From Marantz AV8801 to AVR-X4400H/X3400H, is it a downgrade, or upgrade?

[PENG]

A little bit dramatic there, don't you think ?

You take the wrong drugs and you could DIE.

Like I always say, this audio hobby isn't exactly brain surgery. It's not even medicine.

Sorry I forgot we have a medicine pro here

 

[KEW]

Room modes are the concern and equalizing below the transition frequency can be beneficial. I recommend using REW and a Mic like the miniDSP or a package like OmniMic to measure the room response.

Above the transition frequency, mic positioning produces huge changes and as such "corrections". When there are nulls (cancellations) from room reflections, EQ cannot correct them and any attempt to do so can stress your amplifier with no positive return.

Moving your sub(s) and multiple subs can correct the nulls.

- Rich

Thanks Rich!
I understand that null cannot be EQ'ed, but what about nodes?
So you use sub location exclusively to control the response with something like Omnimic to evaluate it with success?
You mentioned miniSP, but only in the context of measurement microphone - just wanting to be sure I understand that the miniDSP is not in play as a DSP.

 

[AcuDefTechGuy]

Sorry I forgot we have a medicine pro here

Just giving you a hard time.

I had to work this weekend.

 

[AcuDefTechGuy]

- The X4400H has a much better vented top, so two small Infinity fans should solve all potentially heat related issues, including longevity.
- Less crowded internal, the AV8801 is jammed packed inside.

I bet the Denon X8500 w/ 13.2 Ch is jammed packed inside too.

It's funny how a $20K Theta Digital Casablanca pre-pro is relatively light and simple inside, yet these Denon/Marantz Pre-pros are jammed packed inside.

Didn't they teach these Denon/Marantz engineers that horseshoed electronics inside a little box is bad engineering philosophy?

 

[Out-Of-Phase]

I bet the Denon X8500 w/ 13.2 Ch is jammed packed inside too.

It's funny how a $20K Theta Digital Casablanca pre-pro is relatively light and simple inside, yet these Denon/Marantz Pre-pros are jammed packed inside.

Didn't they teach these Denon/Marantz engineers that horseshoed electronics inside a little box is bad engineering philosophy?

AVRs = Economies of scale.

 

[AcuDefTechGuy]

AVRs = Economies of scale.

Throw everything into the box, sell a lot, sell for cheap, and sell often?

 

[PENG]

Just giving you a hard time.

I had to work this weekend.

Before you give them the drug, remind them to turn Audyssey off okay..

 

[AcuDefTechGuy]

Before you give them the drug, remind them to turn Audyssey off okay..

Definitely. Audyssey might boost the effects of the drugs!

 

[PENG]

It is no wonder that many just keep running REQ until the get a result that they like. This is perfectly fine, however, it is not apex of science that many believe they have achieved.

- Rich

I never thought of that, I just followed instructions to the letter, i.e., mic at ear height, or 1-2 inches above the high back of the couch, do all 8 positions, each position within 24 inches from the 1st position, do it when the room is quiet, set the HVAC fan to off, if a car drove by during a test, repeat that test etc., and I get good results, better than I could using REW and minidsp.

REW/minidsp did do a decent job for me after hours and hours of playing around, including trying different XOs, and sub crawling. I think Audyssey is more effective because like Dirac, it does its thing in both frequency and time domains. It gets better now with the app, I can within 5 dB 15-150 Hz, using 1/12 to 1/24 smoothing, and probably 6-8 dB with no smoothing at all. I have never tried averaging. I know you don't believe in EQ'ing above the transition point, the App allows you to pick exactly that point. To me, it is better to EQ it up to around 6-8 K, in order to sharpen up imaging. My theory is that the left and right channel's FR are quite different in my room, Audyssey EQ each channel individually so that helps. REW can do that too, but it seems to me most don't bother because doing it individually takes some thinking and planning, and there limited equalizers/filters you can use.

Below is from just one attempt, without using the Editor app yet, and no smoothing at all. I could never get it that good using REW and mindsp. After editing, I should be able to improve it further by 1.5-3 dB, I know so because I tried it already in my two channel system. The red one is with Audyssey turn off, so no TLSGuy thank but no thanks. Note that the Audyssey off one already looks better due to the axis scale of 10 dB per division vs the 5 dB/division for the Audyssey ref curve.

 

[Out-Of-Phase]

Throw everything into the box, sell a lot, sell for cheap, and sell often?

Yes, you can ‘sell for cheap’ because of the other three steps you mentioned.

 

[TLS Guy]

So, if you recommend turning off Audyssey, what methods, procedures and/or products do you recommend for EQ?

As I said before, I do not use any room Eq, and personally I don't think most rooms need it.
Despite the room measurement showing some rise in LF at the bottom end. The bass of this system is delightfully light. It touches the bass lightly just like live music in the hall. I play a lot of Baroque music and the bass line articulates beautifully. The best way I can describe it is that the speakers sound very much like high end electrostatics, But with more power. In fact, I have had visitors think the speakers contained electrostatic panels.

The first comment by visitors is that the bottom end is light weight. Well that is how live music is. The bass does not dominate, even most pipe organs, and certainly not in the best of them.

Then I can play a movie with a lot of deep bass and you would swear that there were bass shakers in your seats and the floor vibrates. So it is not in any way bass shy, but the quality of the bass is very different from any other speakers and sounds it.

Unfortunately there are no really well designed reverse tapered aperiodically damped TLs on the market and certainly no dual TLs. As far as I know only one dual TL was marketed and that was by TDL. So at this time this is only for the DIY builder and designer.

 

[panteragstk]

I never thought of that, I just followed instructions to the letter, i.e., mic at ear height, or 1-2 inches above the high back of the couch, do all 8 positions, each position within 24 inches from the 1st position, do it when the room is quiet, set the HVAC fan to off, if a car drove by during a test, repeat that test etc., and I get good results, better than I could using REW and minidsp.

REW/minidsp did do a decent job for me after hours and hours of playing around, including trying different XOs, and sub crawling. I think Audyssey is more effective because like Dirac, it does its thing in both frequency and time domains. It gets better now with the app, I can within 5 dB 15-150 Hz, using 1/12 to 1/24 smoothing, and probably 6-8 dB with no smoothing at all. I have never tried averaging. I know you don't believe in EQ'ing above the transition point, the App allows you to pick exactly that point. To me, it is better to EQ it up to around 6-8 K, in order to sharpen up imaging. My theory is that the left and right channel's FR are quite different in my room, Audyssey EQ each channel individually so that helps. REW can do that too, but it seems to me most don't bother because doing it individually takes some thinking and planning, and there limited equalizers/filters you can use.

Below is from just one attempt, without using the Editor app yet, and no smoothing at all. I could never get it that good using REW and mindsp. After editing, I should be able to improve it further by 1.5-3 dB, I know so because I tried it already in my two channel system. The red one is with Audyssey turn off, so no TLSGuy thank but no thanks. Note that the Audyssey off one already looks better due to the axis scale of 10 dB per division vs the 5 dB/division for the Audyssey ref curve.

View attachment 24896

View attachment 24897

One thing I did when I used to have tons of time to mess with Audyssey was move the mic all around the room. I had 5 seats so they all got a mic position. Then I placed the mic on the arms of the couch in the front and one last position in the middle of the couch for the 8 positions. It worked very well. Nice open sound that was much better than me just manual tweaking.

I've yet to try the above method in my current theater since it sounds good already, but I am going to try the recommendation to lift the mic up another few inches to clear the back of the couch. Once I do that I'll grab a mic and measure with REW. I'm interested to see the results.

 

[RichB]

Thanks Rich!
I understand that null cannot be EQ'ed, but what about nodes?
So you use sub location exclusively to control the response with something like Omnimic to evaluate it with success?
You mentioned miniSP, but only in the context of measurement microphone - just wanting to be sure I understand that the miniDSP is not in play as a DSP.

Here is the miniDSP mic that can be used with REW (free) measurement software.

https://www.minidsp.com/products/acoustic-measurement/umik-1

OmniMicV2 which comes with its own software at a higher cost:

http://www.daytonaudio.com/index.php/omnimic-v2-precision-measurement-system.html

- Rich

 

[PENG]

As I stated in the XMC-1 review, I found that Dirac made a large change to the soundstage when the curtain set very low to operate only below 30 Hz. I preferred using REW to measure and PEQ to level out the bass response. PEQ did not have the same change to the soundstage. Measurements did not account for the soundstage observation. Revel representatives are have made many forum posts supporting the notion that well designed speakers performance can be degraded by REQ above the transition frequency.

When I originally tried XT32 it has the effect of drastically brightening the sound an boosting the rears at least 3B too hot. I would expect a number of observations of increased detail and improved surround.
The measurements posted above have no rounding. All measurements above are taken using the Dirac positions with the mic positions remain within a 4x3 foot area representing a couch. These mics are sensitive to position. With continuous measurements, I have found 15dB or more changes when changing the angle of the stand-mounted mic. It is no wonder that many just keep running REQ until the get a result that they like. This is perfectly fine, however, it is not apex of science that many believe they have achieved.

- Rich

Hello RichB, your experience using REW/PEQ and XT32 seems very different than mine. So far I have used XT32/X3400H, XT32/X4400H, REW/PEQ/minidsp in my LS50/R900 2 channel system and the 7.2 HT system, and found that the best I could manage using REW/PEQ/minidsp was to equal what was achievable using the XT32/AV8801 but the XT32/X3400H and X4400H are able to do a slightly better job, without using the Editor App to change the target curve. I just plotted a few more graphs for the subwoofers only scenario.

The phase graphs actually look better with XT32 on, than with it off, same for the impulse. I don't think I will take the minidsp 2X4 HD out of it's box again.

The reason I decided to plot more graphs last night is that going by memory I thought the bass sounded better in both Amazon/Netflix movies and live concert BR so I wanted to see what's going on. The curves definitely look flatter than those I plotted when the AV8801 was in place, it looks like a case of quality over quantity. I plotted those for the FL and FR too. In all cases XT32 improves the low frequencies a lot. I am thinking of shooting a video next time I run XT32 again.

@ADTG, I think you are right, some people do not prefer flat FR, not in the bass area anyway.

Gents, the graphs below are without any smoothing, anyone can tell which one is XT32 Off, right??

 

[AcuDefTechGuy]

Hello RichB, your experience using REW/PEQ and XT32 seems very different than mine. So far I have used XT32/X3400H, XT32/X4400H, REW/PEQ/minidsp in my LS50/R900 2 channel system and the 7.2 HT system, and found that the best I could manage using REW/PEQ/minidsp was to equal what was achievable using the XT32/AV8801 but the XT32/X3400H and X4400H are able to do a slightly better job, without using the Editor App to change the target curve. I just plotted a few more graphs for the subwoofers only scenario.

The phase graphs actually look better with XT32 on, than with it off, same for the impulse. I don't think I will take the minidsp 2X4 HD out of it's box again.

The reason I decided to plot more graphs last night is that going by memory I thought the bass sounded better in both Amazon/Netflix movies and live concert BR so I wanted to see what's going on. The curves definitely look flatter than those I plotted when the AV8801 was in place, it looks like a case of quality over quantity. I plotted those for the FL and FR too. In all cases XT32 improves the low frequencies a lot. I am thinking of shooting a video next time I run XT32 again.

@ADTG, I think you are right, some people do not prefer flat FR, not in the bass area anyway.

Gents, the graphs below are without any smoothing, anyone can tell which one is XT32 Off, right??


View attachment 24925

The red one has to be Audyssey on. It’s way too flat.

 

[PENG]

For those who wonder about the axis scale, yes I did try to make it 5 dB per division instead of 10, but then it wouldn't fit the screen, and I still have not figure out how to make the fonts larger. I could make it larger on the screen, but as soon as I use the photo capture function, the saved JPG file would default back to the small font size regardless. Does any one know how, before I PM the creator John M or post the question on the Hometheatershack forum about it?

 

[RichB]

Hello RichB, your experience using REW/PEQ and XT32 seems very different than mine. So far I have used XT32/X3400H, XT32/X4400H, REW/PEQ/minidsp in my LS50/R900 2 channel system and the 7.2 HT system, and found that the best I could manage using REW/PEQ/minidsp was to equal what was achievable using the XT32/AV8801 but the XT32/X3400H and X4400H are able to do a slightly better job, without using the Editor App to change the target curve. I just plotted a few more graphs for the subwoofers only scenario.

The phase graphs actually look better with XT32 on, than with it off, same for the impulse. I don't think I will take the minidsp 2X4 HD out of it's box again.

The reason I decided to plot more graphs last night is that going by memory I thought the bass sounded better in both Amazon/Netflix movies and live concert BR so I wanted to see what's going on. The curves definitely look flatter than those I plotted when the AV8801 was in place, it looks like a case of quality over quantity. I plotted those for the FL and FR too. In all cases XT32 improves the low frequencies a lot. I am thinking of shooting a video next time I run XT32 again.

@ADTG, I think you are right, some people do not prefer flat FR, not in the bass area anyway.

Gents, the graphs below are without any smoothing, anyone can tell which one is XT32 Off, right??


View attachment 24925

I am a fan of target curves so it would be interesting to see that applied.
Here is what I was able to achieve in a ½ hour to flatten the bass to within 5 dB between 22 Hz and 75 Hz (of the intended curve).
The sound stage was not significantly altered by these PEQ settings.

It seems odd that speakers with reasonable phase performance properly placed would out of phase.
I suppose room interference can cause cancellations and late reflections, DSP processing can definitely alter phase.
I remember reading somewhere that the MiniDSP 2x4HD introduces about 4 milliseconds of delay. Whereas, their Dirac products can introduce up to 40 milliseconds of delay.

This is why I always compare PEQ/REQ to Pure Direct with trims adjusted (as best I can) as a reference.

- Rich

 

[PENG]

I am a fan of target curves so it would be interesting to see that applied.
Here is what I was able to achieve in a ½ hour to flatten the bass to within 5 dB between 22 Hz and 75 Hz (of the intended curve).
The sound stage was not significantly altered by these PEQ settings.

View attachment 24926

It seems odd that speakers with reasonable phase performance properly placed would out of phase.
I suppose room interference can cause cancellations and late reflections, DSP processing can definitely alter phase.
I remember reading somewhere that the MiniDSP 2x4HD introduces about 4 milliseconds of delay. Whereas, their Dirac products can introduce up to 40 milliseconds of delay.

This is why I always compare PEQ/REQ to Pure Direct with trims adjusted (as best I can) as a reference.

- Rich

I was able to do better than that with REW, but XT32 got me with 5 dB from 18 to 110, before I even touch the editor. Different room, different speakers so hard to compare for sure.

So did Dirac do worse for you, compared to REW/minidsp? I am going to try its Stereo version soon.

 

[RichB]

I was able to do better than that with REW, but XT32 got me with 5 dB from 18 to 110, before I even touch the editor. Different room, different speakers so hard to compare for sure.

So did Dirac do worse for you, compared to REW/minidsp? I am going to try its Stereo version soon.

Yes, but it depends on the mic and on the smoothing.
Here are is a chart for the 20 to 80Hz showing all 9 Dirac positions which are cube 2x3x4 representing my couch.
This is the UMM-6 calibrated MIC measurements.
On the left you can see the thumbnails.

I can smooth the charts and seem clear patterns but different MICs show different dips.

Here is the same data from 20 to 20kHz:

Even when smoothed, there were significant difference above the transition frequency with different (calibrated MICs).

After many, many hours of measurements, I decided to follow the Harman recommendation to limit correction below the 100Hz which had some consistency between MIC and measurement sessions.

- Rich

 

[PENG]

Yes, but it depends on the mic and on the smoothing.
Here are is a chart for the 20 to 80Hz showing all 9 Dirac positions which are cube 2x3x4 representing my couch.
This is the UMM-6 calibrated MIC measurements.
On the left you can see the thumbnails.

I can smooth the charts and seem clear patterns but different MICs show different dips.
View attachment 24927

Here is the same data from 20 to 2kHz:

View attachment 24928 Even when smoothed, there were significant difference above the transition frequency with different (calibrated MICs).

After many, many hours of measurements, I decided to follow the Harman recommendation to limit correction below the 100Hz which had some consistency between MIC and measurement sessions.

- Rich

I typically use no smoothing for frequencies below transition and 1/12 for above and the graphs I posted are with no smoothing at all. It is strange that we have opposite results, i.e. manual EQ (REW/PEQ/minidsp) works better for you, and auto REQ (only Audyssey so far) works better for me.

If I understood right Dirac Live has a free 30 day trial, that should give me enough time to find out if it works well in my rooms too. If the PC/Mac version is in fact available for the free trail then I am going to do for sure, and soon.