• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Denon AVR-X3800H is released.

GXAlan

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,866
Likes
5,953
Can I share my YouTube review here? Not sure what the climate for that is around here lately
Sure. People value measurements of some sort and as long as you are not egregiously advertising and monetizing, I think it is ok. I say this with zero authority.
 

Brian6751

Active Member
Joined
Dec 3, 2019
Messages
155
Likes
161
Just my opinion. I will leave the measurements to Amir. My preferences for gear do tend to follow Amirs measurements though so knowing that and $8 should get you a nice coffee.

 

minus3dB

Member
Joined
Oct 27, 2019
Messages
50
Likes
42
In my response I didn’t distinguish between power off and disconnect. In pre amp mode the amp runs cooler and SINAD improves at least for the X3700 and higher models. If that is achieved with power off or disconnect, I don’t know and so far I didn’t bother. It just works. ;-)
I think it is very unlikely any or all power amp modules would be powered down when "deactivated" in the software. If they were then switching surround modes in software (which can happen automatically based on bitstream if so configured) would be very clunky as the individual modules would not power up and stabilize instantaneously. There is likely some BIT that is run including on the PA modules when the entire unit is powered up. So that would consume some time during power up as well.

OTOH, disconnecting them is instantaneous via opening of relays at the inputs to the modules.
 
Last edited:

minus3dB

Member
Joined
Oct 27, 2019
Messages
50
Likes
42
Yeah I know Whathifi is famous (or infamous) for their (non sense) subjective reviews, but "reviewing" an AVR by running calibration then bypassing them is a new level of incompetency.
They usually throw in a "it had good punch and rhythm." smh
 

anphex

Addicted to Fun and Learning
Forum Donor
Joined
May 14, 2021
Messages
662
Likes
870
Location
Berlin, Germany
Maybe, but in my opinion still the "best buy for money" out there with Dirac Live expansion :cool:
I don't see how Dirac should be such a game changer when it's running on the same chipset that procesees XT32. The results should be the same.
 

minus3dB

Member
Joined
Oct 27, 2019
Messages
50
Likes
42
I don't see how Dirac should be such a game changer when it's running on the same chipset that procesees XT32. The results should be the same.
The computationally intensive part of the Dirac processing occurs off board on the PC/Mac during the measurements and calculating the filter coefficients. Not that the filter processing onboard doesn’t place any burden on the DSP, but I would guess it is less of a load than with audessey. That’s just speculation though.
 

HarmonicTHD

Major Contributor
Forum Donor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,829
The computationally intensive part of the Dirac processing occurs off board on the PC/Mac during the measurements and calculating the filter coefficients. Not that the filter processing onboard doesn’t place any burden on the DSP, but I would guess it is less of a load than with audessey. That’s just speculation though.
Not necessarily. Audyssey conducts all its computations in the AVR if you don’t use the iOS App or the MultiXQ software. Of course we don’t know yet how they will implement Dirac.
 

minus3dB

Member
Joined
Oct 27, 2019
Messages
50
Likes
42
Not necessarily. Audyssey conducts all its computations in the AVR if you don’t use the iOS App or the MultiXQ software. Of course we don’t know yet how they will implement Dirac.
I may not understand your point as you appear to be acknowledging mine. Are you just stating that Audyssey can also perform off-board calculations if using those apps? I'm aware of that, but have always assumed the vast majority of Denon owners don't use the app or MultiXQ SW.

Regardless, I suspect most agree that the initial point that Dirac shouldn't be any better than Audyssey since they both use the same processor has been adequately refuted.
 

SKBubba

Senior Member
Joined
Sep 5, 2018
Messages
453
Likes
785
Just my opinion. I will leave the measurements to Amir. My preferences for gear do tend to follow Amirs measurements though so knowing that and $8 should get you a nice coffee.


Good, credible review of features, setup and listening impressions v a way more expensive model in the same decent speaker setup.
 

anphex

Addicted to Fun and Learning
Forum Donor
Joined
May 14, 2021
Messages
662
Likes
870
Location
Berlin, Germany
@minus3dB @Alexium @peng

What does the initial measurement and processing have to do with a noisy DSP?
Even if Dirac provides 10 times the accuracy, when the processing happens on the same chip the issues with the electronics are the same.

Your analogy doesn't make sense since you're comparing digital processing with electric signal flow.

XT32 for its price (MultEQ 20 €) is already pretty accurate and does a great job. Dirac (350 $ if google was right) looks just like a marketing gimmick to me.
 
Last edited:

SupremeAV

Member
Joined
Sep 8, 2022
Messages
11
Likes
35
@minus3dB @Alexium @peng

What does the initial measurement and processing have to do with a noisy DSP?
Even if Dirac provides 10 times the accuracy, when the processing happens on the same chip the issues with the electronics are the same.

Your analogy doesn't make sense since you're comparing digital processing with electric signal flow.

XT32 for its price (MultEQ 20 €) is already pretty accurate and does a great job. Dirac (350 $ if google was right) looks just like a marketing gimmick to me.
Thats not how it works. I can see your point about the hardware signal limitations but the DSP processing and how its calculated and implemented between Dirac and Audyssey aren't anywhere near the same.
 

anphex

Addicted to Fun and Learning
Forum Donor
Joined
May 14, 2021
Messages
662
Likes
870
Location
Berlin, Germany
Thats not how it works. I can see your point about the hardware signal limitations but the DSP processing and how its calculated and implemented between Dirac and Audyssey aren't anywhere near the same.
Sooo brand new chip? I still don't get the point. What does Dirac do so much better than XT32? With an exact calibration the resulting FR should look the same, unless XT32 is far worse than I thought. And if Dirac isn't much better than XT32 it just the same limited by the hardware noise/distorion in terms of signal quality.
 

GalZohar

Senior Member
Joined
Aug 8, 2021
Messages
442
Likes
216
DSP doesn't work like that, as in a higher quality chip won't provide different results than a worse quality chip, provided they run the same algorithm (same filters). Of course, the lower quality chip would have to be fast enough to run said algorithm/filters, otherwise nothing will work at all (not related to quality), or an algorithmic change would be required to make it still work. It's not like a better/worse DAC or amp that can have better or worse signal/noise ratio. Plus we don't even know if Dirac runs on the same DSP chip as Audyssey or not (would be economic if they did, but it's not necessarily possible/simple), plus we don't know if Dirac even requires more filter processing power than Audyssey (in fact, Dirac is supposedly more "economic" with how it uses filters).

Basically, room correction (and any DPS) is done on the digital domain, and its quality will be as good as the DSP algorithms used, and no extra noise can be added except the one added by flaws in the algorithm (or simplifications that needed to be done in order to make it run on slower hardware, which probably won't be the case here).


Whatever Dirac actually does/doesn't do much better than XT32 remains theoretical at this point. There have yet to be an actual test that proves any advantages of Dirac. Dirac with DLBC is a different beast, though, as it has capabilities that you can't get elsewhere (only MSO comes close, but that can usually be applied to subs only as you normally don't have the DSP required to apply it to your speakers, while DLBC applies to mains as well).
 
Last edited:

minus3dB

Member
Joined
Oct 27, 2019
Messages
50
Likes
42
@minus3dB @Alexium @peng

What does the initial measurement and processing have to do with a noisy DSP?
Even if Dirac provides 10 times the accuracy, when the processing happens on the same chip the issues with the electronics are the same.

Your analogy doesn't make sense since you're comparing digital processing with electric signal flow.

XT32 for its price (MultEQ 20 €) is already pretty accurate and does a great job. Dirac (350 $ if google was right) looks just like a marketing gimmick to me.

A general purpose DSP chip does not contribute significantly to the noise level in an audio system depending on how much of the signal chain audio processing is incorporated in its code. It is performing mathematical operations - addition and multiplication - on the bitstream of each audio channel, at a very high rate of speed using stored coefficients that, together with the signal processing algorithm implemented, cause each channel (in the digital domain) to be transformed in magnitude and phase.

There are multiple ways of accomplishing this transformation in terms of the algorithm chosen by the RC system and filter design. As an example, at one time, and it still may be the case, Dirac Live was the only RC to attempt to correct the system impulse response. How the filters for each channel are implemented can affect the end result. Some use IIR filters, some use FIR, some use a combination of both. Some use higher order filters, some lower, some use cascaded filter stages to achieve a higher order. Some use minimum phase filters, some use mixed phase, etc. These are all system design choices that affect the outcome of the RC process. No two RC systems are identical in how they accomplish this algorithmically.

Where are you getting the idea that DSP used in RC systems is "noisy?"
 

Masza

Member
Joined
Oct 15, 2020
Messages
63
Likes
37
Dirac and Audyssey also differ in calculating the filters. Dirac uses PC for measurements and calculation. Only filters are applied in the playback device. Audyssey uses the chipset(s) in the playback device for measurements and calculation of the filters in addition to that.
 

CyrusTheGreat_600BC

Active Member
Joined
May 29, 2021
Messages
182
Likes
150
Dirac and Audyssey also differ in calculating the filters. Dirac uses PC for measurements and calculation. Only filters are applied in the playback device. Audyssey uses the chipset(s) in the playback device for measurements and calculation of the filters in addition to that.
Any algorithm can be implemented on a microprocessor as long as the processor speed and its accessible memory are enough for it. It doesn’t matter if Dirac runs on a PC, because you can interpret any code to run on any appropriate programmable processor which is when firmware updates come into the picture to reprogram the microcontroller. Actually, it might even run faster on the device than your PC, since your PC is doing a lot of background tasks to service GUI, file system interrupts, complicated multi core memory management and allocations, and … to name a few.
 

NewbieAudiophileExpert

Senior Member
Joined
Apr 17, 2022
Messages
306
Likes
135
The difference between Dirac Live and Audyssey isn't in the processor, but in the programming and algorithm that the company uses to help achieve the results.

That is to say that some people believe that Dirac Live is the superior software in this instance, hence why they wish to use it.

Has nothing to do with the microprocessor, but rather the algorithm that runs on the software.
 
Top Bottom