miniDSP Tide16 - Holy Grail with 16 Channel Atmos/DTS:X, high SINAD

[Ultrasonic]

The stereo version of Dirac Live on the SHD (and other miniDSPs) is identical in terms of functionality as on other platforms. I'd therefore very much expect the same to be true for ART.

The extra functionality offered on the stereo miniDSPs with Dirac Live is exclusively via miniDSP's own software and completely independent of Dirac Live. I've not used an HT/HTx but I think the same it true?

 

[Mort]

I've not used an HT/HTx but I think the same it true?

You are correct

 

[3ll3d00d]

Coming back to that non-ideal step response with ART - is it audible?

Any pics/examples of this? A quick search didn't yield anything obvious

 

[Sokel]

It would be more accurate to say that dips should not be filled by boosting the amplitude of the equalisation. Whether a dip can be filled with DSP depends on the cause of the dip. If it is caused by a sound interacting with its reflection, then this hole is infinitely deep and requires infinite SPL to fill it. But if it is caused by two speakers interacting with each other, then the phase of the narrow band can be adjusted in one speaker to remove or reduce the dip in the sum. This is why the eBook contains a procedure to help you diagnose the cause of a dip.

Detecting the true nature of a dip requires skills and gear sometimes way above of what we amateurs use.
High resolution data, and a combination of static and MMM measurement and multi-static can be very good hints though.

Gear with accurate timing help also (a UMIK-2 instead of UMIK-1 for example, or traditional mic-interface with loopback, calibrated etc)
Ok, they won't triangulate space as the Trinnov stuff does or pro arrays do but way better than the all time classic traditional forum poster with its static 1/6 REW charts, with questionable positioning, etc.

So, before we decide to fill dips or not it would be nice to have reliable data about it, and this is harder than we use to think.
I'm firmly at the no-fill-dips camp, as I never met a real expert who fills.
They may well demolish the room to fix positioning, etc, but they don't fill.

 

[thcdru2k]

I think another issue is what you are listening to that would reveal whether or not a non-ideal step is audible. Dirac ART was made for people who watch movies but a car explosion may not reveal it as heavily as techno/tech house/modern EDM where kick fundamentals sit around 30-55hz or trap hiphop with short kick envelopes. I think maybe the population that is into calibration may not even be listening to the type of music that would reveal these issues. Sharing my personal opinion... I think there is a non-audible difference but its nuanced because it's probably not noticeable in other music genres or movies.

Edit: When I think about it these are not audible frequencies so when I say audible it's more of a feeling when I listen to that type of music. I think that is another reason why this is difficult. I feel that it's better and I'm crossing that with hearing when its really not hearing.

 

[EWL5]

Gear with accurate timing help also (a UMIK-2 instead of UMIK-1 for example, or traditional mic-interface with loopback, calibrated etc)

We've needed to help a handful of Dirac users by telling them to de-sensitize their UMIK2s to be more like UMIK1s to get a proper calibration. With tools like Dirac and REW that were tested w/UMIK1s, it sometimes doesn't make a lot of sense to pay more for a UMIK2 if not needed.

 

[Oddball]

I think another issue is what you are listening to that would reveal whether or not a non-ideal step is audible. Dirac ART was made for people who watch movies but a car explosion may not reveal it as heavily as techno/tech house/modern EDM where kick fundamentals sit around 30-55hz or trap hiphop with short kick envelopes. I think maybe the population that is into calibration may not even be listening to the type of music that would reveal these issues. Sharing my personal opinion... I think there is a non-audible difference but its nuanced because it's probably not noticeable in other music genres or movies.

Edit: When I think about it these are not audible frequencies so when I say audible it's more of a feeling when I listen to that type of music. I think that is another reason why this is difficult. I feel that it's better and I'm crossing that with hearing when its really not hearing.

Movies, at least some of them, use drums, and have music. ART was made to reduce decay in any material, but it is more difficult to detect subtle nuances when you have 13 or 15 channels plus subs firing at the same time (or close to that) then when listening to stereo material (even with subs). Plus in standard listening environment you have shared attention between video and audio for movies/shows and that will further reduce your attention level for audio compared to music only.

 

[Dj7675]

Dirac ART was made for people who watch movies but a car explosion may not reveal it as heavily as techno/tech house/modern EDM where kick fundamentals sit around 30-55hz or trap hiphop with short kick envelopes. I think maybe the population that is into calibration may not even be listening to the type of music that would reveal these issues. Sharing my personal opinion... I think there is a non-audible difference but its nuanced because it's probably not noticeable in other music genres or movies.

What experiences have you had other information that would lead you to believe that?

 

[Sokel]

We've needed to help a handful of Dirac users by telling them to de-sensitize their UMIK2s to be more like UMIK1s to get a proper calibration. With tools like Dirac and REW that were tested w/UMIK1s, it sometimes doesn't make a lot of sense to pay more for a UMIK2 if not needed.

I don't know about DIRAC, but to get amateur semi-reliable data at advanced measurements like FSAF, UMIK-2 is the bare minimum.
Cause what good any correction may be if it results in distortion and noise over the roof down low?

Trade-offs is the norm, yes, but it's getter to keep it sane across the line.

 

[thcdru2k]

What experiences have you had other information that would lead you to believe that?

I think I have to be careful because this is subject to huge listener bias. The spectrogram here is from my system https://www.audiosciencereview.com/...nel-atmos-dts-x-high-sinad.68883/post-2500282 , it's just calibration by acoustic timing reference and maximum summation at crossover frequencies for subwoofer integration with only mild cuts no boosting at certain frequencies. UMIK-2 was used for acoustic timing reference as well. I do have two 15in subs for this system. You can see there are no blobs of energy at low frequencies that are typical of uncalibrated/auto room corrected system. I do listen to a lot of techno and occasionally hip hop. The system is enjoyable. My wife figured how much I spent on Flex HTx and surprisingly she did not get upset. She said the system is sounds good. However; when I've had friends who have heard the system before Flex HTx come over to listen they said they cannot hear anything different. Poor sampling and very very subjective to bias but I will say my wife has not been a fan of the continual upgrades to the system over the last year but she really did not object to the Flex. That's why I preface my statement above as personal opinion. I cannot figure out any other way to prove this other than here I took the time to produce a clean spectrogram.

 

[EWL5]

I don't know about DIRAC, but to get amateur semi-reliable data at advanced measurements like FSAF, UMIK-2 is the bare minimum.
Cause what good any correction may be if it results in distortion and noise over the roof down low?

Trade-offs is the norm, yes, but it's getter to keep it sane across the line.

My Dirac knowledge is fairly reliable (1st hand for DL, 2nd hand for DLBC and ART). I think quite a bit of you DIY DSP folks will be pleasantly surprised by ART should you ever give it a try!

I don't know if it helps any but something like Audyssey calibrates to a reference dB whereas Dirac does not. It asks for a margin above the noise floor during the channel leveling step.

 

[Sokel]

My Dirac knowledge is fairly reliable (1st hand for DL, 2nd hand for DLBC and ART). I think quite a bit of you DIY DSP folks will be pleasantly surprised by ART should you ever give it a try!

I don't know if it helps any but something like Audyssey calibrates to a reference dB whereas Dirac does not. It asks for a margin above the noise floor during the channel leveling step.

I admit I do diy electronics, lots of them.
Room, I don't, I left that to pros when I built my house and dedicated.

And I mean room-room, nor gear, that's another matter, it's not fixed to address my 30Hz mountain for example, that's up to DSP for now (on occasion) .
It's also strictly stereo with big-ish active speakers, not subs.

So I feel we speak an entirely different language (apart from my obviously terrible english) if you talk HT for example.
For example, I detest this slow subs rumble some of the HT people are proud of, I value the lighting speed chest punch that decays in an instant in contrast, which lives way above the subs.
Never heard gear with the subs far apart doing it, while fairly big (almost) full range speakers do it just fine, even not so good ones, even decades old ones.
Proportions are over the chart sometimes also, do subs and mains wrong and baritone's mouth image like caves.

So...

 

[EWL5]

I admit I do diy electronics, lots of them.
Room, I don't, I left that to pros when I built my house and dedicated.

And I mean room-room, nor gear, that's another matter, it's not fixed to address my 30Hz mountain for example, that's up to DSP for now (on occasion) .
It's also strictly stereo with big-ish active speakers, not subs.

So I feel we speak an entirely different language (apart from my obviously terrible english) if you talk HT for example.
For example, I detest this slow subs rubble some of the HT people are proud of, I value the lighting speed chest punch that decays in an instant in contrast, which lives way above the subs.
Never heard gear with the subs far apart doing it, while fairly big (almost) full range speakers do it just fine, even not so good ones, even decades old ones.
Proportions are over the chart sometimes also, do subs and mains wrong and baritone's mouth image like caves.

So...

You are the perfect customer for ART. The most common feedback is their bass is tighter and leaner w/ART vs before. The correction in decay is so great that many say the bass response is similar to what you get when wearing headphones (I can't think of a higher compliment than that)!

I have been really jealous of all the folks who are already on ART. I have a capable AVR that can do ART but I can't justify it with a 5.1.2 system consisting of a Martin Logan L/C/R passive sound bar and 4 ceiling speakers. I have an ultra-clean setup but no full-range speakers and only one sub (in case you're wondering, it is WAF-related)!

 

[Sokel]

You are the perfect customer for ART. The most common feedback is their bass is tighter and leaner w/ART vs before. The correction in decay is so great that many say the bass response is similar to what you get when wearing headphones (I can't think of a higher compliment than that)!

I have been really jealous of all the folks who are already on ART. I have a capable AVR that can do ART but I can't justify it with a 5.1.2 system consisting of a Martin Logan L/C/R passive sound bar and 4 ceiling speakers. I have an ultra-clean setup but no full-range speakers and only one sub (in case you're wondering, it is WAF-related)!

At a funny note, if they were targeted for customers like me they would go bankrupted

(unless they published their algos so we know exactly what they do, so they would also go bankrupted but in a helpful way )

 

[thcdru2k]

Using separate subs generally means more crossover filtering in the bass, so there’s more phase rotation/GD to manage than with a big full-range active. I’ve done the best I can with alignment; subs are just the practical solution I can afford right now...

 

[EWL5]

At a funny note, if they were targeted for customers like me they would go bankrupted

(unless they published their algos so we know exactly what they do, so they would also go bankrupted but in a helpful way )

I will give one caveat. ART hasn't been able to figure out corrections for infrasonic bass (i.e. below 20Hz). This has forced some to prefer DLBC bass response over ART. I have a feeling you would prefer the ART treatment (accuracy over SPL).

 

[Sokel]

I will give one caveat. ART hasn't been able to figure out corrections for infrasonic bass (i.e. below 20Hz). This has forced some to prefer DLBC bass response over ART. I have a feeling you would prefer the ART treatment (accuracy over SPL).

Classical listener, so SPL is unavoidable, crest factor takes no prisoners at some works, even if you start at modest 70's dB SPL (C) average RTA reports 90dB SPL (C) for max and 110dB SPL (Z) for peaks.

Can I have both please?

 

[Oddball]

Classical listener, so SPL is unavoidable, crest factor takes no prisoners at some works, even if you start at modest 70's dB SPL (C) average RTA reports 90dB SPL (C) for max and 110dB SPL (Z) for peaks.

Can I have both please?

Just to clarify, ART can give you both SPL and relative accuracy - first hand experience.

Frequency response up to 150hz can be as tight as 1dB band with 1/6 smoothing. No traditional crossovers, although at the point where non-subs kick in for support there is bit of divergence between the channels. Decay is reduced drastically without need to go for complex DBA setup, and works up to 150hz - as opposed e.g. to 80hz when you would generally cross over to your mains.

SPL will depend on how much headroom Dirac uses. When I use +12dB shelf on low end, my max master volume reach is +5dB to reference. That still gives you, at least in HT language, +120dB peaks for the subs (+12dB from Dirac - is your subs can take that much pain) and +100dB for the mains (plus whatever is your bass boost in Dirac - I use 7dB). It's really a lots of noise and doubt anyone would want more. You can reduce the bass boost and free up some headroom to distribute more evenly between the subs and mains if that is more in line with preference.

For subs most can add additional gain at the sub gain level as ART will in general set it pretty low (at least for multiple decent subs). However, this will start impacting accuracy of ART algo and decay will rise. Couple of dB not a problem, but with +10dB or so you basically are close to, or even worse, where you would end up being with Bass control (would obviously depend on room and system).

 

[Lsc]

I think another issue is what you are listening to that would reveal whether or not a non-ideal step is audible. Dirac ART was made for people who watch movies but a car explosion may not reveal it as heavily as techno/tech house/modern EDM where kick fundamentals sit around 30-55hz or trap hiphop with short kick envelopes. I think maybe the population that is into calibration may not even be listening to the type of music that would reveal these issues. Sharing my personal opinion... I think there is a non-audible difference but its nuanced because it's probably not noticeable in other music genres or movies.

Edit: When I think about it these are not audible frequencies so when I say audible it's more of a feeling when I listen to that type of music. I think that is another reason why this is difficult. I feel that it's better and I'm crossing that with hearing when it’s really not hearing.

I found this image via Google (from Stormaudio). https://sp-ao.shortpixel.ai/client/...ontent/uploads/2024/10/Dirac-ART-bandeau.webp

The front left and surround right speakers are negating the right speaker’s first reflected sound.

Based on this ART would be more beneficial for music imho vs. movies where sometimes muddier bass is ok.

 

[Sokel]

I found this image via Google (from Stormaudio). https://sp-ao.shortpixel.ai/client/...ontent/uploads/2024/10/Dirac-ART-bandeau.webp

The front left and surround right speakers are negating the right speaker’s first reflected sound.

Based on this ART would be more beneficial for music imho vs. movies where sometimes muddier bass is ok.

Maybe the pic is unfortunate, but shows the left speaker having an 120° radiation pattern? I hope they mean lows, only, cause lines and reflections make no sense for higher as it shows it.