My first project is a 4 way speaker...

[TEK]

Hey!, we live in a world where everything is possible and changes so rapidly that maybe you can. I would be happy to make you a couple of boxes but you would have to pay also for the transportation .

Veneering would not be an easy task for sure, mainly because of the top and bottom rounded all around not to mention the size of the thing and that's why I dropped this option at the very start.

Actually my carpenter made some mistakes as I designed a 30mm chamfer on these tops and he cut a much smaller radius instead. He also forgot to chamfer some edges of the matrix which are visible on the 4th picture, supporting the baffle and made some holes in the wrong places :evil:

First: Holy shit, that some nice looking monsters :E:T
When it comes to the mistakes. First of all, any chamfering done inside the speaker is - let's be honest - only done for yourself and to impress the rest of the crowd here. No one visiting you will EVER take out your drivers and check if the inside edges are nicely chamfered :thanku:
When it comes to the extra holes. If that is supposed to be sealed - just add some silicone and your good. For your carpender I'll do this: ;x(
Then, chill out and do this: 8x) and this arty: and really just enjoy your design - it's looking great - so this goes to you to ;x(

[cochinada]

I like to think that chamfering inside the speaker has some effect as minor as it may be on the sound diffraction because avoiding sharp edges at 90º can only be a good thing. But you may be right and this be more a psychological issue than a practical one.

The main mistake with the wholes are the ones on the back panel where the cable will pass through. If they were at the same side like planned I could make the crossovers equal but this way I will have to make them symmetric.

Anyway thanks a lot for your words!

Here are two pictures of the terminal covers that I felt like take just for amusement.

Images not available

And another one:

Image not available

And finally a night perspective:

Image not available

[JonMarsh]

Very, very nice!! :T Two thumbs way up there.

Of course, this raises the question, what are you going to do for your SECOND project? :W :B

[TEK]

Jon, be nice now ;-)

cochinada: Could you tell a bit about what finish have applied on these? They seems to have a really deep clear, making me guess that there are several layers of some type of clear coat on them?
Also, that "black thing" (I assume it is the base), what material is it and what finish has been applied?

[cochinada]

Of course, this raises the question, what are you going to do for your SECOND project? :W :B

:B I guess I'll stop here. My house doesn't have space for anything more. Of course I wouldn't mind projecting for someone else but that's not very likely to happen with perhaps one exception. Anyway the most difficult and challenging stage is now about to come and I just hope I didn't bit more than I can chew. :scratchhead:

cochinada: Could you tell a bit about what finish have applied on these? They seems to have a really deep clear, making me guess that there are several layers of some type of clear coat on them?
Also, that "black thing" (I assume it is the base), what material is it and what finish has been applied?

Sure thing! I have to translate the terms to English so this will probably sound strange but here goes: first he used a product to cover the pores (?) and complained that the material was much more absorbing than he was expecting. Then he used glossy PU varnish of two components (polyurethane). Something like this. He told me that this varnish would be more suitable for plywood as opposed to the natural ones for massive wood. I don't think he polished them and I'm not sure if he used one coat or two but I can confirm very soon.

That "black thing" is indeed one of the brackets for the spikes. It is made of solid stainless steel of 10 mm and powder coated in RAL 9011 matt. Each weights 2.29 kg and 2.51 kg to be more precise. They were a little rough on the edges initially as some might remember but this was fixed and after the powder coated it came up fine so I'm quite pleased with the result. :T

Here you can have an overall view:

Image not available

[cochinada]

More problems... :sh:

The holes in the baffles have about 0,5 mm less than planned. As such the drivers don't fit unless I force them, which I won't do because to start with, I will have to take them out during the measuring stage.

Some manual sanding is inevitable... :rant:

[BobEllis]

That sounds like a nice tight fit before the finish was applied. That's always been an issue for me - how much room to allow for the thickness of the finish. Looks great!

To Jon's point, you do have other rooms where you'll want decent sound. Once you get into measuring you'll likely get the bug to build some small speakers for the bedroom...

[cochinada]

That sounds like a nice tight fit before the finish was applied. That's always been an issue for me - how much room to allow for the thickness of the finish. Looks great!

To Jon's point, you do have other rooms where you'll want decent sound. Once you get into measuring you'll likely get the bug to build some small speakers for the bedroom...

In this case the problem was different as there is no finishing on the inside of the holes and therefore no extra thickness. I already made this mistake in the pass when I did not account for 0.25 mm of paint thickness. Here it seems the CNC machine is making ellipses instead of circles! 8O

I live in a very small house with no extra rooms and my bedroom is so little that I could barely put my bed in it so...

Concerning the veneer he used two coats. This is confirmed now.

[BobEllis]

CNC out of round is really frustrating. At least it's not way out of tolerance.

Looking forward to seeing these assembled and watching your crossover development.

[deewan]

Again, amazing work. Congrats on getting the cabinets into your home safe and sound. I'll be watching closely your crossover development.

[cochinada]

I'm having second thoughts about the filling of the cabinets.

As some of you might remember I bought this Muhwolle/1000:

Image not available

Now there is one thing called gravity... In other words, unless I completely fill the cabinets, unavoidably some of this material will fall from the top. I'm not concerned so much about the alignment and respective Qa but more about the need or not to leave some sort of empty chamber behind the drivers so they can breathe. This would be almost impossible with the collapsing. What do you have to say about this need? Is this really necessary or not?
In case it is then I guess I must use some sort of foam like Sonofil for instance which has enough texture and size to stick to the top of the cabinets taking advantage of the inside matrix..

I hope I made myself clear enough... :roll:

[deewan]

My opinion is that the Muhwolle won't be a good option. You may get it to work in a few places but over time it was sag and fall out of place. That will leave some areas in your cabinets bare. If you are going to use that material I would suggest buying some leggings/pantyhose:


Place the Muhwolle along a inside panel of the cabinet and then staple the stretched out leggings over the Muhwolle to hold it in place and against the cabinet wall.

[cochinada]

My opinion is that the Muhwolle won't be a good option. You may get it to work in a few places but over time it was sag and fall out of place. That will leave some areas in your cabinets bare. If you are going to use that material I would suggest buying some leggings/panty hose. Place the Muhwolle along a inside panel of the cabinet and then staple the stretched out leggings over the Muhwolle to hold it in place and against the cabinet wall.

Yes. I'm afraid you're right, unless I completely fill the cabinets so there is no room for falling but this raises the question again of that chamber behind the woofers. Is it necessary? :huh:

I never thought about using panty hose in these speakers. They would surely mark some points in the sexy factor :alol:

By the way this Muhwolle comes in plastic bags. Would there be a problem to use it like this with the bag? This way it wouldn't fall but I wonder if plastic is a good idea... :sos:

[Paul W]

Or pillow cases.

[Lex]

Wow, that was some effort in these doing a quick scan at all those baffles and stuff! WOW

[deewan]

Yes. I'm afraid you're right, unless I completely fill the cabinets so there is no room for falling but this raises the question again of that chamber behind the woofers. Is it necessary? :huh:

I never thought about using panty hose in these speakers. They would surely mark some points in the sexy factor :alol:

By the way this Muhwolle comes in plastic bags. Would there be a problem to use it like this with the bag? This way it wouldn't fall but I wonder if plastic is a good idea... :sos:

I wouldn't suggest using the plastic bag.

The stuffing/Muhwolle isn't a requirement in the cabinets since the crossovers have not been created yet. But I am sure it will help with the internal reflections and tame some of the internal noises. Have you considered using compressed insulation or foam to line the insides of the cabinets?

http://www.acoustimac.com/owens-corning-703-1-6/

http://www.parts-express.com/acoustic-sound-damping-eggcrate-foam-california-prop-65-compliant-ul-94-1-1-2-x-12-x-12---260-510

[Paul W]

Or Muhwolle in $1 mesh laundry bags.

[Juhazi]

Looks like lining the inner walls is out of question. Stuffing is actually better. http://www.tnt-audio.com/clinica/spe...stuffed_e.html

[cochinada]

I wouldn't suggest using the plastic bag.

The stuffing/Muhwolle isn't a requirement in the cabinets since the crossovers have not been created yet. But I am sure it will help with the internal reflections and tame some of the internal noises. Have you considered using compressed insulation or foam to line the insides of the cabinets?

No. I designed this speakers precisely to avoid lining the inner walls. For that I have very uneven walls, not parallel between them and forming different angles or round corners. I'm not saying it would not be possible or it would not benefit from that but that's one thing I personally hate to do so I wanted to keep that at bay from the beginning. I was also influenced by this "Design mantra" at Zaph|Audio and I quote:

"For damping vented enclosures, I use Whispermat, Sonic Barrier or 1/2" carpet padding doubled up. Whispermat is the best value. Sonic Barrier is expensive but occasionally handy since it comes with adhesive and is sold in small (but expensive) quantities. Carpet padding is surprisingly effective and cheap but it varies in performance. I never use egg crate foam because it just plain does not perform well. For sealed, I feel that "Acousti-Stuff" is the best performing, but standard craft store dacron pillow fiber will do oK for about 1/3 the cost. "

So I thought that since the cabinets are all sealed I would be OK with stuffing alone.

Looks like lining the inner walls is out of question. Stuffing is actually better. http://www.tnt-audio.com/clinica/spe...stuffed_e.html

Thanks. I've read the whole article but how did you come to this conclusion? I got the idea that this is actually a lot more difficult as it depends on the density of the materials and where do you put them among other things. :-y

I don't want to complicate this too much and fortunately according to the article it seems that sealed boxes are not as critical as vented so since I already have the muhwolle I'm considering using it on the far recesses of the cabinets where it won't fall and at the same time, as I'm approaching the drivers to use something like Sonofil or any other foam that would serve to keep everything firm. So, a mixture of two materials and no need to use any kind of bag. To leave or not leave space behind the drivers that is question...

What do you think?

[Face]

I would lean towards denim insulation. It's much more effective at lower frequencies than polyfill type material. Just be sure not to overstuff.

[Juhazi]

So, a mixture of two materials and no need to use any kind of bag. To leave or not leave space behind the drivers that is question... What do you think?

:T You must leave some fre space behind the driver in a sealed cab too. Muhlwolle is quite dense, dacron/polyfill is lighter. My stategy would be to pack loosely muhlwolle to sides and back and polyfill in the center of the box. Measure and listen, do some experiments at least with midranges which are most sensitive to in-cabinet reflections.

[TEK]

How's it going? Are you lost in your speakers listening area, or did something come up?
Curious hear :-)

[cochinada]

How's it going? Are you lost in your speakers listening area, or did something come up?
Curious hear :-)

Things slowed down a bit mainly because I had to travel but now I'm going to pick up where I left. Actually cabling is almost done and today I glued some foam on the back of the 8" drivers which was a pain in the you know what because the access is terribly narrow. Next I will finish soldering the cables on the terminals and start filling the boxes with the Muhwolle, so hopefully everything will be set and ready to mount the drivers and start measuring.

As I will do that I will post the measurements here.

Image not available

Image not available

This will give you a better idea of the foam placement:

Image not available

I planned using 20mm thick but since I could only get 10mm I have two options: either I leave it this way or I glue a second layer on top of this one.

[cochinada]

I also had to struggle with the driver's openings like I told but this I was able to fix before I had to travel.

Before:

Image not available

After:

Image not available

It's not a perfect circle unfortunately but they fit.

[sdl2112]

Beautiful build. You are setting the bar with this one :T. I look forward to your measurements

[cochinada]

Beautiful build. You are setting the bar with this one :T. I look forward to your measurements

Thanks

This is how I plan to fill the boxes (muhwolle in red):

side view

Image not available

front view

Image not available

[TEK]

Glad to here from you again
Just glue two layers on top of each other if you want 20 mill. That should be no problem as far as I can tell. I did that in the mid-tones chambers of my Wavecord Ardents.

Are you going passive or active with these?

[cochinada]

Glad to here from you again
Just glue two layers on top of each other if you want 20 mill. That should be no problem as far as I can tell. I did that in the mid-tones chambers of my Wavecord Ardents.

Thanks! The only problem is the extra work. It was already gruesome and I hate to go over that again, :-y

I have an idea. To minimize the hassle how about covering just the portion immediately behind the drivers like this for the second layer?

Image not available

EDIT: this is not the damping I decided to use (see #216)

Are you going passive or active with these?

WWMTMWW passive and the subwoofers on the side will be active.

[cochinada]

Two boxes ready for drivers...

Image not available

Concerning the Muhwolle I was very pleased to see that it held quite well without need for pantyhoses or any other bizarre solutions. The subwoofer box is another story altogether but I'm leaving it for last.

I also glued two extra pieces of foam, one at the top and the other at the bottom as you can see in this picture:

Image not available

[cochinada]

From now on, for sake of clarity I will number the drivers starting from the bottom and so the WWMTMWW will have a W1, W2, W3 and W4 and an M1 and M2.

All 8" drivers are mounted and W1 and W3 are connected and ready to be measured one at a time. For connecting W2 and W4 I will have to remove both and switch the cabling between each pair. That's the purpose of the duct tape also know as man's best friend. By the way, I'm thinking about measuring ALL units but I wonder if it is really necessary... Any thoughts?

Image not available

As you can see the MTM box is already filled with the Angel Hair and the respective drivers are next. It's not very clear in the picture but there is actually a chamber like a nest behind each driver. This material is also much less dense than the Muhwolle.

Image not available

[cochinada]

I'm not very good with a soldering iron but I managed to weld all terminals and finally I'm ready to take some measures!

Sorry about the poor quality of the picture but it was taken counter light:

Image not available

[cochinada]

I have OmniMic as I already wrote and would like to use DATS as a signal generator because "DATS can provide a consistent high-quality signal source for testing".

The problem is how to set that up. Following the manual from OmniMic leads me nowhere.

It says to configure DATS as below:

Image not available

On the Audio Input Device Selection I have these options:
...
Line (USB AUDIO CONNECT) -> DATS V2?
Microphone (3 - OmniMic -> self explanatory

Image not available

and on the Audio Output Device Selection I have:
...
AVM 50v-1 -> my poweramplifier
Speakers (Realtek...) -> small pc speakers
Speakers (USB AUDIO CONNECT) -> DATS V2?

Image not available

If I follow the OmniMic's manual I get this horrible distorted sine wave:

Image not available

The only way I get nice signals is if I chose 'Line (USB AUDIO CONNECT)' as input and 'Speakers (USB AUDIO CONNECT)'


Image not available

But then nothing reaches my power amplifier and I have no sound

Moreover, if on OmniMic I use 'Play From Soundcard (on)' I only seem to have two options: 'pseudo noise' or 'sine sweep'

so, where are the different generator curves of DATS? How can I use them within OmniMic?

[cochinada]

Still didn't understand how OmniMic works together with DATS but nevertheless I tried to measure M1 and here is what I got:

Image not available

As you can see I have a problem in my room. I have placed the mic at 2m from the tweeter and I believe the first reflection comes at around 1.586msec as can be seen on the Impulse Response, which is too fast and doesn't allow me to accurately measure any frequency below 1.16kHz. Since the crossover point will be somewhere around 350Hz - 400 Hz this is not enough. :-y

... and if I'm doing this right the response from the tweeter (blue) is almost to good to be real :roll:

Image not available


Concerning W1, as can be seen from picture in Purchase options? it's practically impossible to measure as it is only about 28cm from the floor so the reflections are unavoidable. W3 however, gives the following curve:

Image not available

I guess that these curves will be more important for calculating the acoustic offset in the z-axis between the drivers than for crossover simulation, specially the W3 but also the M1.

Concerning W1 I'm going to face some unexpected challenges because of floor reflections. After all the crossover simulation program will not take these into effect and so the task ahead is getting even more difficult....

[cochinada]

Still playing around...

Option 'only to' selected
smoothing: 1/48th octave
mic fixed at 2m from tweeter center and height.

Tweeter:

Images not available

M2:


Images not available


W3:

Images not available


Will these curves be enough for calculating the acoustic centers or should I extend them first by using for instance option 'blended'?

[5th element]

http://www.mediafire.com/download/7dsikmz44lnf7br/measurement+guide.pdf

Might be of interest as the basics of this should be applied to pretty much any DIY loudspeaker in simulation packages.

[cochinada]

http://www.mediafire.com/download/7d...ment+guide.pdf

Might be of interest as the basics of this should be applied to pretty much any DIY loudspeaker in simulation packages.

You've already shared this on #158
I've read it already more than once and now I wonder if I can import these curves or is it better to extend them just for the purpose of finding the z coordinate...

[5th element]

You don't find the Z coordinate, with the measurement method outlined in the guide the Z coordinate is included within the relative phase measurements that you make. Correct phase alignment is then arrived at within the simulation software by comparing the softwares theoretical summation of the drivers with the real measured summation.

[cochinada]

Hi!

You don't find the Z coordinate, with the measurement method outlined in the guide the Z coordinate is included within the relative phase measurements that you make. Correct phase alignment is then arrived at within the simulation software by comparing the softwares theoretical summation of the drivers with the real measured summation.

I think I got it: you basically have to import both curves with phase measurements included, enter the x and y coordinates on the program, let it calculate the sum of both drivers in parallel and adjust z by trial and error in order to have a match with the measured curve for both at the same time.

I just have two questions:

a) in OmniMic I found how to define the tail of the gate but not the beginning. Is this possible? You mention in your tutorial that "One thing that is critical here is to remember where you set the start point
of the gate.", so is it all right if it starts at 0ms in all measurements?

b) You also mention "Lets assume that for whatever reason you do not need to measure the low frequency response of the driver, or you don't need to have the effects of baffle-step included in your measurements." and then you go and import your measurements and calculate the z as above. Now I will definitely need to incorporate the low frequency for the woofer and medium response as the crossover point between the two will be around 300-400Hz so must I measure the near-filed and join the curves before importing them or for just to calculate the z coordinate this is not necessary?


One more thing. The Near Field measurements, according to the rule of thumb Fmax = 4311 / radiating diameter (in inches) gives the following values:

22W-8534G00 => 629 Hz
12W-8524G00 => 1260 Hz

So, like I wrote before I have a problem as my Far Field graphs don't get even close to these limits. Unless there is some trick I can use, there will be an area of complete darkness where I can't measure the driver properly

Thanks!

[cochinada]

I found that OmniMic has a new feature for calculating the Z offset between two drivers! Somebody could have told me that... :roll:

Anyway the problem is that I can't find a good match between the calculated and the measure curve. The best I could achieve is this:

Image not available

Image not available

As you can see there is this strange delta in dB between the curves, otherwise I think the match would be very good.
In black is the measured curve with the tweeter in parallel with M2 and in blue is the dynamic added sum of the tweeter being measured alone with the previously measured M2. I say dynamic because the blue curve changes as I adjust the offset.
How do you explain this?

[5th element]

you basically have to import both curves with phase measurements included, enter the x and y coordinates on the program, let it calculate the sum of both drivers in parallel and adjust z by trial and error in order to have a match with the measured curve for both at the same time.

Yeah pretty much. The phase is all relative and you calibrate the system using a real measured combination so that the simulator matches up with the real world measurements.

a) in OmniMic I found how to define the tail of the gate but not the beginning. Is this possible? You mention in your tutorial that "One thing that is critical here is to remember where you set the start point
of the gate.", so is it all right if it starts at 0ms in all measurements?

I don't know. The only answer I can give you is that the group delay inherent within any hardware configuration should be identical between measurements performed. Some are not, such as some HDMI and receiver/DAC combos and these are unsuitable for performing measurements. If the group delay is constant then each measurement will have the same delay built in, this is what you want, the only thing you want affecting the arrival time of the measured impulse is the 'time of flight' path difference between the drivers that you are measuring.

The omni-mic itself is just a piece of hardware and should function just fine with any measurement software. I have absolutely no idea about the bundled software so cannot advise you in anyway, but it might be helpful if you try using something else? I would assume that the start point is the same for each measurement taken, but really you need to be able to trim the start of the impulse otherwise you're going to have to contend with ridiculous amounts of phase wrap.

b) You also mention "Lets assume that for whatever reason you do not need to measure the low frequency response of the driver, or you don't need to have the effects of baffle-step included in your measurements." and then you go and import your measurements and calculate the z as above. Now I will definitely need to incorporate the low frequency for the woofer and medium response as the crossover point between the two will be around 300-400Hz so must I measure the near-filed and join the curves before importing them or for just to calculate the z coordinate this is not necessary?

The high frequency, far field measurements are what contain the important phase data that are required for accurate crossover design. When you merge in the low frequency data you have to make sure to keep the phase data from the high frequencies as the reference and then you match the the near field/low frequency data to it.

One more thing. The Near Field measurements, according to the rule of thumb Fmax = 4311 / radiating diameter (in inches) gives the following values:

22W-8534G00 => 629 Hz
12W-8524G00 => 1260 Hz

So, like I wrote before I have a problem as my Far Field graphs don't get even close to these limits. Unless there is some trick I can use, there will be an area of complete darkness where I can't measure the driver properly

Thanks!

When measuring near field you can place the mic 1cm away from the cone and not have to worry. It's when measuring far field that you need to get the mic far enough away. The first reason for this is that you need (as you correctly mentioned) is because you need to get the microphone far enough away from the driver so that the sound radiating from its cone sums properly. The second is that you need to get the microphone far enough away so that you can measure proper summation from all of the drivers. With small loudspeakers this isn't ever really a problem. 50-80cm measurement distance is fine, but if your loudspeakers are 2 meters tall... well if they are symmetrical that simplifies things as for the high frequencies you can assume the drivers towards the top will measure the same as the drivers towards the bottom, so it can be assumed to be far less tall.

Usually one would recommend measuring the loudspeaker at the distance that it will typically be listened to but this creates problems. The further away you take the mic, the further away all nearby surfaces need to be to give you the same amount of reflection free data. I would highly recommend using the mic at around 80cm or so as this will allow you to open the gate up and get accurate data down far lower.

[cochinada]

I don't know. The only answer I can give you is that the group delay inherent within any hardware configuration should be identical between measurements performed. Some are not, such as some HDMI and receiver/DAC combos and these are unsuitable for performing measurements. If the group delay is constant then each measurement will have the same delay built in, this is what you want, the only thing you want affecting the arrival time of the measured impulse is the 'time of flight' path difference between the drivers that you are measuring.

Great... just what I needed right now. It so happens I'm using an HDMI connection from my video card as I don't have a sound card. .:scratchhead:
The path is: PC (HDMI NVIDIA GeForce GTX 750 Ti - NVIDIA High definition audio) -> Processor -> Power amplifier -> driver

...I would assume that the start point is the same for each measurement taken, but really you need to be able to trim the start of the impulse otherwise you're going to have to contend with ridiculous amounts of phase wrap.

After reading the manual I found there is a way to define the beginning of the gate, just by right clicking on the impulse response at the starting point. It seems that by default the 0ms corresponds to the higher peak and I always see the beginning of the graph at -1ms. Take a look at the impulse response and the frequency response:


Image not available

Image not available

...and if I right click and make the starting point somewhere between -1ms and 0ms (shortening the selected in red) I get this and a shorter valid frequency range: :roll:


Image not available

Image not available

What does this mean? I don't know... :stupid:

... but if your loudspeakers are 2 meters tall... well if they are symmetrical that simplifies things as for the high frequencies you can assume the drivers towards the top will measure the same as the drivers towards the bottom, so it can be assumed to be far less tall. ... I would highly recommend using the mic at around 80cm or so as this will allow you to open the gate up and get accurate data down far lower.

Indeed they are symmetrical :yesnod:. So in my case with these drivers and dimensions (in mm) exactly which ones would you recommend I measure? I already placed the mic at 1,5m instead of 2m keeping the height at the same level of tweeter (89cm). Isn't 80cm too close?

Image not available

[cochinada]

I have a very bizarre question for the experts:

I've downloaded ARTA because I noticed something very suspicious. It seems that any gate length gives much more frequency extension in this than in OmniMic and I simply don't know why. Is as if they are using different methods for FFT (?).

ARTA (gate 2.458msec):

Image not available

Image not available

OmniMic (gate 3.009msec):


Image not available

Image not available

So, as you can see even with a larger gate length, OmniMic Fmin is about 750Hz whereas ARTA is 400Hz!!!
Moreover the phase measurement is completely different!!!

Can anyone explain? 8O

[5th element]

I like ARTA because quite simply - what you see is what you get.

No doubt there are some settings in omnimic that will give the same results as ARTA but without using it I do not know what they are.

Where you position the start of the gate dramatically alters how the phase looks. The more silence there is before the impulse hits the more times the phase will wrap. If omnimic is automatically adjusting things so that the start corresponds with where the impulse is actually measured then it's this kind of compensation that screws up the phase relationship between drivers that is crucial for proper simulation to take place.

The lower end frequency limit is determined by how many cycles can pass, at a specific frequency, for the length of the gate. The lower the number of cycles the lower the accuracy becomes. I am sure there is some statistic about how accurate a measurement with one one cycle possible actually is, but there is a reason for response averaging. In ARTA you can customise it to take any number of measurements and then it will average them out for the best accuracy. This really helps with removing noise from the measurements too. For what it's worth I haven't had any problems with using ARTAs lower frequency limit.

I seem to remember the HDMI system that someone was using had obvious issues with regards to the start of the impulse response. If you use ARTA and make a measurement, presuming the mic and loudspeakers are fixed in place, then every single impulse response measured should look identical in time. This means that within ARTA the time along the bottom, at which the impulse first starts, should be identical with each measurement made. So if it starts at 2.188ms in one measurement, it should start at 2.188ms in the next. If this is the case then your hardware will be fine. In some systems the HDMI receiver (or maybe transmitter?) does some buffering and/or locking on of the HDMI signal (or so it seemed) so the amount of time the hardware took to actually output sound changed with every single measurement. Now when you're trying to determine the relative phase of the drive units by the difference between their time of flight to the mic, having the hardware adding in random delays to this kind of made the results meaningless.

So yes, perform some repeat measurements with ARTA and check that the impulse start time remains exactly the same with each measurement. If it does, then no problem!

[cochinada]

Now I understand why you are the "Supreme Being Moderator" ;x(

I have good news! I made some consecutive measurements with ARTA (more than 5) and the impulse started always at the same time so at least it seems I don't have that problem with HDMI.

Image not available

Image not available

What do you think about this gate?

The problem is that I have a demo version that doesn't allow me to safe the curves

[5th element]

That looks absolutely fine and it's great that your hardware is compatible with making measurements!

One does not need to save the curves with ARTA what you need to do is export them as an frd file or similar for use with some simulation software. Saving simply saves the impulse itself so that you can load it up again within ARTA at a different time. The files you save are only usable by ARTA too, which isn't any use for other programs. Of course having the ability to save the impulse files is very useful but isn't necessary for most things.