Right- I think we are trying to determine why the delta? I have not seen differences between the speakers as don’t really look into full files. If anyone wants to emphasise anything, would appreciate for the benefit of myself and others to post it out.
-
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!
Dirac ART is now running on beta FW for Denon Xx800H AVRs!
- Thread starter PHD
- Start date
FYI all your zip files are immediately triggering virus alerts and getting blocked with Windows Defender (10 pro):
I am no stranger to false flags myself but you should still check...
Curious... I have just scanned them with Windows Defender (10 Pro) and it did not find anything.
I use the same (Windows Defender on 10 Pro) and got no results.
I then checked using this: https://www.virustotal.com/gui/file...5c4dbfa3e87effbe0f9627fff394dd92ec7/detection
Again, nothing. Are you using heuristic checks by Microsoft? Those often produce false positives.
After I forced to allow one of them all other zips now downloaded without issues. It's 7-Zip packaging probably triggering false positives.
Just post your images if you want the attention. As I noted before, I will not go the .mdat files and would say many other will not. My setup and measurements are turning our they way they should be turning out - at as least as I can understand. Pls let me know if mistaken.
This is directional bass that Dirac is talking about with obviously strong support of the LFE channel, that is in turn supported by all bed channels. I did have quite a bit of experience with previous solutions, that were unfortunately not quite adequate or amazing, and what I can hear from ART is exactly what I would have hoped for.
If your mains default to same curve, it will not really fare well for directional bass, and at the same time, it might signify that these speakers are not fit to stand alone in the low end bass. Just a guess....
This is directional bass that Dirac is talking about with obviously strong support of the LFE channel, that is in turn supported by all bed channels. I did have quite a bit of experience with previous solutions, that were unfortunately not quite adequate or amazing, and what I can hear from ART is exactly what I would have hoped for.
If your mains default to same curve, it will not really fare well for directional bass, and at the same time, it might signify that these speakers are not fit to stand alone in the low end bass. Just a guess....
Since you don't have a sub, front mains are acting as subs and other speakers are crossed over with them at about 80Hz. This is why your bass responses are identical. A comparison of your measurements (raw, DLBC, ART) reminded me how bad DLBC calibration is:
Massive boost filters in the bass region certain to trigger distortion at much earlier volume levels. Increased group delay due to all this boosting:
EQ filters up to 20kHz certain to sound boxy. After that anything will be improvement of course but ART is doing its magic in the bass region: No group delay which means flatness is achieved without boost filters, perfect phase response in the bass region. Unfortunately it's still DLBC after the transition frequency ruining a magic bass calibration. Speaker distances are up to 15cm off identical to DLBC, quite damaging to immersive effects and there's no excuse to get these wrong.
I would manually align delays if at all possible with ART and cut correction as early as possible to get the best out of it.
Massive boost filters in the bass region certain to trigger distortion at much earlier volume levels. Increased group delay due to all this boosting:
EQ filters up to 20kHz certain to sound boxy. After that anything will be improvement of course but ART is doing its magic in the bass region: No group delay which means flatness is achieved without boost filters, perfect phase response in the bass region. Unfortunately it's still DLBC after the transition frequency ruining a magic bass calibration. Speaker distances are up to 15cm off identical to DLBC, quite damaging to immersive effects and there's no excuse to get these wrong.
I would manually align delays if at all possible with ART and cut correction as early as possible to get the best out of it.
Are you talking to me? You asked me to do measurements with Dirac Off and I complied. I've always seen people asking for .mdat files to properly evaluate measurements. If the opposite is true, by all means, let me know.
That's not the case. I have four subwoofers. I simply have not measured the LFE channel, which shows bass following the curve you'd expect if you look at all the other speakers.
It doesn't to my ears. But the main problem is that I cannot understand what you would mean by "boxy".
Here I'm not sure I follow you. It depends on which file you are referring to but, yes, latest measurements (the one using ART) have been done after I moved the speakers slightly apart, making them a bit off-axis to the MLP. I had read often that this was suggested for KEF speakers (due to the coaxial tweeter/mid) and I can say that the effect is most definitely more pleasing to my ears, in the high frequencies.
I posted those old results because people were thinking there were problems in my metodology, while I was seeing similar frequency response in the mid-high region from measurements taken many months ago and that were not questioned.
In any case, I'm willing to test your latest developments. With all the versions you published I don't know what version is the one to use, bearing in mind I want custom target curves, as an option at least.
Last edited:
Interesting choice to not include subs in any mdat file. This changes my boost filter comments. It was a bit too much boost on fronts even for Dirac and that explains it now.
You can understand what boxy sound is when if you ever hear a proper calibration.
15 cm differences are obvious in impulse graphs in dlbc/art zip file you posted earlier.
One of the reasons I never want to look into .mdat files. Especially from relatively new members. They often seem to confuse things, like what we are experiencing in this long an not so much productive thread.
What did I confuse? Could you point it out? I don't understand the turn this has taken, to be honest.
That's the point: "groups should be aligned with that" What does it mean? Aligned with what? Group by capability (e.g. all full range, all subs, all other)? Group by its placing (e.g. all fronts, all rears, all subs, all ceilings)? Group depending on the number of speakers because of maximum channel limitation (e.g. every speaker is one group for 2.1/3.1, place more in a group for 7.4.4/9.4.4)?
All I hear is "it depends". But it depends on what?
15 centimeters of difference between what speakers?
In what file? The microphone is centered at the listening position. I have just now measured from the tip of the Umik-1 to the top of the two front R7 Meta, with a laser meter. There's, give or take, two centimeters of difference, hard to be more precise than this (taking the measurements, I mean).
When I took measurements for Dirac I made sure that it was equidistant from the front speakers. Delays, set by Dirac after measurements, are 5.6 ms for each of them, 5.3ms for the center speaker and 6.5/6.7 for the two surrounds (which cannot be perfectly equidistant, for furniture reasons.
I have no idea what you are talking about - and noted before, will not look into your files. If you have something to post as screenshot, will look at it. I do note that you posted multiple iterations of your setup, so even if you posted them as screenshots, I would probably be lost and not able to contribute.
Focus and get your best game here is all I can recommend. Not seeing that at all though...
Well, I have to say that this truly, truly saddens me. Literally, genuinely sad here. The last thingk I was expecting was this. I could react aggressively but that would lead to nothing good and it's not the spirit with which I visit these forums.
I spent pretty much all day trying to come up with ways to make you and others understand the measurements I first posted. I posted other measurements just because Keith was doubting that the measurements were correct. I posted my settings for the same reason. You asked for measurements of the speakers without Dirac, I took them and posted them.
Now I'm being accused of generating confusion. Well, I don't know what to say. I either ignore people asking me for stuff or I don't. I chose not to, maybe I was wrong.
Just one thing I feel has to be clarified. The fact that you do not look at measurement files doesn't mean nobody does. As I've already stated, other people do and actually prefer to have them, instead of just the images posted on the forum.
Having said that, you asked for measurements with no Dirac. Here they are. I'm not even sure why you were asking for them but here they are.
Front left
Front Right
Center
Surround Left
Surround Right
And, as a comparison, same speakers with Dirac ART active
Front Left
Front Right
Center
Surround Left
Surround Right
ban25
Major Contributor
- Joined
- Nov 5, 2022
- Messages
- 1,288
- Likes
- 1,362
While it's certainly a bonus to include the .mdat attachments, I think most here will appreciate if you just post screenshots of your All SPL frequency response chart at perhaps 1/16th smoothing and you can limit it to 20-150 Hz or do full range (for ART we only care about up to 150). The spectrum decay graph is interesting as well. Without those screenshots, though, it's difficult because at least in my case, I'm not even on a laptop with REW installed at the moment, so there's no opportunity to even open the files.
Great. But have no idea why it turns out this way. I guess you should be happy that it does, looks pretty good.
Because I come here to learn and discuss with individuals that are passionate about music/sound and I come with the utmost respect and gratitude toward everybody helping and engaging with me. When, after a whole day spent doing basically nothing else but measuring, remeasuring, making sure settings were correct, it ends up with me seemingly being annoying (my interpretation of your "new members often confuse things"), not on "my best game", plus being questioned about distances between speakers without even telling me which speakers (by OCA, which I beg to let me know what he's referring to and how I can find confirmation in REW, since "real world" measurements with both tape and laser meter don't confirm his findings)... well, it's enough for me to be saddened.
And yeah, it's even sadder because after measuring yesterday, and listening to my system which sounded better than ever, I was more than ready to enjoy it for the whole day. Instead I've spent the day with REW and annoying people I respect.
These are your pre-calibration measurements. Speaker L is at t=0ms because it was the acoustic timing reference:
FL/FR and SL/SR are surely very carefully placed and almost the same distance to MLP (I guess you did some laser work to achieve that - I don't see that kind of symmetry often) but there's not much else to see in this graph other than your speakers actual distances to MLP. With the click of a button (Cross correlation) REW can time align these precisely as below:
Impulse alignment with cross correlation is a technique John added to REW years ago and it was known long before that. Latest REW iteration even align sub samples (those dots on the graph) perfectly using sophisticated interpolation techniques.
And this is post calibration after Dirac (same in DLBC and ART proving ART is still using the same algorithm):
Only Center speaker is correctly delayed, SR is more than 13cm off. There's almost always such errors with both Audyssey and Dirac because they use primitive methods unchanged since decades for reasons I will not get into. People with "up-firing" speakers will get even worse deviations (as high as their ceiling height) totally damaging immersive effects.
I guess you have also never been warned with your $900 calibration software that your surround speakers are both inverted and out of phase!
Oh, ok, thanks. So the "inexcusable" part was referring to Dirac, not to my speakers placement. I try very, very hard to do things carefully and following best practices but I'm never "sure of myself", even after years of doing this.
I cannot generate the exact same kind of graph you have but I came up with this, which I think shows something equivalent. Is this what you are referring to?
In any case, thanks a lot for getting back to me on this aspect. Unfortunately I'm not knowledgeable enough to give things granted for me. I need for them to be spelled out. Thanks for doing that.
As far as equalizing mid/high frequencies, you mentioned previously, if you see the results you can see that Dirac doesn't do much there for the front KEFs. It's needed more to bring the surrounds tonally in line with the fronts, I suppose. Would you concur? I could probably correct the fronts and center till 1kHz and stop there, doing full correction only for the surrounds.
Last edited:
Similar threads
