Incorporating MiniDSP With Audyssey XT32

[agarwalro]

Shite! Ninja'd by KEW on the waterfall plot.

I need to fire up REW and reacclimatise. Might have you send over the .mdat to play with it.

 

[KEW]

Shite! Ninja'd by KEW on the waterfall plot.

I need to fire up REW and reacclimatise. Might have you send over the .mdat to play with it.

Don't let me stop you, I am uncomfortably beyond my ken! You've actually done it, I've only pondered it, LOL!

 

[TheWarrior]

@agarwalro

Room mode measurement provides information which can be meaningful, but in rooms as complex and fixed as Pogre’s it seems like it’d just add to the confusion. My approach is to use REW tools, like the waterfalls which you suggested, which warrior flippantly (IMPO) dismissed with a text wall of mostly irrelevant links (your words). Or other REW tools which can give you an idea of the most problematic modes, and can give a very good solution to them. I’ve tried several, several times how I use these things to Warrior, and he would just beat me up with his (incorrect) definitive assumptions, and condescending Toole fanboyish approach. Bottom-line; it all depends on one’s goals, expectations, and how ideal one’s situation is.

Moreover, and he can tell me I’m wrong, but I believe that Pogre has asked the same basic question several times without an answer to his satisfaction, yet Warrior continues to brow beat... In a sense, this is trolling.

Then let's use your conveniently 'sealed rec room' as 'exhibit A'.

My predictions for your room:

Length: 17.5 feet = 64 hz

Width: 18.5 feet = 61 hz

Height: 8.5 feet = 133 hz

Still need to know the front wall to the back wall of the storage area as that longer dimension will yield an even lower frequency.

As I recall, REW didn't get this right, which is why you came to me in the first place.

So the choice is yours, post shots of REW accurately predicting and recommending solutions in your room and how everything I've said is a waste of time, or continue chasing me across the forum claiming I'm a knowitall jack@@@ that is stubbornly trying to help people in a way that you don't like.

@Pogre

"To compute the frequencies at which axial standing waves occur, simply measure the distance between the walls, and divide that number into the speed of sound in whatever units the measurements were done." 'Sound Reproduction'

That reads the same no matter which edition you have. I'm trying to help you by singling out this one item from the entire chapter to help guide you in making informed decisions. And you are ignoring it because I cant succinctly explain all of the variables involved in the decision process. Perhaps you'll notice that it took Floyd 50 pages just to write that chapter. I'm not here to offer you cliff notes, just lending a hand with a very basic first step. You've made it clear you don't want my help, so I'll stop offering it.

 

[PENG]

If I apply up to 6db +/- to a frequency and it doesn't do the job, then I leave it. I'm assuming since eq isn't taking care of it, then those frequencies most likely are room modes. Is that a reasonable conclusion to arrive at without physically measuring the space? I don't have any issues with imaging or soundstage. I still get up to see if my center channel is on when listening to music! The Ultra towers are very good at both and I've played extensively with positioning them (moving them, and running a sweep after each move). I believe them to be optimized where they are.

I'm only interested in getting it to sound as good I can with the tools at hand. I don't want absorbing or diffusing panels, room treatments, or rearrange my living room and Alex knows that.

View attachment 22470
This is my in room response right now with no smoothing or room curve applied. It's also the graph to which TheWarrior was referring to when he said he could only assume my bass sounds bad, is full of bloat and I need to scrap the work I've done (with @ATLAudio's help) in these last 25 pages and start over.

I'm not seeing it? He keeps telling me that this in no way shows that I have a good bass response without measuring my asymmetrical room with vaulted ceilings and many uneven parallel surfaces to the nth degree. Am I doing something wrong? That's what I feel is being implied.

Thank you very much for saying what I would have wanted to express also. Like you, after spending hours and hours working with REW, moving mains and subs, Audyssey, PEQ onboard the subs, phase adjustments, and minidsp, the best graphs REW could get me were similar to your posted ones, i.e. quite smooth up to around 100 to around 120 Hz. Beyond that, there are dips that I can't do anything further about, setting higher XO (90Hz) for the mains (f3=35Hz) helps the 20-120 Hz range only.

I believe Audyssey, and other automated REQ software such as Dirac Live, Trinnov, Lynddorf's, though far from being perfect, do offer the best chance of achieving cleaner bass response in a lot of rooms, whereas REW/minidsp is highly limited. ATL offered a great idea of combining both Audyssey and REW/minidsp. That approach for sure could help, but in my room I found the minimal gain obtainable is not worth inserting an additional piece of powered hardware and interconnections, so I am now using the mini in my stereo system.

As for measuring and calculating room modes etc., I completely agree with you, if you have REW, minidsp and Audyssey's top version, it is much easier and quicker to run the automated way and then so your fine tuning until you get to the point of diminishing return, where you are face with getting rid of/move some of your furniture and/or sit where you don't want to sit. For me, it may be time to try acoustic panels.

 

[PENG]

This thread had a different title when it was started. It's evolved a couple of times. There's been a lot of good discussion and I've made big improvements in my sq thanks to the contributions of others. I am pretty sure through the course of this thread that I've measured plenty and have a very good idea where my room modes are without busting out a tape measure.

@PENG, where are you right now with the mini? Getting anywhere? Any good results?

I saw this post after I posted this morning, still, I provided sort of an update in my last post.

 

[agarwalro]

I'm not seeing it? He keeps telling me that this in no way shows that I have a good bass response without measuring my asymmetrical room with vaulted ceilings and many uneven parallel surfaces to the nth degree. Am I doing something wrong? That's what I feel is being implied.

It may be lost in the earlier posts and if so, I'm sorry for being lazy. Do you have measurements (with no Audyssey or miniDSP correction engaged) showing each sub and tower separate? Can you or ATLAudio direct me to them please?

 

[PENG]

View attachment 22471
This is the waterfall from the same measurement. Not sure what to make of it.

This is my Subs only waterfall:

 

[agarwalro]

View attachment 22471
This is the waterfall from the same measurement. Not sure what to make of it.

The Z axis, marked 0-300ms are time slices. If you look closely, the line at 0ms corresponds with this,
View attachment 22470

Every subsequent "slice" on the waterfall represents the "FR" (residual sound energy plotted as frequency vs time) for that slice after the impulse tone was played.

What makes a good looking waterfall plot?
1) No frequency bands showing excessively long decay time. (Example, a persistent bump along all slices from 0-300 with minor drop in level.)
2) All frequencies showing even energy decay. (Even decay time results in "parallel slices".)

Have you tried the ETC plot? It is way easier to interpret than waterfall plots at the expense of losing amplitude resolution.

If you post the ETC plot for the above measurement, I'll correlate the three.

 

[TheWarrior]

This is my Subs only waterfall:

View attachment 22508

What is your LFE crossover? 80 hz is 10dB down from 20 hz.

 

[ATLAudio]

What is your LFE crossover? 80 hz is 10dB down from 20 hz.

Looks like 80. It's a subwoofer only measurement, no speakers. AVR crossovers are usually 24 or 12 dB slopes.

I’m not sure the reason for posting my room dims and their predicted modes. Do you know if REW can determine this and reveal an applicable solution based of miced room measurements? I did not say that REW failed in providing proper solutions, sorry if you had that idea.

I feel that since you are the one making a claim about what @porgre should do, and disputing sight unseen what he has done, even while admitting that you don’t have full knowledge what he has done, that YOU should post said screen shots of YOUR processes. Start with your own solution path.

IIRC this subject has been broached in two separate threads where, yes, I have been quite critical of your claims/statements/ and overall approach, but I would hardly call two threads a whole forum. Some of my criticisms have indeed been too hostile, and I apologize for that.

 

[Pogre]

Then let's use your conveniently 'sealed rec room' as 'exhibit A'.

My predictions for your room:

Length: 17.5 feet = 64 hz

Width: 18.5 feet = 61 hz

Height: 8.5 feet = 133 hz

Still need to know the front wall to the back wall of the storage area as that longer dimension will yield an even lower frequency.

As I recall, REW didn't get this right, which is why you came to me in the first place.

So the choice is yours, post shots of REW accurately predicting and recommending solutions in your room and how everything I've said is a waste of time, or continue chasing me across the forum claiming I'm a knowitall jack@@@ that is stubbornly trying to help people in a way that you don't like.

@Pogre

"To compute the frequencies at which axial standing waves occur, simply measure the distance between the walls, and divide that number into the speed of sound in whatever units the measurements were done." 'Sound Reproduction'

That reads the same no matter which edition you have. I'm trying to help you by singling out this one item from the entire chapter to help guide you in making informed decisions. And you are ignoring it because I cant succinctly explain all of the variables involved in the decision process. Perhaps you'll notice that it took Floyd 50 pages just to write that chapter. I'm not here to offer you cliff notes, just lending a hand with a very basic first step. You've made it clear you don't want my help, so I'll stop offering it.

Let's say I measure everything in my room and you predict all of my room modes. Then I measure with rew and the graph correlates with the predictions. What would I do differently outside of eq'ing the highs and lows? My subs and towers are pretty much stuck where they are.

 

[Pogre]

It may be lost in the earlier posts and if so, I'm sorry for being lazy. Do you have measurements (with no Audyssey or miniDSP correction engaged) showing each sub and tower separate? Can you or ATLAudio direct me to them please?

I don't have anything with everything measured separately. Like I said, I'll be redoing everything this weekend and plan to be even more thorough this time.

The green line is with no room eq or the mini. It's an older graph, but probably still pretty accurate.

This is with my house curve.

 

[Pogre]

The Z axis, marked 0-300ms are time slices. If you look closely, the line at 0ms corresponds with this,
View attachment 22470

Every subsequent "slice" on the waterfall represents the "FR" (residual sound energy plotted as frequency vs time) for that slice after the impulse tone was played.

What makes a good looking waterfall plot?
1) No frequency bands showing excessively long decay time. (Example, a persistent bump along all slices from 0-300 with minor drop in level.)
2) All frequencies showing even energy decay. (Even decay time results in "parallel slices".)

Have you tried the ETC plot? It is way easier to interpret than waterfall plots at the expense of losing amplitude resolution.

If you post the ETC plot for the above measurement, I'll correlate the three.

The only plot I'm confident in understanding is the spl only. What you describe about waterfalls is what I had inferred, but have been told to adjust time/phase(?) for a more accurate one. I can post more graphs (ir, phase, etc). I generated them, but don't understand them all.

 

[William Lemmerhirt]

Let's say I measure everything in my room and you predict all of my room modes. Then I measure with rew and the graph correlates with the predictions. What would I do differently outside of eq'ing the highs and lows? My subs and towers are pretty much stuck where they are.

@TheWarrior, this is the question that has been asked before(by me also). It’s not that pogre isn’t asking, see quote above. It’s that there hasn’t been a clear solution path. Even an arbitrary one would do. For example, if you have a low tire, connect air supply, and inflate to proper pressure level. So, knowing where the modes are when you can’t move your LP, or gear will be addressed by doing what? That’s the question. A solution path with arbitrary numbers is fine. We’re just looking for the path.

 

[PENG]

What is your LFE crossover? 80 hz is 10dB down from 20 hz.

That was subs only after running Audyssey. I assume Audyssey did the roll off in trying to integrate them with the other speakers. There are also dips at around 50 and 70 Hz due to the room, that Audyssey improved, but could not do more.

 

[agarwalro]

View attachment 22512
This is with my house curve.

This is why people on AVS cut themselves.

 

[William Lemmerhirt]

so, I'm not trying to muddy the waters but I told J I would share of my progress. this is post audyssey with +5 to the sub trim, and one with deq on top of that. It sounds great but I'm a little worried about destroying my subs with that much energy below 20hz. To my eye, i'm pretty flat to about 15hz which is great but thinking of putting a high pass filter down there. Maybe around 19hz where my subs are tuned.
Also, I tried posting my mdat file for anyone to peruse. It says the file is too large. ideas? I would be interested in someone going through it to provide insight. especially around that 50-60hz dip

here's a waterfall just for fun

 

[ATLAudio]

Here’s my waterfall

 

[William Lemmerhirt]

Here’s my waterfall

View attachment 22521

Haha! Got room for me?

 

[Mitchibo]

Here’s my waterfall

View attachment 22521

Probably some ringing going on inside that barrel. I’d check out Toole’s book on barrel/waterfall graph issues.