Audyssey Room EQ Review

[Lbstyling]

I've done quite a few tests in the last couple of days, and I'm now thinking that keeping Audyssey corrections at least up to 1KHz gives good results. I've tried with 3KHz and more, but I'm really undecided about the best threshold to use, and I can't do anything conclusive until I get my Umik mic and do some room measurements. Without being able to do measurements, the method I'm using is to compare stereo tracks with the same tracks played back through my SennHD600 headphones, primarily to check the tonality. A bit basic as a methodology, but that's what I've got at the moment. Audyssey corrections up at 1KHz also give good results for surround music. The system sounds more balanced and immersive than simply correcting up to 500Hz/600Hz.

Overall, I'm sort of happy with the results (better than no EQ), but I'm dissatisfied and frustrated by the process. Drawing curves on the app, then uploading settings on the receiver, listening to results and guessing what's happening. One step more if I had a mic (measuring results). That's very time consuming, and I've spent many hours to get a decent result. I will immediately try/switch to Dirac as soon as I change my AVR. It's so much praised that it has to be better in the process and results.

DIRAC.... maybe. It's very clever in its ability to reduce RT60, but I'm not entirely convinced that should be the aim.

Personally my best results (KEF meta 7.4.4) are with MEQX, with EQ up to between 1khz and 4khz. I add manually entered EQ correction against spinorama data for my speakers from ASR data above this for each speaker from Pierre.

Anechoic EQ for loudspeakers

www.spinorama.org

Subs are optimized through MSO/ miniDSP with MEQX off for these channels.

Target curve is flat to 4hz, but with dynamic EQ on.

 

[peng]

I've done quite a few tests in the last couple of days, and I'm now thinking that keeping Audyssey corrections at least up to 1KHz gives good results. I've tried with 3KHz and more, but I'm really undecided about the best threshold to use, and I can't do anything conclusive until I get my Umik mic and do some room measurements. Without being able to do measurements, the method I'm using is to compare stereo tracks with the same tracks played back through my SennHD600 headphones, primarily to check the tonality. A bit basic as a methodology, but that's what I've got at the moment. Audyssey corrections up at 1KHz also give good results for surround music. The system sounds more balanced and immersive than simply correcting up to 500Hz/600Hz.

If you read Dr. Toole's comments on REQ/RC, he explained that once you get pass the room transition frequency, that is often in the range 200-300 Hz depending on the room (small to medium large I assume, that most of us ordinary home users are in), the kind of mics used would not be able to do their job accurately and/or consistently. That's just my words and interpretation, you should read it yourself in case I interpreted what I read/heard wrong. So if you let the RC/EQ whether it be Audysssey, Dirac or others, you may like the effects, or you may find the results worse than without RC/EQ. The best way it to try them all out, with the use of even the $20 app, you can store as many target curves as you want and load a different one every day, until you find the one that does the best for you.

Overall, I'm sort of happy with the results (better than no EQ), but I'm dissatisfied and frustrated by the process. Drawing curves on the app, then uploading settings on the receiver, listening to results and guessing what's happening. One step more if I had a mic (measuring results). That's very time consuming, and I've spent many hours to get a decent result. I will immediately try/switch to Dirac as soon as I change my AVR. It's so much praised that it has to be better in the process and results.

If you are frustrated, why not bite the bullet and learn how to use Ratbuddyssey, it is quite simple to use:

GitHub - ratbuddy/ratbuddyssey: Audyssey .ady file editor

Audyssey .ady file editor. Contribute to ratbuddy/ratbuddyssey development by creating an account on GitHub.

github.com

 

[Acerun]

I've done quite a few tests in the last couple of days, and I'm now thinking that keeping Audyssey corrections at least up to 1KHz gives good results. I've tried with 3KHz and more, but I'm really undecided about the best threshold to use, and I can't do anything conclusive until I get my Umik mic and do some room measurements. Without being able to do measurements, the method I'm using is to compare stereo tracks with the same tracks played back through my SennHD600 headphones, primarily to check the tonality. A bit basic as a methodology, but that's what I've got at the moment. Audyssey corrections up at 1KHz also give good results for surround music. The system sounds more balanced and immersive than simply correcting up to 500Hz/600Hz.

Overall, I'm sort of happy with the results (better than no EQ), but I'm dissatisfied and frustrated by the process. Drawing curves on the app, then uploading settings on the receiver, listening to results and guessing what's happening. One step more if I had a mic (measuring results). That's very time consuming, and I've spent many hours to get a decent result. I will immediately try/switch to Dirac as soon as I change my AVR. It's so much praised that it has to be better in the process and results.

If you can afford it, the desktop version is spectacular. The new version of it also enables additional headroom and other really nice features that are pointed out in the video. It's a breeze to make adjustments and then try out your adjustments. I add a tilt to my settings as well which I like quite a bit. Also has PEQ if you want to make adjustments to particular frequencies or lay in a shelf.

 

[Acerun]

You can use offset setting to lower the bass, try 10 or even 15.

DEQ is calibrated for THX movies, if you listen to other source contents you should play with the offsets to get it closer to what you prefer.

For whatever reason I decided I didn't want dynamic EQ and I think that was because it made things too mid bass boomy but I didn't realize I could back it off. My system came alive today next level when I turned on dynamic EQ and then adjusted the setting to 15. Thank you.

 

[Chromatischism]

I've always used my Denon x6400h without Audyssey, with manual calibration of speaker distances and levels, but earlier today I decided to give it a try, after reading this positive review. I will soon upgrade to a Marantz AV10, so it's the right moment to do testing. I am impressed with the results, particularly with 5.1 and Atmos music (I have a 5.0.2 system). So, with surround music listening, I'm 100% sold on using Audyssey. Some of my demo albums are:
- Pink Floyd's Dark Side in Atmos
- Pink Floyd's Animals in 5.1
- Kraftwerk 3D in Atmos
- Tears for Fears' The Hurting in Atmos, Mad World is great to calibrate all speakers
- Gentle Giant's Free Hand in Atmos
- XTC's Black Sea in 5.1 (Living through another Cuba is a demo surround track)
- Steely Dan's Gaucho in 5.1 (to check how main and back vocals gel together)
- Santana's Abraxas in Quad
These are all good albums to check how discrete surround music sound. Without Audyssey, okay, with Audyssey ten times better.

I used the app on the iPad, and I followed Amir's instructions to correct the target curve (not flat; see Amir's target curve in post #1) and also to remove the midrange compensation. I did six measurements instead of eight (I have a wall behind the couch).

Glad to hear you're getting good results so far. I didn't see what speakers you had—some speakers should have midrange compensation on, and with others it should be disabled.

I used the app on the iPad, and I followed Amir's instructions to correct the target curve (not flat; see Amir's target curve in post #1) and also to remove the midrange compensation.

In my experience, adjusting the curve is never necessary. Keep the Reference curve at 100% default, turn Dynamic EQ on, and adjust the filter curtain to 300-500 Hz.

Another question: when I upgrade to the Marantz AV10, would it be wise to use Dirac instead?

This will come down to how much you value 1) how quick Audyssey can get great results and 2) Audyssey's other features such as Dynamic EQ and Dynamic Volume. To me the former is mandatory 100% of the time and the latter is situational but useful. This combined with no real dual sub solution is why Dirac is not for me.

A final consideration: with stereo listening, things have also improved a lot. The difference between Audyssey engaged and disengaged is obvious (Dynamic EQ is off in my case). Still, it sounds so different from what I'm used to that I am still unsure whether Audyssey Reference (with Amir's target curve) or Direct is better. I think Audyssey might be better, but hard to A/B test since the volume changes as one switches on and off.

The way you should go about testing this is to limit corrections to 300-500 Hz. Just let it fix your bass and you'll know straight away how powerful the FIR filtering is. I have never had a setup that wasn't dramatically improved in this region. Your mileage will vary up higher because every speaker is different and Audyssey can either compliment your speakers or make them sound worse due to the interaction between your tweeters and Audyssey's Reference Curve. It's a tool and you need to know your equipment and how to use the tool with it. The best bet is to use speakers that measure well, position them well, and cut the corrections off sooner so you don't need to touch the tweeters at all. This is why my Buchardt S400 MKII's sound so good with Audyssey. I have SBIR minimized through placement then corrected to 400 Hz and nothing above that, and dual subs smoothing out everything below 80 Hz.

IMO if you find yourself endlessly playing with EQ above 1 kHz to get the right sound, it means you haven't yet found a pair of speakers that you truly love the sound of.

 

[Chromatischism]

It's great that you tried it and liked it. Had you not even tried it before?

I'm pretty pissed at Audysey for not updating the app to make drawing the curve easier. There is no excuse for this. People have been complaining about this for years.

Complain to Sound United—they made the app!

 

[Chromatischism]

I think the audioholics advice is out of date.
Audessey level matches speakers at 1khz, so you want to go to at least this.

The main guy at audioholics admitted that post a/b testing with experts he has his set at about 4khz. You want it to EQ up to the point that the natural curve diverges away from the room curve you want.

This is going to be different in every system.

For me, sound quality peaks when I cut off at 400 Hz on my current main speakers, and is worse if I continue to run the Reference curve above that. I'm a very picky listener and play songs many, many times to arrive at my values. And of course I measure.

For my surround speakers that number is 600 Hz. In other systems it may well be 1000 Hz. It's always different, but there is no rule that says you should run it to 1 kHz just because of level matching.

 

[LeoB]

Do not use any "traditional" house curves with Dynamic EQ. They exaggerate bass too much. For me, anything higher than about +2dB at 20-30Hz is too much when using Dynamic EQ.

 

[peng]

Do not use any "traditional" house curves with Dynamic EQ. They exaggerate bass too much. For me, anything higher than about +2dB at 20-30Hz is too much when using Dynamic EQ.

Sure, but then it is subjective, hence the term "house".

 

[LeoB]

Unfortunately, many people do not understand that it is subjective and try to EQ their systems to house curves from the start, then they turn on Dynamic EQ and wander why it does not sound right. I learned from my personal experience that this is a big mistake.

My current eq routine is following:
1. Eq to flat with MultEQ-X and REW.
2. Turn on Dynamic EQ.
2. Gently roll off frequencies above 2500Hz to taste.
3. Start tilting the whole range in 0.1db/oct increments until you fееl that there is too much low frequencies.
4. Go one step back, mark this point and from now on compensate further tilt with LF roll-off below 250Hz.
5. Keep tilting until vocals sounds wrong and go one step back.
6. Adjust HF roll-off to taste.

As a result, you'll create your own house curve, similar to the "traditional" ones, but customized to your particular system and taste.

 

[Acerun]

Do not use any "traditional" house curves with Dynamic EQ. They exaggerate bass too much. For me, anything higher than about +2dB at 20-30Hz is too much when using Dynamic EQ.

This helped me quite a bit. I've removed my tilt and kept my low shelf and it has helped control the bass with Dynamic EQ set to 10.

 

[Acerun]

Unfortunately, many people do not understand that it is subjective and try to EQ their systems to house curves from the start, then they turn on Dynamic EQ and wander why it does not sound right. I learned from my personal experience that this is a big mistake.

My current eq routine is following:
1. Eq to flat with MultEQ-X and REW.
2. Turn on Dynamic EQ.
2. Gently roll off frequencies above 2500Hz to taste.
3. Start tilting the whole range in 0.1db/oct increments until you fееl that there is too much low frequencies.
4. Go one step back, mark this point and from now on compensate further tilt with LF roll-off below 250Hz.
5. Keep tilting until vocals sounds wrong and go one step back.
6. Adjust HF roll-off to taste.

As a result, you'll create your own house curve, similar to the "traditional" ones, but customized to your particular system and taste.

I like your process. For me I feel like I need extra energy at 100 based on the filter graphs (I am crossing over at 100), so I added a +3 db low shelf. I previously had a -1 tilt as well but was battling some subtle boom demons.

 

[Acerun]

I like your process. For me I feel like I need extra energy at 100 based on the filter graphs (I am crossing over at 100), so I added a +3 db low shelf. I previously had a -1 tilt as well but was battling some subtle boom demons.

This has made me question whether I have been boosting enough or the right way if +2 is the target (I have been closer to 0). Below is my Audessey graph with a +6 boost (instead of +3), crossing to subs at 100. Does this look right to you? I'm doing some listening.

 

[Reverend Slim]

Do not use any "traditional" house curves with Dynamic EQ. They exaggerate bass too much. For me, anything higher than about +2dB at 20-30Hz is too much when using Dynamic EQ.

Or just don't use DynamicEQ, especially if you have Atmos. Like it or not, what it does with surround/height levels completely changes where objects are supposed to image in the room, and on a sliding scale that means imaging will get shifted by varying amounts. Plus, it's just generally inconsistent since content is inconsistently mastered, and especially with streaming, often in the opposite direction of what the DEQ offset can account for.

They really need to split the surround presence and bass portions of DEQ. Its pre-immersive development doesn't reflect how the immersive formats work, and no further research was done to account for it. Audyssey just shoe-horned in similar boosts to the heights that they were supplying to the surrounds and called it a day.

 

[LeoB]

To Acerun

It looks like you need to check Disable Auto-Leveling boxes for all speakers at Design Target Curve screen.

 

[LeoB]

Or just don't use DynamicEQ, especially if you have Atmos. Like it or not, what it does with surround/height levels completely changes where objects are supposed to image in the room, and on a sliding scale that means imaging will get shifted by varying amounts. Plus, it's just generally inconsistent since content is inconsistently mastered, and especially with streaming, often in the opposite direction of what the DEQ offset can account for.

They really need to split the surround presence and bass portions of DEQ. Its pre-immersive development doesn't reflect how the immersive formats work, and no further research was done to account for it. Audyssey just shoe-horned in similar boosts to the heights that they were supplying to the surrounds and called it a day.

You are absolutely right; however, I don't think Denon/Maranz will ever fix this.
In my case, I play everything from my PC and do not use Atmos speakers. So, my media library is normalized and I offset surround channels boost by fixed -5dB.

 

[Acerun]

To Acerun

It looks like you need to check Disable Auto-Leveling boxes for all speakers at Design Target Curve screen.

They are disabled

 

[LeoB]

Put the check-marks in Disable Auto-Leveling boxes.

 

[Acerun]

Put the check-marks in Disable Auto-Leveling boxes.

Did it. OK, I've added one of my fronts and my subs below. Thanks for any input. Do these look reasonable? I tried to match the crossover at 100 with the subs at +2

 

[LeoB]

Yes, this looks reasonable.
Why do you limit Auto EQ so much? If I were you, I would use about 40Hz LF Limit for the speakers and about 250-300Hz HF Limit for the sub.