless the speakers are already "close" in FR? Like, if I'm talking Legacy Audio Helix for fronts and Whafedale Diamonds for surrounds I can't see how it would end well.

How close would the speakers have to be in practice?
artur9 is offline  
post #544 of 649 Old 05-23-2013, 03:22 PM
AVS Special Member
 
Stereojeff's Avatar
 
Join Date: Oct 2000
Location: Whittier, CA USA
Posts: 1,501
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 21
It's remarkably capable. While I haven't spent a great deal of time with that feature as I prefer to fix the room issues from 300 Hz down, I've done multi-channel demos with surprisingly different speakers--even for L & R with very good results.

I've even done demonstrations where we discovered during calibration that the tweeter was blown and still managed to create a seamless soundstage.

Jeff
Stereojeff is offline  
post #545 of 649 Old 05-25-2013, 03:47 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Quote:
Originally Posted by artur9 View Post

In one of the Trinnov settings, Audiophlie1 (or is it 2), it says that it "voices the surrounds to match the LCR" or something to that effect.

What does that mean? Is it that it'll try to give the speakers the same frequency response or something like that?

As my surrounds and my LR are the same I don't think it'll make any difference to me.

Audyssey had a feature that did something similar to that.It was supposed to match the voicing of the other channels to the L&R mains.I don't know how well their version worked,but they eventually dropped it from their features list on most of the AVR's that used it.
StevenLansing is offline  
post #546 of 649 Old 05-26-2013, 12:36 AM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by artur9 View Post

In one of the Trinnov settings, Audiophlie1 (or is it 2), it says that it "voices the surrounds to match the LCR" or something to that effect.

What does that mean? Is it that it'll try to give the speakers the same frequency response or something like that?

The reason behind this option is that many people buy speakers for their particular voicing. The value is that the correction keeps the L/R voicing, essentially a custom target curve.

Trinnov corrects for room issues in L/R, and modifies the response of Center and Surrounds to match, within the limits of correction. The "matching to L/R is three steps:(1) Trinnov measures all speakers, (2), then smooths L/R response to their overall curve, and (3) correction is applied all other speakers to the L/R curve (up to the +6/-12dB limit). If the speakers are within the correction limit, voicing will be very similar, depending on a number of other relevant criteria.

Cheers,

___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #547 of 649 Old 05-26-2013, 07:06 AM
Member
 
islandman2020's Avatar
 
Join Date: Oct 2010
Posts: 118
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
I love learning on this thread. It never ceases to amaze me what Trinnov is capable of. I'm afraid though, when all the R-972s are gone there will not be any "affordable" AVR or Pre/pro on the market for people. I heard the Emotiva XMC-1 at their Emofest last year. It's quite an amazing machine, however, I believe Emotiva is struggling to get it to market because the implementation TacT and that it is more complex to do than they surmised. I wish they would abandon Tact altogether and go with a version on Trinnov. It seems to me that Sherwood has already paved the way, and did a lot of the R&D that might somehow relate to making the implementation of Trinnov more cost effective than Tact. Sorry, to diverge here......just a thought.
islandman2020 is offline  
post #548 of 649 Old 05-26-2013, 07:12 AM
Advanced Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 676
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 39
Quote:
Originally Posted by Curt_Trinnov View Post

The reason behind this option is that many people buy speakers for their particular voicing. The value is that the correction keeps the L/R voicing, essentially a custom target curve.

.... . If the speakers are within the correction limit, voicing will be very similar, depending on a number of other relevant criteria.

Cheers,

I'm curious as to what "the correction limit" is, if it is easily expressible outside the language of DSP?

The other question is, is this a custom curve that the higher end units come with as well or is it something that the installer would have to build?
artur9 is offline  
post #549 of 649 Old 05-26-2013, 09:53 AM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 18,537
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 312 Post(s)
Liked: 556
Quote:
Originally Posted by artur9 View Post

I'm curious as to what "the correction limit" is, if it is easily expressible outside the language of DSP?
Wouldn't it be the limit that Curt mentioned in the post you quoted?
Quote:
Originally Posted by Curt_Trinnov View Post

...correction is applied all other speakers to the L/R curve (up to the +6/-12dB limit).

Sanjay
sdurani is online now  
post #550 of 649 Old 05-26-2013, 03:36 PM
Advanced Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 676
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 39
Quote:
Originally Posted by sdurani View Post

Wouldn't it be the limit that Curt mentioned in the post you quoted?

The limit Curt mentioned is a volume adjustment limit. The limit I'm curious about is the frequency range limits.

For example, I could easily see Trinnov deciding to make all speakers sound the same by limiting the fronts to a frequency range of 75Hz-15Khz because the sides cannot do better than that. Or vice versa.

Not that I expect it to do anything as boneheaded as that.
artur9 is offline  
post #551 of 649 Old 05-26-2013, 06:25 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Quote:
Originally Posted by artur9 View Post

The limit Curt mentioned is a volume adjustment limit. The limit I'm curious about is the frequency range limits.

For example, I could easily see Trinnov deciding to make all speakers sound the same by limiting the fronts to a frequency range of 75Hz-15Khz because the sides cannot do better than that. Or vice versa.

Not that I expect it to do anything as boneheaded as that.

It would try to correct the full frequency range,but only within the gain and reduction limits of +6 or -12dB that the in room measurements show that each speaker is capable of being corrected. The further off each speaker's in room response is and it's individual roll off points indicate,play in how closely their responses can be matched.
StevenLansing is offline  
post #552 of 649 Old 06-17-2013, 08:03 PM
Advanced Member
 
pwiss's Avatar
 
Join Date: Oct 2004
Location: Oregon
Posts: 812
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17
Curt or anyone,

I increased my subs gain and performed a new calibration. My initial calibration had my subs at 10 o'clock on the gain and gave me 69 hz on the mains and the subs were at 23hz with a -5db point. My new calibration with my subs at 1 o' clock on the gain gave me 74 hz on the mains and the subs were at 19hz with a +4 db point. I am a bit confused about the sub db point settings. I would think that increasing the gain would lower Trinnovs db point even more not increase it. Could someone explain how this works?
pwiss is offline  
post #553 of 649 Old 06-17-2013, 09:10 PM
AVS Addicted Member
 
cschang's Avatar
 
Join Date: Mar 2003
Location: Manhattan Beach, CA
Posts: 14,729
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 70
Quote:
Originally Posted by pwiss View Post

Curt or anyone,

I increased my subs gain and performed a new calibration. My initial calibration had my subs at 10 o'clock on the gain and gave me 69 hz on the mains and the subs were at 23hz with a -5db point. My new calibration with my subs at 1 o' clock on the gain gave me 74 hz on the mains and the subs were at 19hz with a +4 db point. I am a bit confused about the sub db point settings. I would think that increasing the gain would lower Trinnovs db point even more not increase it. Could someone explain how this works?
Interesting.

Nothing else changed? Sub position? Speaker positions?

How about moving back to 10 o'clock and calibrating again?

-curtis

Owner of Wave Crest Audio
Volunteer Mod at the Ascend Acoustics Forum
Like all things on the Internet, do your research, as forums have a good amount of misinformation.
Help beat breast cancer!

cschang is offline  
post #554 of 649 Old 06-18-2013, 10:01 AM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by pwiss View Post

Curt or anyone,

I increased my subs gain and performed a new calibration. My initial calibration had my subs at 10 o'clock on the gain and gave me 69 hz on the mains and the subs were at 23hz with a -5db point. My new calibration with my subs at 1 o' clock on the gain gave me 74 hz on the mains and the subs were at 19hz with a +4 db point. I am a bit confused about the sub db point settings. I would think that increasing the gain would lower Trinnovs db point even more not increase it. Could someone explain how this works?

Trinnov Reported Levels. The reported levels are relative levels for all speakers on a scales from - to +20dB (like the volume control). Trinnnov measurement indicates you turned up your sub 10dB (from -5 to +4, including "0dB = 10dB of gain change). The goal is to have the sub level be near the level of the speakers. This is because if there are big differences between speakers, the required level correction to make them all equal eats up the system's headroom by the same amount. As an example, if you have a 30dB spread (unlikely, but makes the point), the system would operate 30dB below the loudest level to keep the speakers in balance- so one would be throwing away 30dB of headroom, signal to noise, etc. Sometimes the spread between speakers can't be avoided, because the speakers have different efficiencies or have big distance variations. If one does have an adjustment on the sub level, the best option is to find a level where it is within 0-5dB or so of the speakers.

Trinnov Reported Frequencies. The frequencies will vary, particularly the sub, from measurement to measurement. The measurement is telling you where the sub becomes -6dB relative to the sub's averaged output. There are several factors that come into play besides the sub's output at very low frequencies. Unless you have a purpose built room with excellent outside noise isolation, and known the room mode behavior, it is very difficult to get a repeatable level (this is what you are having reported) in the area of 25hz or so. One can expect to see a variation of 5Hz. In your case, it may simply be an indication that there may be only a few dB of level change between 19-23Hz. Averaging the results, they are both good numbers, but without a measurement graphs, there is really no way to know exactly what is going on. On the plus side, Trinnov looks at both time and energy to determine correction of all speakers, including the sub, so the results can be quite good, especially if one has sorted out the best solution to the room modes.

If there is enough interest from R972 owners, I may organize a go to meeting session on how to best configure the sub with the R972, illustrating with measurements what you all are experiencing in your rooms. PM me if interested. If you've not read the User Notes, you'll find more info there on this and other topics.

Cheers,
Encom Tron likes this.

___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #555 of 649 Old 06-18-2013, 01:01 PM
Advanced Member
 
pwiss's Avatar
 
Join Date: Oct 2004
Location: Oregon
Posts: 812
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17
Thanks Curt for the reply. So it looks like both the 10 o'clock and 1 o'clock positions are within the +/- 5 db range. I really like the way the 1 o'clock position sounds especially on movies. I have the +6db files installed and previously felt that the bass response was good overall but now it is more pronounced but not even close to sounding boomy or standing out unnecessarily. By the way I am using the flat curve. I might try a calibration with the subs at 12 o'clock to see how that sounds.
pwiss is offline  
post #556 of 649 Old 06-22-2013, 10:01 AM
Advanced Member
 
pwiss's Avatar
 
Join Date: Oct 2004
Location: Oregon
Posts: 812
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17
Quote:
Originally Posted by Curt_Trinnov View Post

Trinnov Reported Levels. The reported levels are relative levels for all speakers on a scales from - to +20dB (like the volume control). Trinnnov measurement indicates you turned up your sub 10dB (from -5 to +4, including "0dB = 10dB of gain change). The goal is to have the sub level be near the level of the speakers. This is because if there are big differences between speakers, the required level correction to make them all equal eats up the system's headroom by the same amount. As an example, if you have a 30dB spread (unlikely, but makes the point), the system would operate 30dB below the loudest level to keep the speakers in balance- so one would be throwing away 30dB of headroom, signal to noise, etc. Sometimes the spread between speakers can't be avoided, because the speakers have different efficiencies or have big distance variations. If one does have an adjustment on the sub level, the best option is to find a level where it is within 0-5dB or so of the speakers.

Trinnov Reported Frequencies. The frequencies will vary, particularly the sub, from measurement to measurement. The measurement is telling you where the sub becomes -6dB relative to the sub's averaged output. There are several factors that come into play besides the sub's output at very low frequencies. Unless you have a purpose built room with excellent outside noise isolation, and known the room mode behavior, it is very difficult to get a repeatable level (this is what you are having reported) in the area of 25hz or so. One can expect to see a variation of 5Hz. In your case, it may simply be an indication that there may be only a few dB of level change between 19-23Hz. Averaging the results, they are both good numbers, but without a measurement graphs, there is really no way to know exactly what is going on. On the plus side, Trinnov looks at both time and energy to determine correction of all speakers, including the sub, so the results can be quite good, especially if one has sorted out the best solution to the room modes.

If there is enough interest from R972 owners, I may organize a go to meeting session on how to best configure the sub with the R972, illustrating with measurements what you all are experiencing in your rooms. PM me if interested. If you've not read the User Notes, you'll find more info there on this and other topics.

Cheers,

Do the bass files play a part in the Trinnov measured sub level? I ran that calibration with the +6db file installed. If I were to install the +3db file would my sub level go from +4 to +1 even without re-running Trinnov?
pwiss is offline  
post #557 of 649 Old 06-22-2013, 10:20 AM
AVS Addicted Member
 
cschang's Avatar
 
Join Date: Mar 2003
Location: Manhattan Beach, CA
Posts: 14,729
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 70
Quote:
Originally Posted by pwiss View Post

Do the bass files play a part in the Trinnov measured sub level? I ran that calibration with the +6db file installed. If I were to install the +3db file would my sub level go from +4 to +1 even without re-running Trinnov?
I'm not Curt....but no, it doesn't.

-curtis

Owner of Wave Crest Audio
Volunteer Mod at the Ascend Acoustics Forum
Like all things on the Internet, do your research, as forums have a good amount of misinformation.
Help beat breast cancer!

cschang is offline  
post #558 of 649 Old 06-22-2013, 12:02 PM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by pwiss View Post

Do the bass files play a part in the Trinnov measured sub level? I ran that calibration with the +6db file installed. If I were to install the +3db file would my sub level go from +4 to +1 even without re-running Trinnov?
Correct. No need to recalibrate when loading a change in sub level via the files.

___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #559 of 649 Old 06-22-2013, 03:17 PM
Advanced Member
 
pwiss's Avatar
 
Join Date: Oct 2004
Location: Oregon
Posts: 812
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 17
Quote:
Originally Posted by Curt_Trinnov View Post

Correct. No need to recalibrate when loading a change in sub level via the files.

I was hoping so. The +3db file still keeps the bass sounding great. I tested with some of my favorite music and also played some scenes from Battle Los Angeles and the Avengers and these settings are a definitely a keeper.
pwiss is offline  
post #560 of 649 Old 07-08-2013, 04:23 PM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20

The question came up from someone contacting Sherwood support.  Tinnov is supposed to work in bi-amp mode?

 

 

One can activate bi-amp mode where 4 amplifier channels are connected to two speaker's hi-low binding posts (which are normally jumpered).  The result is that the amps carry the low impedance load on the associated frequencies. The amps for the hi frequencies are not burdened by the low impedance load on the low frequencies, etc.   One would have to determine if a particular amp/speaker combination benefits from such a configuration.

 

However, it's important to note, the R972 will not act as an active crossover.  Nor will the Trinnov individually calibrate the hi and lows.  Trinnov will calibrate each speaker, when bi-amped, will be fed by two amps.

 

Cheers-


___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #561 of 649 Old 07-08-2013, 09:30 PM
Member
 
asangamnerkar's Avatar
 
Join Date: Dec 2006
Posts: 67
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Curt,

Maybe this has been discussed here but in this link

http://www.hometheaterhifi.com/audio-calibration/audio-calibration-reviews/anthem-room-correction-arc-system-part-1/page-9-room-gain.html

" I expect many and perhaps all Audyssey MultEQ XT or XT32 enabled products, including those at higher price points, will also down-sample high-resolution material to 48k samples/sec when the room correction is enabled. The Sherwood R-972 with Trinnov also down-sampled high resolution material to 48k samples/sec."


there is a comment about Trinnov downsampling input signal before applying the algorithm. Any insights from your side?
asangamnerkar is offline  
post #562 of 649 Old 07-09-2013, 08:56 AM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20

 The R972 has two TI DSPs. The first processes HD audio, etc, the second- Trinnov.  Any audio rate above 48k  is down-sampled to 48k before the second DSP/Trinnov using TI's downsampling engine.  This is typical: the Trinnov algorithm has been optimized to use the full processing power of the DSP when running 7.1 channels.  During development, we looked at the 96k sampling rate- but the trade-off in lost filter resolution just was not worth it, as every doubling of sample rate quadruples the DSP load, meaning the filters would be 1/4 of their current effectiveness. 

 

Cheers,


___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #563 of 649 Old 07-09-2013, 09:07 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
That's one of the inherent problems with implementing advanced DRC into AVR's.They just don't have enough DSP available like a stand alone dedicated DRC box can.Especially if the" brute force" approach is going to be used for filters.
StevenLansing is offline  
post #564 of 649 Old 07-10-2013, 10:29 AM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by StevenLansing View Post

That's one of the inherent problems with implementing advanced DRC into AVR's.They just don't have enough DSP available like a stand alone dedicated DRC box can.Especially if the" brute force" approach is going to be used for filters.

Not true.  As an example, the R972 has similar DSP capabilities dedicated to Trinnov processing as the Audyssey stand alone box.  The R972 has  two TI  DSPs, the second being used exclusively for Trinnov processing.  Audussey used a similar, single  TI DSP in their stand alone box.  DSPs pale in comparison to PCs for processing power, particularly evident in the significantly shorter time it takes to compute the filters.  This is why we see DRC systems that rely on PCs for computation, leaving the DSP for runtime use only.  Even then, DSP is limited in the filters that can be run.  This becomes even more evident as the channel count goes up.  In the R972, we chose long computational time over lower resolution or necessitating the use of a PC.  

 

As a R972/DRC user, is there anyone who would want to give up fidelity just so the computation (the computing time after you've run the cal mic) happens faster??

cschang likes this.

___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #565 of 649 Old 07-10-2013, 11:21 AM
AVS Special Member
 
M Code's Avatar
 
Join Date: Mar 2003
Location: Joshua Tree, CA
Posts: 9,773
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 32 Post(s)
Liked: 107
Quote:
Originally Posted by StevenLansing View Post

That's one of the inherent problems with implementing advanced DRC into AVR's.They just don't have enough DSP available like a stand alone dedicated DRC box can.Especially if the" brute force" approach is going to be used for filters.

A couple of things to keep in mind..

1. The 972 uses (2) TI DA710 DSPs which are now 5 years old..
2. Think about the DSP power much like the Intel or AMD processors in PCs, about every 2 years the power of the processors double while its cost decreases. Also clock frequency and on-board memory are crucial as well as these are increased...
3. The most powerful audio DSP processor available today is the Cirrus Logic CS49800 available in a tri or quad core (fixed point, 32 bit). As now used in some of the higher end Integra/Onkyo AVRs. While other brands are using AD sharcs or TI..
4. In order to handle the higher bits rates > 48kHz significantly more MIPs are required.

Bottom line..
The DSPs used in todays HDMI 1.4 AVRs are powerful in MIPs compared to previous audio DSPs however its most crucial challenge is to provide enough processing power to handle DTS Master Audio and additional post-processing modes (NEO X, PL2Z) in 7.1 and beyond.. 😳
But since AVRs are being strongly driven by price, there is a tendency to implement an audio DSP with just enough MIPs.
Going forward...
Due to the increasing demand of more digital sources and DSP processing for the main system and secondary zones this will require more powerful DSPs..

Just my $0.02... 👍😉
cschang likes this.
M Code is online now  
post #566 of 649 Old 07-10-2013, 12:06 PM - Thread Starter
Senior Member
 
Curt_Trinnov's Avatar
 
Join Date: Sep 2010
Location: Huntington Beach, Ca
Posts: 291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20

M Code-  Good points re DSP development.  I recall the R972 actually has DA708s.  I'd like to see Cirrus Logic move to floating point DSP, particularly  for DRC apps. 


___________
Curt Hoyt
3D Audio Consultant

Trinnov Audio
USA

Curt_Trinnov is online now  
post #567 of 649 Old 07-10-2013, 02:21 PM
AVS Special Member
 
M Code's Avatar
 
Join Date: Mar 2003
Location: Joshua Tree, CA
Posts: 9,773
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 32 Post(s)
Liked: 107
Quote:
Originally Posted by Curt_Trinnov View Post

M Code-  Good points re DSP development.  I recall the R972 actually has DA708s.  I'd like to see Cirrus Logic move to floating point DSP, particularly  for DRC apps. 

Hi Curt, Hope all is well... 😃

Update...
In checking the 972 Service Manual, it uses a DA710 for the DSP decoding and a DA708 for the Trinnov. Both processors use the same basic engine except with different clock speeds, the DA710 is 275/300 MHz while the 708 is 266 MHz....

Just my $0.02... 👍😉
M Code is online now  
post #568 of 649 Old 07-10-2013, 07:14 PM
AVS Special Member
 
Ryan1's Avatar
 
Join Date: Apr 2001
Posts: 1,850
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 36
Quote:
Originally Posted by StevenLansing View Post

That's one of the inherent problems with implementing advanced DRC into AVR's.They just don't have enough DSP available like a stand alone dedicated DRC box can.Especially if the" brute force" approach is going to be used for filters.

I don't know if I would call it a "problem," necessarily.

I would guess that the higher filter resolution produces audible benefits, while the real-world audible benefits of sampling rates higher than 48k would be at best debatable.
Ryan1 is offline  
post #569 of 649 Old 07-11-2013, 09:44 AM
AVS Special Member
 
Stereojeff's Avatar
 
Join Date: Oct 2000
Location: Whittier, CA USA
Posts: 1,501
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 21
FWIW, during development of the R-972 we were given many choices regarding available and necessary tradeoffs. We could implement 96 kHz signal processing with acoustic correction to 48 kHz but with only 512 tap FIR filters (+ 7 IIR filters for the bass) or we could process at 48 kHz with acoustic correction to 24 kHz with 1024 tap filters (and the same IIR filters). There are a number of reasons we chose 48 kHz including the lack of economical microphones flat to 48 kHz, but the number one reason was the more robust 1024 tap filters. That gave us the same preferred FIR resolution as the multi-kilobuck Trinnov professional system available then.

Jeff
Stereojeff is offline  
post #570 of 649 Old 07-11-2013, 04:34 PM
Advanced Member
 
StevenLansing's Avatar
 
Join Date: Mar 2007
Posts: 629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
I was just speaking in general about DRC and DSP limitations related to AVR's,not Trinnov or the R-972 in particular. There were systems that used a lot of filter taps that required lots of processing power for their correction due to a "Brute force" approach. The Audyssey stand alone used the whole chip just for generating the filters and everything else off loaded to a PC. It was also limited to 48KHz for the same reasons.

I never said there was something wrong with those choices.Just that those are some of the limitations that are inherent in the AVR implementations.Off loading calculations to a PC is another alternative to having the calculations performed on board with DSP's,not that it's necessarily a more cost effective or a more intuitive way to do so for AVR's versus using the DSP's for the tasks.

I'm sure that others made the same choices for the same reasons Stereojeff has mentioned they did with the R-972.

I've yet to see an AVR that has the DSP horsepower to run what the current production stand alone DRC boxes/processors are running for DRC filters.

There is always something more poweful in DSP's coming off the line,that doesn't mean the AVR companies are going to be using them.
StevenLansing is offline  
Reply Receivers, Amps, and Processors

Tags
Sherwood Newcastle R 972

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