Thank you very much keantoken
Sorry for the late replay
I really appreciate your efforts and the time you spend solving the problem.
For me is still unclear how I've got those reflections at 250 and 500ms but the mic doesn't lie. I'm not blaming the manufacturer for any issues. I suppose that is my fault.
Probably the baffle cutout for the tweeter is still too tight as I had to expand it a bit after the varnishing which may couses some twist or preasure over the mounting plate.
I tend to rework the baffles (and the XO obviously) and I gonna use waveguide for the tweeter. My livingroom is not too big = quite reflective and using the waveguide would be very useful for achieving a better directivity.
Greetings
Sorry for the late replay
I really appreciate your efforts and the time you spend solving the problem.
For me is still unclear how I've got those reflections at 250 and 500ms but the mic doesn't lie. I'm not blaming the manufacturer for any issues. I suppose that is my fault.
Probably the baffle cutout for the tweeter is still too tight as I had to expand it a bit after the varnishing which may couses some twist or preasure over the mounting plate.
I tend to rework the baffles (and the XO obviously) and I gonna use waveguide for the tweeter. My livingroom is not too big = quite reflective and using the waveguide would be very useful for achieving a better directivity.
Greetings
I'm using small stand which comes along with UMIK-1 fitted on the edge of old ironing board. The mic itself is about 30 - 40cm out of the board.
You don’t have to remove the tweeter from the cabinet. Just move the mic right upto the tweeter. As close as you can without touching it (risk damage to dome or mic)
Nearfield measurement of the tweeter? What's the purpose of such a measurement?
Sorry I'm confused but I'll do it probably tomorrow evening.
Sorry I'm confused but I'll do it probably tomorrow evening.
You measure just the tweeter FR and it's useful to see, by comparison with the farfield measurement, if there are problems related to the baffle.
That stand certainly isn’t reflection free. If you have acoustic foam or so at hand, cover the stand with a layer of 3cm. Better would be a mount on the end of a tube (I fitted an XLR in/on a 4’ rod that I clamp on a photo tripod).
I didn't think that such a thin stand could affect the measurements.
I have on hand some pieces of 1" Dacron and definitely gonna use two layers of it next time.
I find what cause the "ringing". It comes from way you display waterfall. If you choose Burst Decay mode then usually time axis is not time but periods of sinus signal at particular frequency.
From REW help: The Burst Decay mode provides a way to more easily distinguish resonances of similar Q but different frequencies. It does this by showing the way a shaped tone burst at each frequency would decay, but along an axis denoted in periods of the frequency rather than time. On a period axis the extent of a decay is the same for resonances of the same Q regardless of the frequency of the resonance.
If you select Fourier mode, you get standard waterfall what is also way how producers and others display the results. There is no ringing until level where noise will be visible.
From REW help: The Burst Decay mode provides a way to more easily distinguish resonances of similar Q but different frequencies. It does this by showing the way a shaped tone burst at each frequency would decay, but along an axis denoted in periods of the frequency rather than time. On a period axis the extent of a decay is the same for resonances of the same Q regardless of the frequency of the resonance.
If you select Fourier mode, you get standard waterfall what is also way how producers and others display the results. There is no ringing until level where noise will be visible.
There are still 3 peaks/dips on graph from 10 to 20kHz and the first two can and will be heard. There is smth wrong with sound reproduction hardware or measurement setup hardware/software. I checked my measurements - there are no such peaks on any of the worst drivers.
I just checked
I changed the view from Burst Decay to Gaussian
Basically spectrogram of tweeter w/o XO still have some 'fires" above 10kHz (as svp noted) probably some errors due equipment or noise.
Then I checked the measurement of the finished speaker with crossover
Ta-Ra!
All the ringing disappear!
Thank you very mach for the advice! I never thought there was a difference between the two views and that one shows the spectrogram in cycles and the other in time.
Looks like I'll need much more practice with REW.
Great news that you started to master REW.
I wanted to emphasize that this is not how the SPL curve should look of pretty good and expensive tweeter:
The dips/peaks are +-5dB, or more, or 10dB overall. It is not how should plot a good tweeter.
Here is quick and dirty measurement of some Mackie studio monitor (cheap) with non-prepared setup
Here is measurement of Sica coax tweeter only - which has much worse tweeter than yours:
and so on...
I wanted to emphasize that this is not how the SPL curve should look of pretty good and expensive tweeter:
The dips/peaks are +-5dB, or more, or 10dB overall. It is not how should plot a good tweeter.
Here is quick and dirty measurement of some Mackie studio monitor (cheap) with non-prepared setup
Here is measurement of Sica coax tweeter only - which has much worse tweeter than yours:
and so on...
Sorry for the delay, but I had a busy week
Last weekend I prepared a new baffle with the same size like the current one but without the step and I did some measurements.
First I removed the current baffle from the box and fitted it on a stand firmly.
The first measurement was farefield 1m
After that I placed the tweeter on the new stepless baffle keeping the same distances from the edges.
On the third measurement I moved the tweeter off center.
And finally I put the tweeter back on its place and moved the mic 2mm close from the phaseplug
In the frequency range above 7kHz the baffle doesn't have significant impact over the tweeter's behaviour. The tweeter's ringing isn't linked to the baffle step or the tweeters position.. The possibly of defective driver is rather insignificant, as the both units acting the same way despite they was bougth at different times.
I have doubts about the quality of the used equipment as some of the guys assumed. Probably the amplifier or the sound card are producing some hissing noise as the same ringing appears within woofer's waterfall at the frequencies over 13kHz.
Last weekend I prepared a new baffle with the same size like the current one but without the step and I did some measurements.
First I removed the current baffle from the box and fitted it on a stand firmly.
The first measurement was farefield 1m
After that I placed the tweeter on the new stepless baffle keeping the same distances from the edges.
On the third measurement I moved the tweeter off center.
And finally I put the tweeter back on its place and moved the mic 2mm close from the phaseplug
In the frequency range above 7kHz the baffle doesn't have significant impact over the tweeter's behaviour. The tweeter's ringing isn't linked to the baffle step or the tweeters position.. The possibly of defective driver is rather insignificant, as the both units acting the same way despite they was bougth at different times.
I have doubts about the quality of the used equipment as some of the guys assumed. Probably the amplifier or the sound card are producing some hissing noise as the same ringing appears within woofer's waterfall at the frequencies over 13kHz.
Attachments
What your ears say? If the sound is good - probably it is good.
For me at 2mm the drop of 10dB from 3kHz to 15kHz is not how it should be.
Yes, it could be hissing of some component, but that component can be analog part of inside USB mic also.
Or it can be measurement procedure. For analog mic I use this one, as most of us here: https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_Measurement_REW.pdf
You can start by changing components one by one:
Soundcard - one of the best ones for any amount of money: Apple USB-C to 3.5mm dongle. 10-15EUR. Another option - Asus Xonar U3 USB, up to 30EUR.
Amp: That Denon can have some coloration in it. Maybe you are using some phono correction or some "on-by-default" EQ by accident? Maybe try another decent Class D amp with no built-in EQ. TDA 325X is a now king . With PFFB - even better. TDA 311X can be an option. I am using both. Would not recommend anything bellow 60-70EUR. For 100-150EUR you can have some box with PFFB.
For me at 2mm the drop of 10dB from 3kHz to 15kHz is not how it should be.
Yes, it could be hissing of some component, but that component can be analog part of inside USB mic also.
Or it can be measurement procedure. For analog mic I use this one, as most of us here: https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_Measurement_REW.pdf
You can start by changing components one by one:
Soundcard - one of the best ones for any amount of money: Apple USB-C to 3.5mm dongle. 10-15EUR. Another option - Asus Xonar U3 USB, up to 30EUR.
Amp: That Denon can have some coloration in it. Maybe you are using some phono correction or some "on-by-default" EQ by accident? Maybe try another decent Class D amp with no built-in EQ. TDA 325X is a now king . With PFFB - even better. TDA 311X can be an option. I am using both. Would not recommend anything bellow 60-70EUR. For 100-150EUR you can have some box with PFFB.
Thank you for the replay svp.
In my ears and in ears of those who have heard them, the speakers sound fantastic, very opened and detailed.
The problems appear in very low levels, so they are rather inaudible.
For the measurement I follow this guide
https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/
Well I definitely have to invest some money in better and more reputable equipment. At the beginning I'll need some Focusrite 2i2 (2nd or 3rd gen can be found for as much as £60), some XLR mic and of course some better amplifier.
In my ears and in ears of those who have heard them, the speakers sound fantastic, very opened and detailed.
The problems appear in very low levels, so they are rather inaudible.
For the measurement I follow this guide
https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/
Well I definitely have to invest some money in better and more reputable equipment. At the beginning I'll need some Focusrite 2i2 (2nd or 3rd gen can be found for as much as £60), some XLR mic and of course some better amplifier.
Well... the OP already has a 200+ EUR amp, my idea is to have another transparent ready to play out of the box "backup" amp for least amount of money.
Can you link that thread, i am interested too.
If everyone say your speakers are fantastic - probably they are fantastic. Mission: success!
Measurement at very low levels can be the cause of such artefacts. I do not have very accurate SPL meters, but for my ears from 70 dB to 80-85dB levels show pretty much everything speaker has to offer.
I read also that ASR thread - it is more for measurement finished products, that one manual from Kimmo is for DIY designer (as you are) and is much more concentrated, and for measurement of separate drivers, before you build crossover for them.
You measured drivers with crossover, right? When measuring tweeter, what you did with woofer and vent? Covered it with pillow or smth?
I do not say, that current one you have is a "bad" one - you need just a backup one for test. Borrow from friend another one to check sound subjectively and/or compare measurements