• 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!

Introducing Directiva - An ASR open source platform speaker project

OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Hi folks! I have family visiting from out of town for the weekend so have limited time for forum responses.

@ctrl is the best audio analyst I know and so I put my faith in his opinion regarding speaker measurements. IMO, there could be many reasons for differences between Amir’s and mine. At the very least, all the evidence points to the different woofer revisions as @McFly documented. I have also noted many times that my test rig is not anywhere close to a Klippel even on a good day!

More later but hope you get to enjoy the weekend in any case!
 

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,086
Location
.de, DE, DEU
As I have previously explained, diffraction from the horizontal baffle edges, i.e., the top and bottom edges, in theory creates on-axis response peaks and dips at specific frequencies; these predicted peaks and dips and readily apparent in Amir's measurement but are not apparent in Rick's measurement.
to me to point out that the major discrepancy between the two measurements are explained thusly, that one (Amir's) shows the effect of diffraction associated with the top and bottom edges of the baffle, while the other one (Rick's) does not. Of course this raises the follow-up question of why Rick's measurement does not show this diffraction effect.

Rick can't eliminate the influence of the speaker cabinet edges (edge diffraction) in his gated measurement - how should this be possible? And I would also rule out the possibility that Rick sets too large a gate and thus allows room reflections to creep into the measurement.
But conjecture is not circumstantial evidence, so there is more circumstantial evidence in the following sections.


When calculating the theoretical frequency ranges in which the effects of edge diffraction are to be expected with the simplest means, one must be careful to estimate their influence correctly.

Therefore I have shown in Post#694 the BEM simulation for the woofer in the loudspeaker cabinet and the large deviations between BEM simulation and Amir's NFS measurement that occur when normalizing to the axial frequency response. Maybe this was too abstract and complicated to understand.

Therefore here is the simple baffle simulation of the Purifi woofer on-axis in the Directiva enclosure that VCAD provides (the influence of the cabinet depth is not taken into account here):
1630829977522.png


VCAD now offers the possibility to include a half space measurement (infinite baffle measurement) and to simulate it based on the baffle simulation in full space (4pi).
For the infinite baffle measurement from HifiCompass of the Purifi woofer we get this:
1630830381929.png
In this way, I took the on-axis half space measurements of the woofer from HifiCompass and the manufacturer and simulated them in the Directiva enclosure for full space, exported them, and compared them to Amir's and Rick's measurements:
1630831544052.png


Also the VCAD simulation, like the BEM simulation, ignores the protruding surround of the woofer.

Shadowing theory
Now you have to realize that the theoretical maximum around 860Hz for the vertical edge diffraction corresponds to a wavelength of 40cm. The surround protrudes about 1cm. So everybody can imagine how big the influence of the shadowing should be. Nevertheless, let's assume that the shadowing has a significant effect on the vertical edge diffraction.

1) If in Amir's NFS measurement the protruding surround was not in the way vertically, then the NFS measurement should agree well with the simulation.

2) If in Amir's NFS measurement the protruding surround was vertically in the way, then the maximum at 860Hz is attenuated (because of shadowing) and frequency curve should be below the simulation in this frequency range.

No matter if 1 or 2 in both cases the NFS measurement follows the course of the simulation less well than Rick's measurement does.
Which again is a confirmation that Rick's measurement seems to be okay and that the NFS measurement does show unexplainable deviations compared to the simulation (apart from far-fetched theories like protruding measurement platform edges).
 
Last edited:

ctrl

Major Contributor
Forum Donor
Joined
Jan 24, 2020
Messages
1,616
Likes
6,086
Location
.de, DE, DEU
@ctrl is the best audio analyst I know and so I put my faith in his opinion regarding speaker measurements.
Thanks for the flattery.
...and how many "audio analysts" (didn't even know the term existed) do you know? ;)
Even after more than twelve years as a hobby LS designer, I still get it wrong from time to time and feel like a beginner. Every new day offers the chance to fail.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Thanks for the flattery.
...and how many "audio analysts" (didn't even know the term existed) do you know? ;)
Even after more than twelve years as a hobby LS designer, I still get it wrong from time to time and feel like a beginner. Every new day offers the chance to fail.

Not sure, but was meant sincerely! Maybe technical audio analyst pumps it up even more? I think the forum uses Technical Expert as a formal title...
 
Last edited:

witwald

Senior Member
Forum Donor
Joined
Jul 23, 2019
Messages
478
Likes
502
How about no ports because they're stupid and make the phase messy and makes subwoofer integration a nuisance?
Why would a vented enclosure make subwoofer integration a nuisance? For example, with a speaker like the JBL LSR308P and the LSR310S subwoofer, subwoofer integration is relatively easy. Site the subwoofer close to the the main speakers, and adjust the output of the subwoofer to achieve a flat response. The built-in 4th-order in-phase Linkwitz-Riley filters with the pre-set 80 Hz crossover frequency blend the outputs together very well.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Hi Directiva fans! Our guests are gone but still recovering from a bit of overindulging over the holiday weekend. :facepalm:

Still planning to share some of my process of designing a crossover on the r1 build thread. Will show what might be done without the help of Klippel measurements and an experienced designer.

As for having Amir help complete and review the r1 speaker, appears we have hit a snag. Right after it got shipped back to me, Amir tried to recall it. Unfortunately, the shipping company did not seem to be able to do until it was almost on my doorstep. For most of last week, we were not sure where it was. Amir put in a call and was told to wait 5 days and put in a claim. The claim has been made. Amir says this sometimes shakes the tree and we may still recover the speaker. I am less optimistic and am preparing to build another. If I do have to build another, will use as an opportunity to improve the build instructions. Stay tuned and will let you know as soon as we get an update. :)
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Some more promising news...

About a half hour ago, the shipper sent an email indicating the speaker is on its way back to Amir!
 

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
That's good, that's a pricey woofer to lose!
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
That's good, that's a pricey woofer to lose!

Hopefully, not an issue, but would be covered by insurance in any case. At this point, it is more about the time to build another one if needed. :)
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Hooray! The speaker is back with Amir. We are working on getting his minidsp HD wired and configured.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Hi folks!

A couple of quick notes for those following this thread. Thanks to the effort of @AdamG247, I had some content moved from the build thread to here. The build thread was getting cluttered by some interesting discussion but would be distracting for someone trying to build a r1 speaker. This move may have caused some limited discontinuities as the build content merged here based on chronology. I regret this but keep this partitioning in mind if you post content in the future.

As I have in the past, will try my best to service both threads, but am pretty busy rn trying to get the build thread caught up and arranging for Amir to test r1. So, in keeping with my earlier direction, will use this thread to track Amir’s progress while I beef up the content of the build thread. Sorry but I may not respond to every post. If you feel your post may have been neglected and needs my attention, send me a PM. Thanks!
 

AdamG

Proving your point makes it “Science”.
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,636
Likes
14,919
Location
Reality
@Rick Sykora yeah keep this in mind for future Move requests. The forum Overmind will happily move posts from one thread to another. But then it decides on its own mind you, to place each post based on the associated time stamp. I know this is not an optimal outcome but the Overmind will punish me if I complain too much. :oops:
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,370
Likes
234,411
Location
Seattle Area
First try at NFS scan is running on Directiva. Had to troubleshoot a lot of stuff to get to this point. Will finish past midnight so processing and analysis will be tomorrow.
 

GimeDsp

Senior Member
Joined
Oct 30, 2020
Messages
418
Likes
362
Location
Earth
First try at NFS scan is running on Directiva. Had to troubleshoot a lot of stuff to get to this point. Will finish past midnight so processing and analysis will be tomorrow.
Can't wait too see the results.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Amir completed a scan and the result was pretty good, but not the expected one. Along with an errant bass filter that caused the troubleshooting he refers to earlier, I found a couple of filters from an earlier design that should have been disabled as well. These caused the targeted voicing to be more pronounced than @ctrl had intended.

We have fixed these issues and sent an updated XO to Amir for his official review. He is rerunning the Klippel tests and should have something to share soon. With the better results coming so soon, see little value in posting the earlier round.

Thanks for your patience and understanding!
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Got a partial result from Amir overnight...

The good news is it shows the sim is correlating well to his measurements and the response is fairly flat. Bass response is impressive for a fairly small speaker. Until I hear (and see) more from him, still not sure we are quite where we want to be for him to do a full review. Please excuse the wait, but we are dealing with 3 major time zones diffs. Should know later today.

Stay tuned!

Rick
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,028
Location
Stow, Ohio USA
Ok folks, here is the plan...

Amir has deemed that Directiva r1 is a "major accomplishment" and he will be completing his usual review process and posting. Since we have kept you waiting, here is an early look at the results...
1633289555591.png


Barring any changes based on Amir's listening tests, I will be posting this crossover configuration to the build thread.

Thanks to all who shared their expertise to get us here, but notably to @ctrl for his outstanding contributions to Directiva's design and @amirm for his support and Klippel measurements!

As Directiva was conceived as an open project, we hope to see the r1 speaker design refined and improved over time. As some of you know, I would like to do a larger Directiva speaker, but still need to wrap up the first one! Notably, the build thread as it will serve as the r1 project documentation.

Cheers,

Rick
 
Last edited:

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Ok folks, here is the plan...

Amir has agreed that Directiva r1 is a "major accomplishment" and he will be completing his usual review process and posting. Since we have kept you waiting, here is an early look at the results...

index.php

Barring any changes based on Amir's listening tests, I will be posting this crossover configuration to the build thread.

Thanks to all who shared their expertise to get us here, but notably to @ctrl for his outstanding contributions to Directiva's design and @amirm for his support and Klippel measurements!

As Directiva was conceived as an open project, we hope to see the r1 speaker design to be refined and improved over time. As some of you know, I would like to do a larger Directiva speaker, but still need to wrap up the first one! Notably, the build thread as it will serve as the r1 project documentation.

Cheers,

Rick

The early look is missing :p
 
Top Bottom