Triangle Titus EZ spinorama measurements (CTA-2034)

[kimmosto]

@Ageve could you save on-axis response with phase angle to zip file, and publish phase response and time domain including impulse response, step response and ETC. RMS of Step response could also be revealing with some speakers. With that content you are permitted to use VituixCAD for processing data and publishing charts on ASR. Otherwise using VituixCAD for is not permitted on this forum (including DIY projects). Thanks for your co-operation.

 

[Toni Mas]

Otherwise using VituixCAD for is not permitted on this forum (including DIY projects).

Never heard of this... Just for information, is this specific to this forum, or is this general?

 

[Ageve]

@Ageve could you save on-axis response with phase angle to zip file, and publish phase response and time domain including impulse response, step response and ETC. RMS of Step response could also be revealing with some speakers. With that content you are permitted to use VituixCAD for processing data and publishing charts on ASR. Otherwise using VituixCAD for is not permitted on this forum (including DIY projects). Thanks for your co-operation.

I have attached on-axis response+phase angle, and pictures of impulse response, step response and ETC.

 

[Azrael]

Never heard of this... Just for information, is this specific to this forum, or is this general?

I would like to add my voice to this question: there are screenshots in many DIY speaker forums where VituixCAD has been used. Was this possibly illegal? Did they not read the terms and conditions carefully enough?

*edit*:
Ok, it seems to be ASR-specific. Quote from the licence.txt file:

You may not publish any material produced or processed with VituixCAD on Audio Science Review (ASR) Forum.

Is there any history to this that I don't know about?

 

[Toni Mas]

I would like to add my voice to this question: there are screenshots in many DIY speaker forums where VituixCAD has been used. Was this possibly illegal? Did they not read the terms and conditions carefully enough?

*edit*:
Ok, it seems to be ASR-specific. Quote from the licence.txt file:


Is there any history to this that I don't know about?

Ok It must be a recent update

 

[Toni Mas]

I have attached on-axis response+phase angle, and pictures of impulse response, step response and ETC.

Mmmm... I don't understand what's going on at about 5ms ... A reflection? Isn't it a gated measurement to avoid this?

 

[RickS]

@Ageve could you save on-axis response with phase angle to zip file, and publish phase response and time domain including impulse response, step response and ETC. RMS of Step response could also be revealing with some speakers. With that content you are permitted to use VituixCAD for processing data and publishing charts on ASR. Otherwise using VituixCAD for is not permitted on this forum (including DIY projects). Thanks for your co-operation.

Sent you a direct message and hope we can find a better way to move forward.

 

[Ageve]

Mmmm... I don't understand what's going on at about 5ms ... A reflection? Isn't it a gated measurement to avoid this? :rolleyes

It’s the first reflection. You gate the measurement by limiting the ”window”, in this case 4.5ms. If you see gated measurements with ripples, then the window is too wide. I did that mistake with some of my first measurements (Dali Zensor 1 and a few others). The overall response will still be accurate though, as long as you don’t go too far with the window width.

5ms is usually fine in my garage (3m ceiling), but I had to go a bit lower with this one for some reason. By doing so you loose resolution at lower frequencies (close to the merging point).

 

[Toni Mas]

It’s the first reflection. You gate the measurement by limiting the ”window”, in this case 4.5ms. If you see gated measurements with ripples, then the window is too

5ms is usuallybut I had to go a bit lower with this one for some reason. By doing so you loose resolution at lower frequencies

Imho, this is due to insisting in making the measurements at 1m indoor. In my experience it systematically makes the response flawed betwen 500hz and 1khz.
To make sure and avoid this i would measure closer, 50cm... Even if It might be a problem with other regards...

 

[Ageve]

To make sure and avoid this i would measure closer, 50cm... Even if It might be a problem with other regards...

Frequency response won't be accurate at just 50 cm, well maybe for very small speakers, but still. Losing resolution at the merging point is less of a problem than getting an inaccurate response.

I measured at 2m as well (attached to the first post):

This is the quasi-anechoic response of both. Similar, but lower resolution at 2m of couse:

Sent you a direct message and hope we can find a better way to move forward.

Thanks.

I didn't know about these requirements. If the information provided above is adequate (or maybe a mdat-file from REW as well), I'm fine with posting that.

 

[Toni Mas]

Frequency response won't be accurate at just 50 cm, well maybe for very small speakers, but still. Losing resolution at the merging point is less of a problem than getting an inaccurate response.

It all depends on what you call an accurate/inaccurate responses. Applying tons of smoothing maybe acceptable to get a global image though loosing resolution, but detecting and fully understanding local problems in isolation also matters.
It is clear for me that there IS a problem with these measurements around 1khz, but It is still unclear to me what the problem IS, measurement artifact or whatever.
I also think that when such a big issue IS detected, it should be investigated in isolation, and before panthers begin to make them heard...

 

[Ageve]

It is clear for me that there IS a problem with these measurements around 1khz, but It is still unclear to me what the problem IS, measurement artifact or whatever.
I also think that when such a big issue IS detected, it should be investigated in isolation, and before panthers begin to make them heard...

I'm pretty sure it's not a measurement artifact, for a number of reasons:

1. Measurement by Audio.com.pl looks the same.
2. It's present at both 1m and 2m distance.
3. It's also present when doing an in-room measurement at 4m distance.
4. It's audible.
5. I have compared measurements I've done to others, like this one:

As for outdoor measurements:

It's been very hot and humid here today, but I did a sweep of my Infinity just an hour ago, in the dark.

This is the impulse response:

The first major reflection (floor) is at 8ms, but there's noise > 6ms that will cause ripples in the gated frequency response.

This is the difference between my outdoor measurement (6ms window) and in my garage (4.7ms in that case):

 

[kimmosto]

I have attached on-axis response+phase angle, and pictures of impulse response, step response and ETC.

Thank you!
Few references how you could plot time domain in the future using VituixCAD only:

Excess group delay should be adjusted to 0 ms at the top octave with Tg parameter of the generator. In this case +51 us.

 

[kimmosto]

Here are some measurements of the Triangle Esprit Titus EZ bookshelf speaker.

Tip 1. VituixCAD is able to show contour lines if you like to imitate steps of 3 dB in Klippel's polarmap.

 

[kimmosto]

@Ageve Tip 2:
REW or any other tool than VituixCAD is not mandatory for plotting CTA-2034 and directivity. Those two charts have required features with fast access.
For example
CTA-2034: Basic + Predicted In-room

CTA-2034: Basic + Predicted In-room + Slope target zones + Data labels

CTA-2034: Early Reflections

CTA-2034: Early Reflections + Reflection Totals

 

[Toni Mas]

I'm pretty sure it's not a measurement artifact, for a number of reasons:

There is still something unclear for me because of the huge discrepancy between the semianechoic curves published and the nearfield one (cone only). First i am curious to see the result of summing the port response to the woofer's one. Second, i you don't gate your measurements properly, the results won't be reliable.

Btw, i think that when reviewing a product, the reviewer should load not only the CTA curves that contain heavy processing of the various measurements they are based on, but also the impulse responses of these ones so that anybody might make a clearer idea of what the reviewer is showing... and also what is not showing.

 

[Ageve]

There is still something unclear for me because of the huge discrepancy between the semianechoic curves published and the nearfield one (cone only). First i am curious to see the result of summing the port response to the woofer's one. Second, i you don't gate your measurements properly, the results won't be reliable.

I think you're overanalyzing the nearfield measurements here. Issues can be seen at 800-1000Hz, but there's no guarantee it will look identical in far-field. It would be stranger if the woofer response was smooth in near-field.

This is the summed and corrected bass response + the gated part. Green line = merging frequency.

You can see a bump in near-field as well.

This is the uncorrected port+woofer sum, compared to measurements by Audio.com.pl:

The most important thing for me (when it comes to the 900 Hz peak), was the in-room response at 4m distance. If there was something wrong with the gated measurements, you wouldn't get such a close match between the real and estimated in-room response.

I actually did a few more in-room measurements (not included in the average here, but I saved them if you want to see them), but I could never get rid of the peak.

< 500Hz can be ignored as usual (room).

 

[billou5767]

the question : the titus are the same since the 1ST lunch ?

 

[RickS]

Did you try to seal off the port and see how the speaker measures?

 

[Ageve]

Did you try to seal off the port and see how the speaker measures?

No, unfortunately I didn’t have time to try it. I only had the speaker for a couple of days.

The larger Comete 40th has a similar peak btw, but centered at around 700 Hz (larger woofer, dual ports).

Distortion peaks at 1-2 kHz also look similar.

(Different tweeter in the 40th anniversary models, seems to have issues at 3-5 kHz)

SoundStageNetwork.com | SoundStage.com - NRC Measurements: Triangle 40th Anniversary Comète Loudspeakers

www.soundstagenetwork.com