AVS Forum | Home Theater Discussions And Reviews (https://www.avsforum.com/forum/)
-   Dedicated Theater Design & Construction (https://www.avsforum.com/forum/19-dedicated-theater-design-construction/)
-   -   Room EQ questions for the average joe equipment (https://www.avsforum.com/forum/19-dedicated-theater-design-construction/2926754-room-eq-questions-average-joe-equipment.html)

HalfSpec 10-17-2017 05:44 AM

Room EQ questions for the average joe equipment
 
Hello

As I approach the final stretch with my budget theater build I find myself questioning how folks here are making their EQ adjustments (questioning how the process is done, not questioning people's decisions) so I wanted to throw this thread up to try and get some clarification on the processes / equipment people are using. I am an electrical engineer, but my career doesn't center around audio, so I'm always a little rusty and if I goof up terms, you have my apologies in advance.

Like a lot people in their 30s, I've owned cars much longer than I've owned a home or a home theater so my experience with car audio predates my theater build by almost 2 decades. I've managed to stay fairly up to date with advances in car audio, especially when it comes to tuning the response of modern systems, so I'm familiar with REW or RTA measurement, and making DSP adjustments to get crossovers set, response time aligned, and ideal curve ~matched in multi speaker setups.
Coming into HT audio, I expected to see a lot of the same tools in place, but aside from a healthy dose of REW / RTA usage, I haven't seen much documented in this subforum concerning the actual adjustments made on the AVR side. I.E., discrete channel DSP adjustments for crossovers, time alignment, and frequency response. This has created one big question I've been wanting to clear up. On average, do normal folks rely on their AVR's built in automated room EQ systems to correct their room's responses after making changes (I.E., room treatments, sub moves, etc) and REW / RTA measurements are only used to compare before and after said change(s)? OR, have I completely missed (or just not looked hard enough) the fact that everyone doing corrections has the type of high end equipment to do discrete channel EQ corrections / time alignment using premium AVRs, preamps with DSP's in-line with their discrete amps, or perhaps or piggyback / man in middle devices like miniDSP's nanoAVR?

Hopefully that frames the question correctly. I'm asking because I'm using an Onkyo AVR that's probably 8 years old with the baseline Audyssey room EQ system for that generation which is probably only good for rough level matching and time alignment so I've been looking at upgrading. As I look at offerings in my price range I'm not seeing what I was expecting to see. It seems like most AVR's hand over all of their adjustment capabilities to auto EQ systems like XT32 or Dirac Live. Maybe that's just because I'm looking at AVRs under $1000, but then again, I've been looking at older used higher end equipment and haven't seen much better.

Am I blind or just lazy?

Lane

BIGmouthinDC 10-17-2017 05:58 AM

read about the capabilities of this unit for $700. (last years closeout) AutoEQ has come a long way since your unit was manufactured. I just picked this up but haven't installed it yet.

https://www.accessories4less.com/mak...eceiver/1.html

Audyssey MultEQ Editor App
The latest Denon audio video products use Audyssey MultEQ for simple, accurate set-up and calibration of your system to the room in which it's used. But now you can go further with the Audyssey MultEQ Editor app, going "under the hood" to view and adjust settings for detailed tuning - allowing you to customize the sound more precisely to the specific problems in your room, and tailor the sound to your personal preferences. With this comprehensive app, you can harness the power of Audyssey MultEQ to take total control of the way your home cinema sounds.
https://assets.denon.com/assets/imag...0317055713.jpg
This app will allow you to:
  • View the speaker detection results, to check correct installation.
  • View before and after results of the Audyssey calibration, making it easy to identify room problems.
  • Edit the Audyssey target curve for each channel pair to suit your tastes.
  • Adjust the EQ frequency adjustment range for each channel pair.
  • Switch between 2 high frequency rolloff target curves.
  • Enable/Disable midrange compensation to make the vocal region brighter or smoother.
  • Save and load calibration results.

https://assets.denon.com/Assets/Room...ction_Ipad.jpg

HalfSpec 10-17-2017 08:03 AM

Thanks for the reply Jeff.

You've definitely landed close to the corner of the market I was looking at. Up to this point I'd been considering the AVRX3300W which, like the AVR-X4300H, comes with the MultiEQ editor. Now, seeing the prices of the AVR-X4300H b-stock has me reconsidering. This is also helping to answer my question which seems to be, invest in newer equipment with better EQ/DSP options.

My only complaint about Denon's offerings and MultEQ Editor's abilities is that it still seems like it's not full featured. It gets 80-90% there, but I'm still not seeing the level of adjustment that I'm used to seeing. Primarily, response corrections are limited to speaker pairs, not discrete channels, or at least, that's what I'm getting from the spec sheet / manual / youtube reviews. This simplifies things for most and most likely works flawlessly for most, but for setups like mine that aren't exactly symmetrical (my front right channel is in an alcove) discrete channel eq is something I'm looking for and haven't found in my price range yet.

That isn't to say the AVR-X4300H / AVRX3300W aren't among the best bang for the buck options I've come across. Like I said, it's almost what I'm after, so it's probably what I'll get when my budget allows.

I'm still curious to see how the majority handles tuning their AVR to their room. Do you leave it up to the auto room EQ after making changes or do most dig deep and put together solutions that allow full manual corrections?

Lane

Mpoes12 10-17-2017 09:17 AM

Hi Lane,

First, as I understand Audyssey does not eq in pairs. It’s a FIR filter based correction system that separately shapes the response of each speaker. It is considerably more complex than the basic types of eq we normally use. In general it is a bad idea to use minimum phase filters to fix the response of a speaker above the transition frequency of the room, as amplitude and phase/timing are not locked. You can fix the amplitude only to muck up the phase. What this means is that it is impossible to use a minimum phase filter set to fix the response of a speaker as it will be only right in a tiny space, and wrong everywhere else. The fix to this is to use a special inverse impulse filter that corrects the response for both phase and amplitude and relies on a weighted average over a larger listening window. This is only possible with FIR filters. That means that it is not possible to understand the filters as a simple set of frequency, Q, and amplitude sets. That means it is not possible to give average users control over the filters or even to explain the filter. We typically see responses as peaks and dips in the amplitude and most people ignore the timing/phase issues. An FIR filter is best displayed as an impulse response and most people can’t read those usefully.

The new editor is a great addition because a lot of people don’t like the factory set target curves. It also gives us a lot more information about what is going on with each speaker. Unlike DIRAC, it was hard to know what curve was being used, what corrections made, or how it measures our speakers. While the filters can't be seen, we can at least see before and Afters.

The one problem with Audyssey is that it uses a filter type that is very processor ineffecient at low frequencies. FIR filters are measured in taps and it takes more taps the lower in frequency. While Audyasey restricts the tap usage by octave to avoid the problem I'm about to mention, it none the less it limited in the resolution it can control at low frequencies. Since low frequencies are minimum phase and steady state (the same everywhere in the room, all variations can be attributed to room reflections and resonances) you CAN use minimum phase filters. It allows for correction of time and amplitude correctly and by the laws of minimum phase it also automatically fixes ringing, the main problem with modal resonances. It isn't perfect and in fact well below the schroeder frequency there is evidence in small rooms of stochastic behavior meaning what I said won't be true, but it's generally ok to accept it as true. Just don't think EQ can replace room design, speaker position, or acoustic treatment.

What this all means is that for some of us we can support Audyssey's own corrections by making our own corrections first. This gives a massive effective boost to our rooms of correction filters and allows a far flatter response. It also allows EQ of response areas too low for Audyssey to efficiently touch (some people like a room that is flat to 5hz for example). Many of us use amplifiers with dsp or external dsp like Mini dsp. We generate filters in various ways, but I think the most common is to use REW's ability to generate filters for us with amplitude, Q, and Frequency. We then plug them into our dsp and go. To do this we need to take measurements and design our filters from that. I think everyone has their own preferred approach that makes them happy. There are a lot of tutorials of how to do this. I personally use one measurement at the MLP for all distance alignment and use the alignment of the impulse response to do this. I use averaged measurements of my MLP and other listening positions to set eq. That way I apply eq only to those frequencies that are consistent over a wide range and avoid eqing things that are very position dependent.

After I have set the distances and alignment using my dsp and the receivers built in bass management I then run Audyssey. I then usually undo the automatic bass management settings it comes up with and restore them to what I want. I take more measurements. I then do final eq to tweak. Take final measurements and call it a day.

Big take away, full manual correction is my preference at low frequencies, but nobody should be really touching the frequencies above say 200hz with eq unless they really are experts in filter design and really know what they are doing.

For what it’s worth Ealr Geddes and Floyd Toole are both huge proponents of room eq for bass but both think that any form of room correction above the bass range is a bad thing. They strongly believe that it is simply not possible to correct for the room at those frequencies and that speaker design, setup, and room treatment are what matter. I have found problems with Audyssey correction in the midrange and do think my speakers sound better and image better with it off. I don’t find that true of DIRAC. Worth noting that I don’t have XT32 which is supposed to be a big improvement.


Sent from my iPhone with a keyboard that predicts what I mean very poorly

HalfSpec 10-17-2017 09:59 AM

Quote:

Originally Posted by Mpoes12 (Post 54968436)
Hi Lane,Sent from my iPhone with a keyboard that predicts what I mean very poorly

Well now. I owe you an even bigger thanks for your very helpful reply knowing you banged it out on an iphone! :p


Quote:

Originally Posted by Mpoes12 (Post 54968436)
First, as I understand Audyssey does not eq in pairs.

I didn’t mean to insinuate that Audyssey EQs in pairs. I was simply saying that it appears manual adjustment is only allowed in pairs via the MultiEQ editor. Sorry for the confusion.



I believe you’ve answered just about all of my questions simply by explaining the shortcomings of minimum phase filters. I didn’t realize how complex the side effects could be with unlocked phase/timings above the room’s transition freq. I do know that in a comparably tiny automotive cabin the same issues are present, but it’s generally not severe and the car owners just live with it and let the passenger listeners deal with minor phase/timing issues. I just didn’t extrapolate those problems to a larger room / take the time to realize it’s an issue that compounds with more area. I started this with the understanding that minimum phase filters could take me the distance and was frustrated and confused why the tools I was looking for didn't exist but now I understand the shortcomings and understand a FIR filters a little better now or at least understand why they are preferred. Thank you very much.


Your approach to using Audyssey for 100Hz+ and full manual correction for sub frequencies makes a lot of sense and definitely sets me at ease. I feel like I can handle manually correcting low frequencies, but I would be lying if I said I wasn’t intimidated at creating minimum phase filter sets for main / surround speakers.

In summery you’ve given me a path with enough options / flexibility that I feel happy with it and think it can take me where I need to go. Originally I created this thread to try and drum up a lot of response so I could determine my options and see what the majority of people are doing, but I don’t think that’s needed anymore :D


I appreciate the help. Really. The help I received from you and Jeff is going to save me hours of time I would have spent floundering before I figured out a direction to hone in on.

Lane

Mpoes12 10-18-2017 04:32 PM

That’s great Lane, glad I could help.

By the way did my reply disappear? I can’t see it anymore.

You can use PEQ to fix the response of the speaker at any frequency but not to correct room distortion. It still potentially has flaws compared to the best FIR filter solutions, but a good case can be made for the benefits of minimum phase too. No pre-ringing, less processor intensive, arguably no benefit to sound over min-phase.

If you can manage to measure your speaker anechoicly then you can adjust the PEQ to correct problems in the response of the speaker itself. This is actually done in many speaker crossovers using lcr filters known as notch or trap filters. It’s better to do it through eq before the amp because you don’t waste amp energy, impact the damping factor as much, and it’s far more precise. You could have 100 eq filters if you wanted. 100 trap filters would be expensive and cost a fortune.

Here is a pick of one of my speakers crossover. It has three lcr trap filters, two types of shelf filters, and a notch filter in the 2nd order high pass 1 octave below the crossover point. The speaker is 14”x30” as a sense of scale doe that crossover. It’s just a 2 way crossover too.

https://uploads.tapatalk-cdn.com/201...33e64f2fb9.jpg


Sent from my iPhone with a keyboard that predicts what I mean very poorly

HalfSpec 10-19-2017 06:12 AM

Quote:

Originally Posted by Mpoes12 (Post 54977724)
By the way did my reply disappear? I can’t see it anymore.

No, it's still here. Or at least I can see it above.

Quote:

Originally Posted by Mpoes12 (Post 54977724)
If you can manage to measure your speaker anechoicly then you can adjust the PEQ to correct problems in the response of the speaker itself.

Unfortunately I'm fresh out of anechoic chambers in Starkville, MS. My plan is still to try and buy a trio of DIYSG 88 Specials for my front sound stage and maybe a HSU VTF-15H or two. As much as the EE inside me would love to sit in my shop and design and build crossovers (actually... I wouldn't... see below), family life is not going to allow that. Aside from a quicky sanity check, I'm afraid I'm just going to have to trust that DIYSG and HSU did their homework on their crossover design and EQ for the room distortion.

Quote:

Originally Posted by Mpoes12 (Post 54977724)
It’s better to do it through eq before the amp because you don’t waste amp energy, impact the damping factor as much, and it’s far more precise.

Agreed. This was another thing that disappointed me when moving over to HT audio from mobile audio. In a car, you just get a minidsp C-DSP 8x12 and you've got 12 discrete output channels you can tinker with via the DSP to your hearts content. I'd gotten very used to the tech and not needing physical crossovers... Now I'm seeing that the clunky bastards are back with a vengeance in my theater :p

Thanks again for the help

Lane

Mpoes12 10-22-2017 10:53 AM

Quote:

Originally Posted by HalfSpec (Post 54980164)
No, it's still here. Or at least I can see it above.



Unfortunately I'm fresh out of anechoic chambers in Starkville, MS. My plan is still to try and buy a trio of DIYSG 88 Specials for my front sound stage and maybe a HSU VTF-15H or two. As much as the EE inside me would love to sit in my shop and design and build crossovers (actually... I wouldn't... see below), family life is not going to allow that. Aside from a quicky sanity check, I'm afraid I'm just going to have to trust that DIYSG and HSU did their homework on their crossover design and EQ for the room distortion.



Agreed. This was another thing that disappointed me when moving over to HT audio from mobile audio. In a car, you just get a minidsp C-DSP 8x12 and you've got 12 discrete output channels you can tinker with via the DSP to your hearts content. I'd gotten very used to the tech and not needing physical crossovers... Now I'm seeing that the clunky bastards are back with a vengeance in my theater :p

Thanks again for the help

Lane



No problem Lane.

You can approximate an anechoic response by filtering out the reflections through gating. That could be used as a means to establish eq filters. The problem is again that you need to know what to look for and how to do it. You don’t want to mess up the power response or directivity index by applying eq such that it’s good at one angle but not others. That means you need to take many measurements around the speaker to ensure it’s response at various angles remains smooth (or rather becomes smoother).


Sent from my iPhone with a keyboard that predicts what I mean very poorly


All times are GMT -7. The time now is 08:42 PM.

Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
vBulletin Security provided by vBSecurity (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.