"Official" Audyssey thread (FAQ in post #1) - Page 289

I'm sorry to barge in on the conversation with my unrelated blubbering, but I'm afraid my question will go unanswered so I'll repeat it.

I just read that you should be using Audyssey flat for movies that contain near sound field mixes. If that's that case, shouldn't we also be using flat for CD's and videogames?

If we are to get into the business of quotes from Gene, then I have one for you from an email he sent me last year. The comment you have above is 4 years old and refers to the AVR5805. MultEQ is three generations older now so that comment has little relevance. We can go on forever, but I assure you I have seen the code. There is no added boost that damages speakers at low frequencies.

Below is an excerpt from Gene's email and the measurement using MultEQ in his room (Velodyne subs).



Chris;

This is the in room bass response I am getting in the front and back row of seating with the T30 system + 2 Velodyne DD-15s all run through Audyssey Mult EQ Pro. This it the best I've ever gotten the bass in my system to sound. Also note the T30s are running stereo bass with the bottom portion crossed over at 80Hz and the top portion running full range.

Notice the extension below 20Hz! In the front row there seems to be nothing I can do about the 80Hz suckout which I think has to do with the width of the room. L shape room 27 feet wide up front and 17ft wide in back. Other than that, it sounds fantastic!

Best Regards;

Gene DellaSala

Hi,

Yes, if the movie mix was redone for the DVD it is likely better using Flat. The problem is that there is no information given about the curve used in the mix when they do that. Similarly, for CDs. Many are mixed to the same curve as film, but others are mixed to flat. Videogames??? No idea.

So, what we recommend is to use the standard Audyssey curve for most content. If you are in THX mode, then use Flat and let THX re-EQ do what it is supposed to do.

I have fixed the sub phase and recalibrated, bass is all good now, however i've just had a look at the eq graphs on the denon and the centre channel shows no correction past 1khz. Other channels show full range correction.

Hakka.

Thank you so much for your help.

Nevermind, it seems to be a bug in the denon ui. If i flick through the graphs sometimes the centre shows full range correction, sometimes it only shows to 1khz, sometimes to 6khz.

Hakka.

Thanks Chris, that quote was from 2 years ago and I had thought that it may be irrelevant to current implementations


I don't know what else to say but there is a definite boost once Audyssey is engaged which I have shown in the previous graphs.
But as you mentioned earlier this is not due to the filters audyssey uses but the level correction for perceived equal loudness of audyssey on or off, which in my case it seems the sub overall level was increased by 10db and mains by 3db to compensate for the filters, which has the net effect of boosting anything that had no filter adjustment. Is this correct, if not is there something else going on?

cheers

Does anyone have a list or know where one is for AVR's that have Audyssey dynamic EQ and volume?

Thanks,

Hi Pete,

Here you go:
http://www.audyssey.com/products/index.html

That's a new one... I will forward to Denon. Thanks

Yes, no more BFDs, only Audyssey. When I look at my prior graphs with the BFD compared to Audyssey, my BFD graphs appear a bit flatter. But overall the Audyssey sounds so much better, probably because of EQing the upper bass and high freq vs me only taming the sub range before.

Edit: I should add that my measurements were done with REW and RS SPL meter, which may not be the most accurate.

Is it possible to run Audyssey setup without a sub connected, then add one afterwards?

The reason I ask is so I could use a BFD for the sub and allow Audyssey to correct my main speakers giving me the sound I like, but protecting my sub from the low boost I seem to suffer in my setup. I could manually put the crossovers to the value set by Audyssey when setup was run with the sub connected. These were 40Hz centre, 60Hz fronts, 120Hz side surrounds and 150Hz rear surrounds if it helps.

Chris, I ran some more tests using the 8 positions(roughly) that I used for the Audyssey auto setup and averaging them to show the difference between Audyssey on and off(without changing master volume) with the filter trace there to see what they have done to the measured response
All traces have 1/3 filtering applied


Blue = Audyssey filters (from previous graph)
Green = Audyssey off (8 position average)
Red = Audyssey on (8 position average)

The results of Audyssey are pretty impressive given the pre and post graphs

That 65Hz is caused by the 2nd axial room mode(and the 35Hz bump being the 1st axial mode) which I would correct but alas this is a lounge room setup and don't have any options for placement and treatments.......but won't be a problem once we have a dedicated room.

BTW that increase down around 5Hz is actually audible due excessive distortion, ouch, reason enough to keep those frequencies from your mains, eh

cheers

I have a Denon AVR 2809CI and was wondering if I should turn ON Dynamic Volume for Watching Blu-Ray movies or while listning to CD's ? Sound level in these play back modes are not a concern. I do have Dynamic Volume set to "DAY" for Tuner and Tivo/TV play modes. Also is the DAY volume level the same as if Dynamic Volume were turned OFF ? I'm mainly concerned that if watching a movie and Dynamic Volume were turned ON (set to DAY) I may lose some audio impact (explosions and other noises associated with an action movie). I guess what I'm asking are the benefits of Dynamic Volume ON or OFF while watching a Movie or listening to CD music. Again volume or db level is no problem.

THANX - Joe

I am relieved to see that


Yes, definitely do add treatment before running MultEQ (if it's practical). I must admit that I've never seen the 2nd mode cause a problem bigger than the 1st mode. To me, this huge bump you have at 65-70 Hz is due to some sort of corner or cabinet loading. Is the L speaker stuck in a corner?


Sorry, I don't buy that...First of all, what is the room background noise at 5 Hz? You showed a graph with data down to 20 Hz and it was already rising down there. I would imagine that it would be well above the speaker response by several dB if you could measure it down that low.

Secondly, with the exception of very few multi-thousand dollar mics with multi-thousand dollar instrumentation mic preamps (e.g., B&K) it is impossible to measure accurate acoustical data at 5 Hz even in an anechoic room with an extremely low noise floor. With a 1/2" mic capsule you will be running into the mic noise at frequencies well above that. It's the same thing I tell my students at the university: "just because the computer shows you some data doesn't make it real".

Thirdly, there is no evidence in the published literature that supports the audibility of distortion at 5 Hz. If you are talking about harmonic distortion generated at higher frequencies because of that, then an anechoic THD measurement should show it.

Yes, I would like to see the filter being totally flat (i.e. on the 0 line) down there. But, I think what you are seeing is a gentle bump that comes from running out of resolution below 15 Hz or so. A standard FIR filter with the same number of taps would run out resolution at about 150 Hz. One of the core elements of our patents is the nonlinear stretching of the frequency axis before applying the Fourier Transform so that the MultEQ filter can correct down to nearly 15 Hz without requiring any more computing power.

Finally, let me say that it's great to see someone who is so into the intricate details and has the time and inclination to take measurements. It's this kind of real-world data that really helps identify issues and gives us feedback for improvements. This past summer we had an intern from the university working on the analysis of room data that we have collected through our installer program from more than 600 rooms around the world. It is the largest such database (still growing) that I know of and allows us to do do a statistical analysis on enough samples to make it meaningful. We will be adding your room to it.

Hi Joe,

The decision of when to turn Dynamic Volume on is totally up to you. If you want to control the dynamic range or level the volume so that it doesn't jump around as you change songs then turn it on. Day is the lowest setting that only applies a limited amount of control, Evening is in the middle, and Midnight applies the most. Day is not the same as Off. If you don't want the control during actions movies then you should set it to Off.

That cut has been applied to all three fronts as they are about 1foot from the front wall and being that 65Hz is the 2nd Length mode for this room, their placement and the fact the listening position is in the peak position for that mode has caused it to be excessive.

As mentioned earlier there is not much I can do about it with positioning or treatments due to the rooms multifunction needs(this will change as we are planning a dedicated room so I'm not too concerned). Previously I have used the 80Hz xover to remove most of that energy from the mains and direct it to the sub where I used the subs built in EQ to iron out that region.


I mentioned I could only hear it due to the driver producing distortion. The distortion I was eluding to was Harmonic and slight mechanical distortion which was in the audible range.........obviously the 5Hz fundamental is inaudible(unless your using a TRW-17 and outputting clean 5Hz at 110db+, now that one is on the to do list)

The Mic I'm using has correction values down to 5Hz which are loaded into the software to correct the measurements, here are the values loaded:
5Hz -24.79
6.3Hz -16.22
8Hz -12.54
10Hz -9.23
12.5Hz -6.64
16Hz -4.68
19.95 -3.06
.....etc
These corrections were taken from a number of the same model mic being professionally calibrated and then testing a number of other same model mic's which were apparently pretty close, so the mic to mic variations for that model were relatively close. Not as good as having my mic tested and and having correct corrections, but close enough maybe.

But you are probably right and that raise below 10Hz is most likley the noisefloor as if you extrapolate the previous noisefloor measurement out and take into account the corrections for the mic that would see the noisefloor take over the measurement around the 10Hz region at that db level.

I could probably run a distortion test at 5Hz and see what frequencies are being excited by the THD and mechanical distortion............but I really don't want to damage my speakers, so I might just leave that one alone........all I can say is that the sound produced was some form of distortion as you could also see the woofers pounding in and out trying their hardest to reproduce the sweep. Anyway I digress


Thanks again for the insight Chris, it's great to see people like yourself taking an interest in and offering your knowledge and experience, it really is appreciated

These forms of testing have always intrigued me and I have spent countless hours performing all kinds of test on equipment to find out the crux of what goes on. In some cases it I have stumbled onto some quirky issues. One of them was an issue I found with my old Denon AVR1603 which did something strange to the sub output when in a particular mode(here) And more recently some funny business with movie soundtracks(here) and many others.

cheers

Chris (et al):

I just set up a Denon 1909 and ran the Audyssey cal. Worked well, including the expected tightening up of bass (5.1 system).

However, the effect of Dynamic EQ seems exagerated (too much bass as volume is turned down - I realize Dynamic EQ does more than just the old "loudness" function).

The Audyssey cal set relative speaker gains very close to what I would have done based on a quick manual test using a Radio Shack SPL meter prior to running Audyssey (I did NOT make any manual changes. I just wanted something for reference prior to running Audyssey). However, Audyssey set the average gain to about +8 db. I understand that Audyssey's intent was to set gains as appropriate to achieve the desired absolute sound level for application of Dynamic EQ. BUT, it seems to have set the gain too high based upon my "perceived" result (opinion shared by 3 others in the room).

1) Is there a "typical" explanation for this apparent scenario?

2) Would it work (and would it be advised) to manually reduce all the speaker gains, while preserving their "relative" difference, to adjust the virtual impact point of Dynamic EQ?

3) Other ideas/comments?

Thanks, Tim

Hi Tim,

The gains are not set based on perceived level or opinion. They are set based on a standard that corresponds to the mixing level used for film. That standard is 85 dB SPL for –20 dBFS bandlimited noise. So, if you use the internal 1909 test tones and measure with an SPL meter you should get 75 dB SPL (C-weighted slow). Note that the internal noise is –30 dBFS so your reading will be 75. That sets the level to mix reference when you have the volume control at 0. Dynamic EQ maintains the reference balance of high and low frequencies, but at lower volumes. Without it, the perceived balance of highs and lows that was used to make the mix is significantly changed as you lower the volume.

The value of the trims depends on many factors, including the efficiency of your speakers and the listening distance.

I don't quite understand the reason to reduce the level trims. If you reduce them all by the same number, that's identical to turning down the master volume. Changing the trims does not change the Dynamic EQ reference. It will simply think you are at a lower volume and apply the needed correction.

The "too much bass" phenomenon is usually attributed to two factors: (1) some TV content is not mixed to film reference and in many cases not bass managed during the mix. Very high levels of bass (especially during local station content) creep in because they are not properly monitored. In fact, many of these are produced without a subwoofer in the control room and with monitors that roll off at 60 Hz. Very sad. (2) incorrect levels on the LFE track. This often happens when the Bluray player is decoding the content and sending it as multichannel PCM to the 1909. The required 10 dB boost to the LFE track is applied twice and that causes "too much bass". If you look a few posts up, you will see instructions on where to look to change that in the AVR if it is needed.

Chris, Thanks so much for your very quick and thorough reply.

I understand the gains are set not based upon "perception," rather to achieve an objective, desired absolute SPL per all of Audyssey's work with listeners and professional sound mixers. My (admittedly preliminary) "too much bass perception" was observed with TV and Elton John's 60th BluRay disk. That disk's audio is uncompressed multichannel PCM so perhaps there could have been multiple applications of the 10DB LFE gain. The BluRay player is a Panasonic BD35k. From my prior forum browsing, I thought errors in applying LFE gain were usually (or always) when the player was converting to PCM from DD or other audio format. I'll do more extensive testing and look for the posts you referenced.

NOTE: My intent in manually lowering all gains was thinking it would reduce absolute volume "AFTER" Audyssey Dynamic EQ processing, resulting in absolute SPL lower than Dynamic EQ predicts and thus virtually reducing its impact. I'm not suggesting any issue with Audyssey's research regarding "perceived" frequency response at different absolute SPL levels. I'm just trying to think through how to tailor and thus be able to use Dynamic EQ's much desired effect if I can't identify and correct a root cause (hey, maybe the mic supplied with the Denon 1909 had 8 DB too much gain... ha ha).

Per above, I will use your suggestions to look for a "root cause." But, if I wind up at a loss, would my concept of manually reducing each channel's gain achieve what I'm trying to accomplish?

THANKS AGAIN!!!

--- Tim

That was true for the analog output, but the opposite is true for several models when sending PCM over HDMI.


I understand, but it can't be done because the trim settings talk to Dynamic EQ. If you have a preference for less bass, then why not just turn down the sub level trim and leave the others where they are?

I can't get no bass or highs from audyssey. ran it like 10 times same ressult. oh well. Hello Yamaha and YPAO

Good luck with that.

Hi everyone,

(This was originally posted on the Denon 1909 thread, but I was recommended to repost it here)

Firstly thanks for all the posts on this forum, it greatly helped me choose the components of our first home theatre system! (yay), which are:

Denon 1909,
Paradigm Cinema Series 90 Front and Centre Speakers,
Paradigm Cinema ADP Surrounds,
Rel T2 Sub-Bass,
Samsung Blu-Ray BDP1500,
Sony PS3,
Panasonic Viera TH-46PZ800A

My question is in relation to setting up the Rel T2 correctly:

The Rel is connected to the Denon FL+FR binding posts using the High Level (Neutrik) connector AS WELL AS the .1/LFE phono interconnect from the Denon Sub out. This is the recommended configuration in the Rel users guide. The Rel users guide also has detailed instructions on the setup procedure including positioning the speaker, setting the phase, as well as how to set the correct Lo/Hi levels and the correct crossover frequency.

However..... the Denon manual states that before running Auto AudysseyEQ to set the Sub levels to 12 o'clock and the crossover to the HIGHEST frequency possible (in this case 120Hz).

Which of these two different setup procedures should I follow? If I run the Audyssey with the T2 crossover set to 120Hz, should/can I subsequently turn it back down to the frequency determined by the Rel setup procedure, or will that negate the Audyssey calibration? Please can anyone help?

Many Thanks,
Dave

P.S. The system already sounds fantastic! (I've never had a sub before) I would just like to make sure it is as good as it can be ;-)

Can't you apply your BFD filters and then run Audyssey? That way you can get the flattest response possible.

I believe you want to set the physical crossover on the sub itself to the highest setting. I have a Hsu VTF 3.3 sub and in my small room 19x13x8, I had to turn the sub level(volume) to 9 oclock, because the sub is overkill for my room and Audyssey would set it to -12db which is an invalid trim level from what I understand reading through this thread. I then came up with reasonable and normal sub levels of -7 or -8db after running Audyssey.

Chris, Here is one comparing the subs output between Audyssey on and off with also the previous filter traces and a 1/24th noisefloor reading captured directly after the sub measurements.


Red = Audyssey off Line level (1/3 filtered) reference line for audyssey filters
Green = Audyssey filters line level (1/3 filtered)
Blue = Audyssey off (1/3 filtered)
Purple = Audyssey on (1/3 filtered)
Aqua = Noisefloor (1/24 capture)

I ran a number of measurements to make sure they are repeatable and they tracked identically. You can see the effect of Audyssey on and off and how it is increasing the sub 15Hz region then reducing the 15-60Hz region and then increasing the 60Hz+ region as was shown in the line level measurements.

cheers

I am not so sure the Rel is wired right.

I have a REL Storm which also feature HL input and LFE but it is not possible to use both.

The HL input acts like a parallel wiring, it is mostly an hifi setup letting you use the sub and the speakers wired directly to the amp. In that setting the sub crossover will not interfere with the speakers filters. In that setting you will need to manually set the REL crossover and level.

The LFE input is used for HC listening, with that input and using up to date HC amps featuring Audyssey, the amp will control all from crossover to sub level. In that case it is recommanded to follow Audyssey instructions. Also check the sub level at the end of the calibration, if it goes up to 0 then the sub level needs to be higher, if it does down to -12 the sub level needs to be decreased. Then re-do calibration and check.

I've got a REL Storm III and a Denon 2807 so I'll assume your REL has basically the same options as mine.

RELs have both high and low level inputs. You're hooking it up with a low level input, basically a long interconnect. They also have 3 settings: phase, gain (actually 2 gain controls, one for the low level input and one for the high level input), and crossover. What you need to do is as follows:

First, the quick summary if you just want to follow instructions blindly is to hook up the REL with the low level input and to set the REL's settings as follows:

Phase: This is called a Mode switch on my REL. Set it to 0 degrees, LFE input option (there are only 4 options: an LFE input option and non-LFE input option for each of two different phase angles, 0 and 180 degrees).

Crossover: set to the highest frequency or just ignore it.

Gain: set to low level gain setting to 12 o'clock and adjust if necessary (ie the trim setting shown on the Denon is at the high or low extremes). You can ignore the high level input gain setting.

If you're only using the low level input on the REL, stop at this point unless you want some explanations.

-----------------------------------------------------------

Second: the same info with explanations and an additional option:

Phase (Mode Switch): REL offer only 0 and 180 degree phase options but 2 settings for each. One is the LFE option and the other isn't. The non-LFE option is there for use with amps without sub outlets and crossover control or if you want to use both high and low level inputs from a receiver and swap between having the receiver control things and running your speakers full range with the sub blending in underneath the speaker as REL recommend for music systems. Use the 0 degrees LFE input option if you're using a low level input only. If you're using both low and high level inputs and swapping between them to swap between the standard movie setup for the sub and REL's recommended setup for music listening you'll need to change the phase setting every time you do that swap. See the details below for that.

Crossover: Set the crossover to the highest frequency option and forget it if you're only using the low level input. Doing that won't achieve anything different in results since you switch the crossover out of the circuit with the phase switch but it does have one advantage. If you set it as high as possible and you accidentally bump the phase setting from 0 degrees LFE input to 0 degrees non-LFE input, you probably won't notice a difference but if you have the crossover set very low on the REL and bump the phase setting to the non-LFE option, the REL will roll off a lot of the input from the Denon which you don't want to happen. If you want to also use the REL music setup approach, see the instructions below.

Gain: set the low level gain control to 12 o'clock as Denon suggest since the Denon will be matching levels for you. The only reason to change the setting from 12 o'clock is if this results in a sub trim that is close to or at the maximum trim level for the sub in the Denon settings. If that's the case, adjust the REL's gain to get a trim setting that's closer to 0 dB but don't bother getting anal about getting it to 0 dB since that's not important. If the trim level that the Denon shows is high and positive, the sub's gain level is too low and needs to be boosted, while if the trim level is high and negative the sub's gain level is too high and needs to be reduced. Ignore the high level gain setting unless you're going to be using the REL music setup as well in which case see the instructions below.

That's it with all explanations.

Stop here if all you wanted was an explanation of the quick summary and do not intend using the option of swapping between low and high level inputs on the REL for movie/music playback respectively.

----------------------------------------------

Still reading? Well, now things get complicated.

As I said above, you can set things up so that you can swap between having the Denon control the sub for movies and using REL's recommended setup for music.

This setup requires both the low level connection between the Denon and the REL that is used above, AND a high level connection with the REL also connected to the Denon's L and R speaker terminals. It also requires setting changes at both the REL and the Denon in order to swap between these two options and it requires 2 separate setup processes, one for the low level input for movie playback and Audyssey EQ, and a totally different setup process for the REL for the music setup where the REL controls the handover between speaker and sub.

First make both the low level and high level connections between the Denon and the REL.

Then set things up for the standard REL music option in the following way. I'm putting this setup first because you either need to do this BEFORE running auto setup to set up Audyssey or, if the auto setup has already been run, you need to turn Audyssey off.

Then:

In the Denon's speaker setup, set the L and R speakers to Large and tell the Denon there is NO SUB.

Then, at the REL, set the phase to the 0 degrees non-LFE input option at the REL's phase setting.

Now follow REL's instructions to establish the correct phase, crossover and high level input gain settings so that the sub blends with your speakers.

You now leave the crossover setting at that frequency at all times since the crossover gets switched out of the circuit when you swap to the movie setup. You also leave the high level gain setting where it is at all times since that setting only works for the high level input. You have a separate low level gain input which sets the gain for the movie setup.

You do need to note the phase/mode setting you ended up with. It may not be 0 degrees and it is also the non-LFE input option. Whatever it is, you will be swapping between this setting for music and the 0 degree LFE input option for movies so make sure you know what phase/mode setting you ended up with. This is the only change required at the REL to swap between the 2 setups.


NOW, it's time to set things up for movies and do the Audyssey setup.

As I said, the only setting to change on the REL when swapping between setups is the phase/mode setting so set it to 0 degrees LFE input. At this point you also need to set the low level input gain setting to 12 o'clock.

Then change the speaker settings on the Denon to tell it you have a sub.

Now run Auto setup to set Audyssey up. You may need to change the low level gain setting as described above if setup sets the sub's trim setting at the high or low extremes. Once you've finished the auto setup process change any speakers set to Large by the setup process to Small and turn Audyssey on.

That completes the 2 different setup processes. You swap between the 2 setups as follows:

For movies with full Audyssey EQ of sub with the Denon controlling the handover between sub and speakers:

On the REL set the phase/mode switch to 0 degrees LFE input.

On the Denontell the Denon that there is a sub and set the L and R speakers to small.


For music using the REL setup with the REL controlling the handover between speakers and sub:

On the REL set the phase/mode switch to the non-LFE input option for whatever was the correct phase angle in the REL setup process.

On the Denon tell the Denon there is no sub and set the L and R speakers to Large.

You may also want to turn Audyssey off when swapping to the music setup but you can try it both ways. It's hard to predict which will be best since the sub will not be getting equalised by Audyssey but it will pick up whatever EQ Audyssey will be applying to the L and R speakers at low frequencies. Unfortunately that EQ will be inappropriate because the sub has a different frequency response to the speakers and is supplementing them so strange things may happen to the bass response curve. The results may be pleasing or not pleasing with Audyssey on but they won't be accurate and the setup process you've done for the music setup was done with Audyssey off. Since you can only store one lot of settings for Audyssey on the Denon, the process above ensures that those settings are for the movie soundtrack setup so they won't be appropriate when you swap to the music setup. There's no way around that because you can only store one lot of filters for Audyssey and getting Audyssey to EQ things correctly for both settings would require the Denon to store 2 different lots of settings that you could swap between.

REMEMBER: You need to choose the right combination of settings for the REL and the Denon whenever you swap between the 2 different sub integration approaches. If you get the wrong combination you'll end up with too much or too little bass. It's not quite simple by any means since you have to change one setting on the sub and 2 settings on the Denon every time you swap between these 2 approaches.

Frankly I wouldn't worry about using this option because you don't get separate Audyssey EQ control of the sub with the normal REL approach and there's a lot of fiddling with settings involved. I'd only use the low level input connection between REL and Denon and set the REL to the LFE option so that the Denon controls the crossover, and run with the same setup options for movies and music. That ensures Audsyssey EQ of the sub at all times and avoids a messy process of changing multiple settings. In my view you get better results and sub integration using the REL's LFE input and letting the receiver handle the handover between speakers and sub but the option is there if you want to try it.

Yep, just haven't had time to play around with it more.

And when using the RS SPL meter, who knows how accurate that really is since I don't have a calibrated mic to compare to.

There's only one hump around 60hz in the response that I would correct and it's not real bad. Keeping it simple and keeping the BFD out of the loop may be best.