I may have "wrong" opinion about purpose of sticky threads. Individual can value and like anything - just like some dogs like to eat sh*t so opinion of author and any (absolute) number of members - experienced or not is not good reason to elevate to sticky. Members who don't know that it's not reliable method could step and stay in the same pile of sh*t while having resources and motivation to more accurate and reliable.
My experiences with designing crossover without measurements are not good. Target was just to study whether combination of horn and driver is worth to purchase and test in full size prototype cabinet. Manufacturer's data was total bollocks so whole projects was just waste of money and time.
More accurate manufacturer's data and less quality variation in component individuals exist, but we can never be sure without doing properly.
Designing with dual plane far field and near field measurements is actually quite fast though prototype or final cabinet should exist to proceed. I would say that it's as fast/slow as tinkering with data processing apps and calculations without measurement data. I've designed final XO in few hours including measurements.
So I would not elevate that thread to sticky. Add warning about reliability/accuracy and quick description about recommended method to the beginning. Then let it flow to history and save the link in case someone needs and asks information in the future. Hopefully no one.
Same type of problem with 'So you want to design your own speaker from scratch!'. If we can agree that single axis XO simulators are not the most recommended (for other than very near field monitors) due to too narrow scope to reality, it's also easy to decide that everything which support single axis far field only should be replaced in that thread with methods and tools supporting at least quasi (dual plane) half and full space. wintermute as an author should study those tools or ask someone else to write instructions. Not to publish just what he prefers or has experience with.
And generally, I'm done with mentioning anything about problems with single axis methods to beginners who know and start with PCD or XSim etc. I really do respect authors of those tools, but I do not respect decades old limited traditions still strong in North America, and culture in diyaudio which does not encourage to elevate and kick those tools out in every thread right away when someone starts with them.
I'm also fully aware that good speaker concept and acoustic design enables totally blind design with on-axis only, but gathering knowledge from individual designs which enable that probably requires methods and tools which are able reveal problems in full space behavior. Usually there are some practical restrictions which prevent optimal acoustic design so designing XO with quasi full space data has at least chances to compensate e.g. by selecting XO frequencies and slopes producing the most balanced total. Sorry but I don't believe a second that beginners using PCD or XSim would be able to cope acoustical design, c-c / XO f etc. so well that on-axis would produce excellent result to full space in reflecting environment.
Last edited:
Member
Joined 2003
I think a big reason that many people on this side of the world get sucked into that single axis design path is that USB mics are popular and more commonly available. Ominimic, UMM-6, UMIK, etc. This limits you to a single mic position for measured data in order to capture delay between drivers accurately. Many DIYers want a plug and play design solution, which is why they gravitate towards a USB mic. An XLR mic, 2 channel measurement jig, phantom power, can all sound rather complicated so they buy a USB mic without understanding the limits of a single channel measurement system.
Umik-1 is popular also in Finland, but most of the owners use it for room acoustics measurements and LF equalization with DSP gear/app. Few individuals also for iterative XO design, driver measurements and quality control of complete speakers. I guess no one for driver measurements so that excess phase is removed with minimum phase extraction.
Cheapest calibrated mics and dual channel sound cards are as cheap as Umik-1, but much more versatile so this fashion looks strange from diy perspective where everything should be free or cheap. I should not mind, but unfortunately I do.
Cheapest calibrated mics and dual channel sound cards are as cheap as Umik-1, but much more versatile so this fashion looks strange from diy perspective where everything should be free or cheap. I should not mind, but unfortunately I do.
I've not actively updated VituixCAD thread at HH.org in few years. diyaudio has been the main forum. Some local forums - especially in Germany are quite active, but without me due to high language barrier.
Hi Kimmo,
A few comments, for what is it worth, from a experienced builder of other peoples designs and renovating existing builds, but a total newbie to own designs.
- I "hate to see you go" (the sentence you often get when you are leaving some earlier accepted info delivery). If you feel exhaused/feed up/etc and like to paus/leave I can fully understand you (and other fellows) as you must put down endless of hours into your "baby". I have always admired you guys having the drive to do such a marvelous thing for other people, and at the same time defend the choices you have done.
- As a newbie I ask questions that could be previously answered, but sometimes not always easy to find, i know that. So maybe a "support thread" could be a good idea as already suggested. Hosted and supported by all the ones that actively use the software and support all the work that you put in to it already by understanding and programing.
- Another thing for a newbie and also as new to the software. If I read the manual and look around on the kimmosaunisto homepage I find a lot of great information of the content and functionality of the software. That alone take quite some time to create and update. A Q&A takes time to create (yes I have done several) but also reduces the support needs. Maybe such a thing could be continuosly brought together as a part of the support thread. No (or few) new answers on the same clear and distinct stuff in the thread, but instead a continuosly updated Q&A. The answer in the thread will be a link to the Q&A. Of course a healty discussion reagarding outcomes of usage, different problems etc should be there in the thread. But less "standard questions"
- One hard thing I guess would be to be helpful to all and everyone with new features as you and others in the same situation is trying to do. That´s really super great, and maybe challenging in a positive way but also I can imagine it can be exhausting. Maybe a "change prio list" could be a part of the support thread. When a suggestion is coming in a "board" of some few experienced users can sort them out, prioritize it (maybe together with your advise on programing complexity).
- Another thing when you go into a new thread is to try to understand the "position" of it. I am a member of a few alternative cancer treatment communities. The ones that are working the best (and also are based on medical studies alone, no "guesswork") have quite a detailed info. As soon as the first (and already answered) question or proposal of doing "something in a better/other/etc way" comes up that is outside the original purpose (of the creator of the community) the answer is (by several moderators working together) "please read this...". Reading the intro, the manual and the kimmosaunisto homepage give me a lot of information. BUT maybe a clear initial "Statement" First in the thread, First on the homepage, First in the manual and as a separate linkable document would have helped me better understading the position and standpoint (also versus other softwares). Within the community of VituixCad we prefer/believe/etc X y z, we do z y z and do NOT x y z. Just a thought.
Again this will involve some work, by experienced people, which excludes me. If I can do anything to help, being the newbie I am (but also a 30+ years of successful professional project/program management in complex areas, with "risk & structure" as favourite areas) I will certainly try.
A few comments, for what is it worth, from a experienced builder of other peoples designs and renovating existing builds, but a total newbie to own designs.
- I "hate to see you go" (the sentence you often get when you are leaving some earlier accepted info delivery). If you feel exhaused/feed up/etc and like to paus/leave I can fully understand you (and other fellows) as you must put down endless of hours into your "baby". I have always admired you guys having the drive to do such a marvelous thing for other people, and at the same time defend the choices you have done.
- As a newbie I ask questions that could be previously answered, but sometimes not always easy to find, i know that. So maybe a "support thread" could be a good idea as already suggested. Hosted and supported by all the ones that actively use the software and support all the work that you put in to it already by understanding and programing.
- Another thing for a newbie and also as new to the software. If I read the manual and look around on the kimmosaunisto homepage I find a lot of great information of the content and functionality of the software. That alone take quite some time to create and update. A Q&A takes time to create (yes I have done several) but also reduces the support needs. Maybe such a thing could be continuosly brought together as a part of the support thread. No (or few) new answers on the same clear and distinct stuff in the thread, but instead a continuosly updated Q&A. The answer in the thread will be a link to the Q&A. Of course a healty discussion reagarding outcomes of usage, different problems etc should be there in the thread. But less "standard questions"
- One hard thing I guess would be to be helpful to all and everyone with new features as you and others in the same situation is trying to do. That´s really super great, and maybe challenging in a positive way but also I can imagine it can be exhausting. Maybe a "change prio list" could be a part of the support thread. When a suggestion is coming in a "board" of some few experienced users can sort them out, prioritize it (maybe together with your advise on programing complexity).
- Another thing when you go into a new thread is to try to understand the "position" of it. I am a member of a few alternative cancer treatment communities. The ones that are working the best (and also are based on medical studies alone, no "guesswork") have quite a detailed info. As soon as the first (and already answered) question or proposal of doing "something in a better/other/etc way" comes up that is outside the original purpose (of the creator of the community) the answer is (by several moderators working together) "please read this...". Reading the intro, the manual and the kimmosaunisto homepage give me a lot of information. BUT maybe a clear initial "Statement" First in the thread, First on the homepage, First in the manual and as a separate linkable document would have helped me better understading the position and standpoint (also versus other softwares). Within the community of VituixCad we prefer/believe/etc X y z, we do z y z and do NOT x y z. Just a thought.
Again this will involve some work, by experienced people, which excludes me. If I can do anything to help, being the newbie I am (but also a 30+ years of successful professional project/program management in complex areas, with "risk & structure" as favourite areas) I will certainly try.
Last edited by a moderator:
This thread is the main reason I visit this site.
And VituixCad is the main reason I feel confident about developing loudspeakers that perform well in actual spaces and overcome some limitations that have always bugged me about existing designs. I could not do this with any other software.
And VituixCad is the main reason I feel confident about developing loudspeakers that perform well in actual spaces and overcome some limitations that have always bugged me about existing designs. I could not do this with any other software.
Member
Joined 2003
Yes, of urse USB mics have their place, I own an Omnimic myself, it's great plug and play solution for setting up EQ, but not for design use.
I don't think cost is the issue, I think it is the complexity of the jig setup for new users, and lack of understanding of what 2 channel measurement is and does. New users often have issues with understanding of phase, delays, destructive , so don't know what they're missing when they buy a USB mic for speaker design, and follow the herd that points them to instructions on determine delays with single channel systems.
Maybe another sticky is welcome, detailing differences between single channel and dual channel measurement method, and promotion against tracing manufacturer data with big warning of potential for error.
Hi Kimmo,
I said it many times before, and I reapeat myself:
Thank you so much for vituixCAD!
I really enjoy working with it, because it is designed very well.
And even if I don't always use it the way you expect it to be used,
the software increased the quality of my speaker designs a lot
(even if there is even more potential in the software).
I understand, that you might not want to go on the current way.
But I hope that it will go on in any way.
Best Regards
Matthias
I said it many times before, and I reapeat myself:
Thank you so much for vituixCAD!
I really enjoy working with it, because it is designed very well.
And even if I don't always use it the way you expect it to be used,
the software increased the quality of my speaker designs a lot
(even if there is even more potential in the software).
I understand, that you might not want to go on the current way.
But I hope that it will go on in any way.
Best Regards
Matthias
Hi Kommosto can I ask you to load the newish and quite popular Eminence sub the 21"monster NSW6021-6
Many thanks if you can!
Last edited by a moderator:
You can add the Thiele & Small parameters yourself via the enclosure tool using the edit function.
See how to do this by using the online manual.
See how to do this by using the online manual.
Thanks for your feedback! I got impression that interfering what methods and tools other members use and recommend is sensitive topic i.e. VCAD users are not eager to say opinion what is better/best for the others. Yesterday I read 'Happiest Man on Earth' by Eddie Jaku. I started to think am I hater or helper. Communication may look the first one at least sometimes, but it's still closer to helping for free than just shouting. I could take my coat and keep mouth shut, but that would not help anybody (except me). VituixCAD thread is four years old, but not much happened. Same old on-axis without measurements etc.
---
One tread should be enough for all discussion about VCAD and measurements valid also for VCAD (method is valid for any XO simulator having movable drivers).
Discussion about individual diy projects to separate threads in multi-way area. I won't comment if measurement data is invalid i.e. simulated, traced or on-axis only if project is more than just preliminary study. I hope VCAD users would inform designers about proper usage.
Feature request primarily to e-mail. Small, easy and fast changes can be done, but I usually resist almost everything for a while. I also try to avoid reversing. E.g. moving short, open, invert, mute, rotate commands from context menu to buttons and then back was educative but not smart.
---
One tread should be enough for all discussion about VCAD and measurements valid also for VCAD (method is valid for any XO simulator having movable drivers).
Discussion about individual diy projects to separate threads in multi-way area. I won't comment if measurement data is invalid i.e. simulated, traced or on-axis only if project is more than just preliminary study. I hope VCAD users would inform designers about proper usage.
Feature request primarily to e-mail. Small, easy and fast changes can be done, but I usually resist almost everything for a while. I also try to avoid reversing. E.g. moving short, open, invert, mute, rotate commands from context menu to buttons and then back was educative but not smart.
I'm definitely a fan of ports on the rear over ports in the front in the majority of cases. There are many advantages and few disadvantages. The biggest advantage is a cleaner midrange in a 2 way system.Thank you Kimmi,
So once again as almost always it´s a trade off.
I guess, at least from a theorethically perspective, I, as many DIYs and manufacturers seem to, prefer the A version. For it´s possibly less problematic mid doe to less "leaking and interacting mid signals" from the port. There are so many discussions on the internet, with various "thinking" but less actual theory that doesn´t really help if one like to understand.
This mean that the most important thing now is to integrate the measures from the port position in A in a correct way into the XO design.
The ~40 litre 2 ways I use now have the midbass driver in the middle of the panel, tweeter half way between the middle and top and originally had a single 75mm diameter port in the front symmetrically opposite the tweeter half way between the midbass and bottom of the cabinet.
They were like this for a while however I realised the port area was insufficient to avoid chuffing but didn't want to put an even bigger diameter hole on the front as the existing port already let you see right through to the lining on the back panel...
With nowhere else to add another 75mm port due to the tweeter taking up the only other visually symmetric position I installed a second port on the rear panel directly behind the tweeter. (And increased the length of both ports to re-tune to the same box frequency)
Chuffing problem solved and significantly better bass performance at higher SPL due to the port not suffering from turbulence losses.
They stayed like this for years as various crossovers were tinkered with, but from time to time I would take measurements with either front port or rear port blocked off comparing that to both ports open and realised that aside from the expected changes in bass that blocking only one of two ports would have, there was a very obvious change in the midrange as well around 600-700Hz, which is the pipe resonance of the PVC tube used in the port.
And that's measured a metre away with a proper gated measurement. The difference was not subtle with a narrow band peak and notch like a zigzag of more than +/- 2dB in the 600-700Hz region. And the insides of the cabinet were lined with thick wool felt including the rear panel visible through the port.
Both ports blocked - no zig zag in the midrange, rear port only open - no zig-zag. Front port open, obvious zig-zag. When doing A/B comparisons there was an obvious subjective difference - the midrange sounded harsher and more pushy with the zig-zag in the response and smoother without it, pretty much as you would expect. Again, not subtle.
While the rear port in theory will have the same pipe resonance issue, because it is radiating backwards and the resonance is well above the baffle step frequency of the cabinet as a whole, the cabinet will prevent that midrange signal travelling around the cabinet, and any reflection you might get from a wall is attenuated and delayed to the point where it can't be heard or measured.
Even with the speaker at a normal distance from the wall behind it I was unable to measure any trace of the midrange pipe resonance from the rear pipe in front of the speaker.
Many more years went by and when I finally finished up the crossover and started laminating the cabinet I decided to finally get around to moving the bottom port to the rear so I glued a wooden plug into the hole at the front and laminated over it and installed the port at the rear, so it now has two ports at the rear - one at the top third and one at the bottom third on either side of the midbass and I'm glad I did.
Other advantages of rear ports - more visually appealing IMHO.
No hole to look into from the front, with a less cluttered cleaner look, no hole to tempt small children to insert their toys etc. (very important
)It also makes the front panel stronger having no port cut outs in it and removes a sharp(ish) edge for midrange or treble to diffract from. (It makes me wince seeing small commercial speakers with their ports right beside a tweeter)
Depending on what frequency the port is tuned to and whether the speaker is near (but not too close to) a wall having the ports on the rear tends to increase the bass performance of the speaker slightly as the radiation from the port is closer to the boundary and you also have two sources of bass through the port/cone overlap region that are spatially separated by the depth of the cabinet - on a larger cabinet this can be a significant effect and noticeably reduces boundary cancellation in the bass as the radiation from the driver and the port don't suffer from boundary cancellation at the exact same frequency like they do with a front port.
The only time I would avoid fitting ports to the rear is on a bookshelf speaker where there is a chance they will be pushed very close or up against a wall. In that case a better option would be to have them on the side instead - the side pointing to the opposite speaker, so on the right hand side of a left speaker, so that they also won't be too close to a wall if the speakers are close to a sidewall. But my default position for free standing speakers would be to install ports on the rear unless there is a good reason not to.
Last edited:
I can to a 100% relate to what you write Dcibel. I guess many of us newbies would like to do it “the right way” but already selecting the mic with eventual extra hardware is a long bridge to pass. Reading all these different forums searching for “the answer” it’s easy to get quite confused, not even on a higher level. And then if one realise that the stuff will be used for very few projects, or even worse, only one, most people I guess are eager to get as Big Bang for the buck as possible.
And then move on to finding a way how to do it, which truly isn’t easy as a total newbie. And THEN move on to actually do the measuring.
Last edited:
Thanks for the interesting story and information. This is something I’ve read about and good to hear. I have used Hornresp for designing the cabinet with the ScanSpeak D8531 midbass and I think I got it really good as a theory. Now I have to select mic and hardware, learn how to and then measure far and near field on the midbass, near field on the port and far field on the tweeters. And then learn how to combine the stuff in VituixCad and start thinking of the xo design. So quite many newbie steps.
Wise words and good thoughts! Keep up the good work, you are without any doubt a helper!!!!
And please everybody; never forget that different cultures have VERY different definitions of rude vs. honest, the dutch ( and fins probably) are much less sensitive and more direct (some will say rude) than others. This is neither good or bad, we are all just used to different communication.
There is a paypal donate button for Vituixcad - I suggest anyone who has used the software to create an actual working speaker design seriously consider using it to show their appreciation for this software. I have.
I've been tinkering with speaker design since the early 2000's and I would have done anything to get software as good as this back then - with the exception of a couple of very expensive commercial packages that I could never afford or justify there was nothing like this available then.
I got by with (ahem) a bootleg copy of SpectraLAB, a spice based circuit simulator I think called e-Sketch ? (paid for but relatively cheap) for very basic passive crossover modelling (but with no integration with driver measurements) a few spreadsheets and other tools like WinISD, The Edge, and a lot of hand tuning of crossovers and trial and error.
Nowadays Vituixcad will basically do everything you need for speaker design with the exception of taking the actual measurements (which I use the paid version of ARTA for) and CAD for cabinet design if you're that way inclined and don't just use pencil and paper. (I used Sketchup when it was still owned by Google)
The fact that one program can do so much, do it so well and in such a user friendly fashion is incredible, especially when it is donationware. This is definitely not the norm for freely available software.
It let me do a brand new, relatively complex passive crossover design for a speaker that had been on the boil for 15 years with various temporary crossovers in use, and get it done more or less from start to finish in about 3 weeks of spare time.
The simulation accuracy gave me such confidence that I ordered the final components (around £130 worth) without ever prototyping the the design in hardware first - the components arrived, I assembled it all and apart from one or two components that were out of tolerance and had to be padded/swapped, the measured response was an almost perfect overlay over the predicted response, and in several months of subjective listening tests following that I only made two or three very subtle tweaks to the network from my initial starting point until I was happy.
Incredible. I was so blown away that I felt compelled to donate to show my appreciation for the software, which I wish I'd had 10-15 years earlier...
Definitely.
I do sense an undercurrent of frustration and burnout from Kimmo and I can relate to that.
For a number of years I was highly involved in an open source project - as both a developer, debugger, tester and also providing end user support on the forum.
Although I occasionally got free prototype hardware (needed to actually do the development and debugging etc) I wasn't paid in any way or otherwise compensated for my time.
At the peak it was consuming most of my free time month after month and I started to burn out. The first place I burned out was on the support forum. There were just too many people with too many problems asking too many questions who needed individual help to debug and troubleshoot problems and not enough support members on the forum with sufficient knowledge to give that support.
I felt individually responsible for all these people having problems (after all they were running code I'd written and tested) and compelled to answer whenever I could but eventually just felt like I was trying to hold back the floodgates in a storm as the number of threads unread by me only ever increased.
I started to get pretty unhappy and resentful about the time sink and feeling that the job was just never done. I was spending more time on the forum trying to help people that I was actually doing coding/developing.
Around that time we had a kid and my free time plummeted. At that point I had to let go and I more or less stopped visiting the support forum altogether apart from the very occasional quick browse to see what was happening and to be honest it was a huge weight off my shoulders and made me feel so much happier. I still had enough time to do a little bit of developing and testing but I reduced my time there as well to find balance.
In this thread there seem to be two different conversations going on - how to use Vituixcad, and how to design speakers. The latter one probably doesn't really belong here and I'm as guilty as anyone at stoking this kind of discussion along. (Just look at my previous post...)
I can see how it would be easy to get sucked into an endless debate about design philosophy and techniques when this thread should really just be about the software itself. Perhaps focusing just on the software would help reduce the burnout that may be happening.
Anyway like the others here I'd like to thank Kimmo for this great software and all the time he has put into it working on it. Working on free software can be a thankless job - I have first hand experience of my own...
Last edited:
Dear Colleagues, I have a question about measuring YG Acoustics Hailey 2 style floorstanding speakers where the bass driver is located right at the floor. While the measurement of the tweeter and midrange is clear (measurements on the axis of each speaker), what should the procedure be for the bass driver? Should the speaker be raised to a height of about 1 m to take the measurement on the axis of the bass driver? What ideas do you have for the crossover design to "account" for the bass boost from the floor surface in typical listening? Maybe for the bass measurement the microphone should be on the axis of the tweeter, midrange, or maybe make an additional measurement of the pseudo ground plane with the microphone at the floor?
Attachments
There is a paypal donate button for Vituixcad - I suggest anyone who has used the software to create an actual working speaker design seriously consider using it to show their appreciation for this software. I have..
.....Perhaps focusing just on the software would help reduce the burnout that may be happening.
Anyway like the others here I'd like to thank Kimmo for this great software and all the time he has put into it working on it. Working on free software can be a thankless job - I have first hand experience of my own...
Good and interesting story. I’ve been there myself, but I didn’t stop until my brain stopped me. 1,5 years of recovery. So to Kimmo, really, take care!!!
Donate, a good thing.
“Focusing on the software” is a good thing but maybe it need to be a little bit more than that. No philosophical discussions regarding whatever things like “what is the best speaker to build”, or too wide “I like to build a ...., which one is best” etc. Looking at myself as a design newbie I can say that three areas are tightly linked together, with (if one like) VituixCad very central.
- Hornresp was great for designing what I looked for, MLTL for a 6-7” midbass with SS D2608 + a Wave Guide. It is powerful and quite serious.
- Then measuring and XO design comes next. Where VituixCad is a serious choice. And if you are here you are probably interested being somewhat serious.
- Mic’s are more hard to see what is fairly “Best buy” from Budget to Std to Serious if you would go with VituixCad. Not directly linked, I know, but in a way yes. Would need some dialogue - OR a short “Speaker design with Vituixcad” white paper with a few choices.
- Measuting software. There are quite a few where REW and ARTA seem to stand out. Which one to prefer together with Vituixcad etc could be a nice dialogue - OR a short paragraph in the white paper which one(s) to choose.
As I have been a project manager for complex softwares (in banking) I love the information that reduce errors. The short “white paper” would be excellent to shorten down my process to get started with Vituixcad. And make the dialogue here more focused. Just some thinking.
Last edited: