Couldn't tell you as I only have the base DL license on my 3800. I do agree moving the questions to AVS may be smarter as it's way more active over there!I thought so as well so why is it, that phase is so different between Dirac BC and ART (and also Audyssey)? Frankly, I never quite understood how phase works
-
Welcome to ASR. 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!
Active Room Treatment (ART) by Dirac
- Thread starter Flak
- Start date
So you may have been onto something. I did not realize I was using an outdated version of Dirac. Updated and recalculated the filters without making any changed on my side. I'll share retest when the house is quite again.
Thanks again @TimoJ
Attachments
Looks like that fixed the sub measurement New Measurements (ver 3.13.16)
Totally agree on support level… -18 for the default seems high to me. Things like default settings for support levels and how low speakers support I think are areas Dirac should improve on. The motto… first do no harm should be theirs as well. And your thought of less is more is also important. When I first signed up for ART Beta when I had my Storm I accepted all the defaults (it corrected up to 300hz in the beta vs 150hz now) and with the defaults it sounded so much better vs DLBC when comparing presets of both. Curiosity and wondering if it could be better leads you to trying things. For example my surrounds were fairly close to my main listening area and I wanted to hear what the correction cancellation signals sounded like. So I listened to some stereo music (not upmixed), muted all channels except for the surrounds to hear what the cancellation signals sounded like. It isn’t something I really noticed during stereo playback much at all if I recall correctly, but isolating it and hearing what it sounded like lead me to think to try lower levels of support, on more of the channels, in particular the ones that are close to the listening area and it did improve things for me because they were not audible even when listening for it. Other recommendations then made sense, for example not using the center channel to support any other speaker or group. Clarity for the center is critical and also the most heavily used in movies so it made sense to exclude it. My system had also expanded to 4 subs and I was curious about removing them. Because my surround and surround back speakers could support down to 50hz, ART used my 2 front subs, side surrounds, and rear surrounds to such good effect I removed those rear subs.
So with following some general guidlines found in the Stormaudio knowledge base, trial and error and some common sense ART is great. I would encourage people to definitely experiment and also read up on best practices as they can improve on the default settings. I sold my storm but have a Denon A1H to use ART with once my new room gets completed.
Any graphs?
My laptop crashed with all my measurements so just going off memory now … I have been without my Storm about 4-5 months and no finished room to take any additional measurements unfortunately. Going to a new room of in wall KEF 5160RLM/3160RLM with Perlisten in wall subs. Also the room will be so different.. new room is open with vaulted ceiling and old room was a dedicated theater. Look forward to seeing how it turns out with ART.
Curvature
Major Contributor
- Joined
- May 20, 2022
- Messages
- 2,040
- Likes
- 3,118
I'm not clear on what you mean, because they don't emit sound even in normal operation.
Are we looking at the same measurements? They significantly reduced decays. From that review:If you own one of these and like it I'm not going to tell you not to like it but they don't seem to do much in objective testing:
Fresh From the Bench: PSI Audio AVAA C214 Digital Active Bass Absorber
This article is a detailed review of the intriguing AVAA C214 Digital Active Bass Absorber from Swiss studio monitor and high-end speaker specialist PSI Audio. The concept is more or less the same as the AVAA C20, which PSI launched in 2017. the first commercially available analog active bass...audioxpress.com
If you have measurements with the thing on an off that are different please share.
Here's another set of measurements by Anselm Goertz of the older C20.
Keith_W
Master Contributor
Thank you for posting the measurements. I had a quick look.
Your MDAT contains individual measurements of FL, FC, FR, BL, BR, and LFE comparing DLBC and ART. Individual speaker measurements do not tell the full story of what is happening in the bass, especially with a scheme like ART which relies on main and support speakers. All the guys in this thread who are looking at individual speaker measurements from your MDAT are wrong (sorry guys). Fortunately all the measurements have an acoustic timing reference. So the first thing I did was vector sum them. Select all the measurements you want to sum in "All SPL", then right click and Vector Sum.
Here we can see an obvious problem. DLBC and ART have different target curves. There is substantially more bass with ART. In fact both DLBC and ART have way too much bass. Now, some people like that huge hump below 40Hz especially HT enthusiasts because that's where you feel the boom in explosions without colouring the upper frequencies too much, but the ART bass boost goes too high. It would make bass sound thick and muddy. I think this is most likely an error in your settings rather than a problem with ART. Would you be able to check, please?
What the different bass response means is that all decay measurements need to be normalized to peak at each frequency otherwise ART will be unfairly disadvantaged because there is more bass.
Here we have spectros of ART (left) vs. DLBC (right), with the same X and Y scales, and normalized to peak at each frequency. All I can say is WOW!! Looks like I have to eat my words! Before ART came out, I was predicting the trainwreck of the century. It really does work!
Now, bear in mind that this is not exactly a FAIR comparison, because DLBC has done a horrible job with the subs as we will see next:
Now let's look at the step response of DLBC vs. ART of the vector sum of all speakers. The third graph is a zoomed in view of the initial step comparing DLBC (red) and ART (purple). The objective is to look at the time domain performance. Those peaks I have arrowed are bass frequencies. A few things are obvious:
1. The initial peak of the step is much better with DLBC compared to ART. ART appears to be getting the woofer to play before the tweeter. There is about 17ms of woofer pre-ringing* (see third graph). You can barely see the tweeter impulse buried in that giant bass wave. I was concerned that REW had not aligned the impulse properly (which would give the appearance of more pre-ringing), but I examined the impulse and the t=0 was indeed correct. This is probably not audible - the pre-masking threshold is about 20ms for the limit of audibility (varies between frequencies), but it is not nice to see. If it WAS audible, it would sound extremely unnatural and objectionable. You would hate it and complain like mad! Since I don't see you complaining, it's probably OK.
* OK I know that technically it's not "pre-ringing" because this is a causal DSP filter and not an acausal one, but I can't think of a better term. I have to insert this disclaimer for the ASR pedants who might read this!
2. Those late peaks in the DLBC graph (top left) are very undesirable - they occur very late (71ms, 96ms, 135ms) and they are very loud, louder the main impulse itself. This would be well above the audible threshold. The audible threshold is somewhat debated, but no bass peak should arrive more than 50ms later than the main impulse. This would make the bass sound flabby and disconnected from the main speakers ... or "slow". The ART step response is better in this regard. Once again, some HT enthusiasts like a lot of bass ringing, it makes the bass sound powerful because it lingers longer.
Overall, I prefer the appearance of ART but it would be nice if there was a bit less pre-ringing.
And finally, this is the of FL and FR with DLBC. The others are not quite the same, but similar looking. I can see a lot of early and loud reflections which suggests a small room, or maybe a larger room with a lot of furniture. The ETC of the left speaker is very different to the right, which suggests asymmetric room placement. The FC (which I have not shown) has an early and loud peak which is missing in FL but not FR. It's either floor bounce or coffee table. My money is on coffee table.
The point is: ART managed to do a decent job even in a room with so many uncontrolled reflections. So yeah, I am impressed. And maybe I have a bit of egg on my face for all my ART naysaying.
Last edited:
Thanks for taking a look and the feedback. Please note I uploaded a remeasure after realizing my DiracLive version was out of date. Also, you do need to spl align the sub. I think there are a couple things I need to share before giving a verdict to what you are seeing.
Keith_W
Master Contributor
SPL align the sub to what? I would have thought that the measurement would already have the sub SPL aligned. In any case, if I SPL align it, it doesn't reflect what you have at home and what DLBC / ART did.
I’ll do some more data sharing and get back to you. The bass truly is not disproportionate in real live and need to find out why it appears that way.
An update on this. I bought the Denon 3800 and failed to get it to work with my two pairs of ACTIVE speakers. There simply is no way to switch between two pair of active speakers as front speakers (using the preouts) while sharing subwoofers. I would like to be corrected but unfortunately, I’m planning to sell it just a day after my purchase.
Last edited:
Just got ART loaded in my Denon 3800. I have a 7.4.4 setup. I did a 9 point measurement and the subs were all grouped together. I would like to separate them into 2 groups. Do I need to remeasure? I can separate them in Bass Control but ART won't let me separate them. Since it's using the same data I would think this would work.
Subs, ART and diminishing returns (on multi-sub setups)...
Just some thoughts I have been mulling:
Multiple subs have traditionally (last couple of decades) been deployed due to either a lack of SPL's or a need to balance out the nodes / modes of the room by using differently located subs
The first item (SPL's) can more or less be ignored, as many of todays subs can reach the desired home use levels without concern (typical for 75db listening level, 95 db peak for full range 105db peak for bass)
So we move to looking at resolving room related issues - which is where ART solves many of these issues.
So in an ART based setup, possibly with full range mains in the base layer (say down to 35Hz...) - would there be any real gain to having more than 2 subs?
Where previously diminishing returns on the bass would kick in at 3 or 4 subs, now with ART perhaps a setup with 2 subs can get the same results.
Does ART bring about a paradigm shift that moves us back towards full range speakers, as well as reducing the number of subs needed?
Note: I am not saying subs beyond 2 bring no benefit - I am however wondering if that 3rd and 4th sub, are past the point of substantially diminishing returns.... In that great balancing act of performance vs $$$ spent, is ART providing a paradigm shift as to where and how to spend our limited budgets?
Just some thoughts I have been mulling:
Multiple subs have traditionally (last couple of decades) been deployed due to either a lack of SPL's or a need to balance out the nodes / modes of the room by using differently located subs
The first item (SPL's) can more or less be ignored, as many of todays subs can reach the desired home use levels without concern (typical for 75db listening level, 95 db peak for full range 105db peak for bass)
So we move to looking at resolving room related issues - which is where ART solves many of these issues.
So in an ART based setup, possibly with full range mains in the base layer (say down to 35Hz...) - would there be any real gain to having more than 2 subs?
Where previously diminishing returns on the bass would kick in at 3 or 4 subs, now with ART perhaps a setup with 2 subs can get the same results.
Does ART bring about a paradigm shift that moves us back towards full range speakers, as well as reducing the number of subs needed?
Note: I am not saying subs beyond 2 bring no benefit - I am however wondering if that 3rd and 4th sub, are past the point of substantially diminishing returns.... In that great balancing act of performance vs $$$ spent, is ART providing a paradigm shift as to where and how to spend our limited budgets?
I’d say 4-6 subs. All corners and the middle of the side walls. But I only have 2...
Question is, what was the increment of improvement from 1 to 2, from 2 to 3 etc...
was 90%+ of the gain in the 1st sub?, 2nd?, 3rd?
Yes we all would like a "ne plus ultra", ultimate system - but most of us have various constraints... so the question becomes where is the optimal balance?
Applying the old Pareto principle (80/20 rule) that 80% of effects are from 20% of causes... (which is a pragmatic old rule of thumb... an observation of life and practical reality) - we can focus on the 20% to achieve the 80% improvement... and thereby we have a mechanism for threading the needle of such quandaries as how to best optimise our audio systems.
Hence my suggestion (and question) as to whether in an ART based setup, the optimal configuration might shift - spend more of the limited budget on the base layer, get full range speakers, spend less on subs, get fewer subs, etc...
Are there more available for those willing to spend greater $$$ in causes with lesser effect - of course there are! The neverending search for perfection....
But for the more pragmatic amongst us, those with a more limited budget and / or WAF constraints - might there be a better optimum balance - once you implement ART?
You should not need to remeasure.
3 is enough to enable ART.
I would expect these to be measurements of channels - not physical speakers - and would include any signals from their respective supporting speakers. Could ART users who verify their results with REW please confirm this?
Similar threads
