Jump to content

Recommended Posts

Posted (edited)

Latest set of measurements, hot off the press. Taken with Acourate with my new set of filters. IACC was 89% in a single listening position. 

 

aris-amplitude.thumb.jpg.e5e6f6d893df55f3df83570f34d3312f.jpg

 

Frequency response +/- 3dB following the Harman Synthesis target curve (thanks Aris). Harman Synthesis curve is 0dB at 20Hz, with a straight flat line down to -10dB at 20kHz. Notable dip at the crossover point at about 500Hz which I have not yet diagnosed. 

 

aris-groupdelay.thumb.jpg.f859e05de662e3b43c3d80edf67ba593.jpg'

Significant amount of group delay in the bass which can potentially be improved in the future with refinements to my bass time alignment technique. Right now I am using a rather coarse method (trying to spot the frequency peak of the subwoofer by naked eye is very difficult!!!). Uli himself describes a method by sine wave convolution which I have not yet tried. 

 

aris-phase.thumb.jpg.8399927cb691adb3ce6d60379a2f0e9c.jpg

 

Nice flat response, except for the bass. 

 

aris-step.thumb.jpg.ba5172be1b9cd678e3de4c0c7b22745e.jpg

 

And a nice looking step response which to my eyes looks ideal. 

 

aris-reverb.thumb.jpg.b004b18d4be46781ae99675dfc5d8495.jpg

 

Nice fast decay with minimal reverb. 

 

So it should sound good, right? 

 

Unfortunately not. To my ears there is still a frequency peak around 3-4000 Hz. At least I think it's around here. I can hear it, but I can't see it in my measurements - and it's driving me bonkers. I'm not the only one who hears it as well, a friend came over and he could hear it too. 

Edited by Keith_W

Share this post


Link to post
Share on other sites
4 hours ago, Keith_W said:

Latest set of measurements, hot off the press. Taken with Acourate with my new set of filters. IACC was 89% in a single listening position. 

 

aris-amplitude.thumb.jpg.e5e6f6d893df55f3df83570f34d3312f.jpg

 

Frequency response +/- 3dB following the Harman Synthesis target curve (thanks Aris). Harman Synthesis curve is 0dB at 20Hz, with a straight flat line down to -10dB at 20kHz. Notable dip at the crossover point at about 500Hz which I have not yet diagnosed. 

 

aris-groupdelay.thumb.jpg.f859e05de662e3b43c3d80edf67ba593.jpg'

Significant amount of group delay in the bass which can potentially be improved in the future with refinements to my bass time alignment technique. Right now I am using a rather coarse method (trying to spot the frequency peak of the subwoofer by naked eye is very difficult!!!). Uli himself describes a method by sine wave convolution which I have not yet tried. 

 

aris-phase.thumb.jpg.8399927cb691adb3ce6d60379a2f0e9c.jpg

 

Nice flat response, except for the bass. 

 

aris-step.thumb.jpg.ba5172be1b9cd678e3de4c0c7b22745e.jpg

 

And a nice looking step response which to my eyes looks ideal. 

 

aris-reverb.thumb.jpg.b004b18d4be46781ae99675dfc5d8495.jpg

 

Nice fast decay with minimal reverb. 

 

So it should sound good, right? 

 

Unfortunately not. To my ears there is still a frequency peak around 3-4000 Hz. At least I think it's around here. I can hear it, but I can't see it in my measurements - and it's driving me bonkers. I'm not the only one who hears it as well, a friend came over and he could hear it too. 

Great work, Keith!  :thumb:

 

Andy

 

Share this post


Link to post
Share on other sites
Latest set of measurements, hot off the press. Taken with Acourate with my new set of filters. IACC was 89% in a single listening position. 
 
aris-amplitude.thumb.jpg.e5e6f6d893df55f3df83570f34d3312f.jpg
 
Frequency response +/- 3dB following the Harman Synthesis target curve (thanks Aris). Harman Synthesis curve is 0dB at 20Hz, with a straight flat line down to -10dB at 20kHz. Notable dip at the crossover point at about 500Hz which I have not yet diagnosed. 
 
aris-groupdelay.thumb.jpg.f859e05de662e3b43c3d80edf67ba593.jpg'
Significant amount of group delay in the bass which can potentially be improved in the future with refinements to my bass time alignment technique. Right now I am using a rather coarse method (trying to spot the frequency peak of the subwoofer by naked eye is very difficult!!!). Uli himself describes a method by sine wave convolution which I have not yet tried. 
 
aris-phase.thumb.jpg.8399927cb691adb3ce6d60379a2f0e9c.jpg
 
Nice flat response, except for the bass. 
 
aris-step.thumb.jpg.ba5172be1b9cd678e3de4c0c7b22745e.jpg
 
And a nice looking step response which to my eyes looks ideal. 
 
aris-reverb.thumb.jpg.b004b18d4be46781ae99675dfc5d8495.jpg
 
Nice fast decay with minimal reverb. 
 
So it should sound good, right? 
 
Unfortunately not. To my ears there is still a frequency peak around 3-4000 Hz. At least I think it's around here. I can hear it, but I can't see it in my measurements - and it's driving me bonkers. I'm not the only one who hears it as well, a friend came over and he could hear it too. 

Does the frequency peak correspond to a crossover point? And, if so, then does the effect go away if you move the crossover out of the 3-4K range?

Share this post


Link to post
Share on other sites

It seems to persist no matter what I do. I have asked Uli and some others on the Acourate forum and they have given me leads to chase up. Unfortunately I am a little bit too drunk tonight to do anything. Maybe tomorrow. 

Share this post


Link to post
Share on other sites
11 minutes ago, Keith_W said:

Unfortunately I am a little bit too drunk tonight to do anything.

Nothing unfortunate about that...

Just one way to say...enough for today...

Share this post


Link to post
Share on other sites
On 13/04/2018 at 11:42 PM, Keith_W said:

Unfortunately not. To my ears there is still a frequency peak around 3-4000 Hz. At least I think it's around here.

The problems in the octaves below could be making that region sound 'wrong'.

 

Eg.  If you turn off the bass, it sounds treble heavy.

If you have cancellations happening in the 500-1000-2000 octaves, it can make 2000-4000 stand out.

 

You don't appear to be achieving the overall target of -10dB @ 20khz   (you look like you're getting 4dB drop from LF to HF)  ....   If you did, those octaves would be a little bit lower in SPL too.

Share this post


Link to post
Share on other sites


Posted (edited)

It's hard to be definitive without more data.... but remember the chart posted on 30 March ?!

 

When measuring the horn near field, there is a boost in the low end (500 to 1000) as expected.    That boost isn't present further way from the horn ..... so this could be (part of) what is responsible for what you're seeing in the 500 to 1000 ish range.

Edited by davewantsmoore

Share this post


Link to post
Share on other sites

Just FYI Dave. 

 

horn-before.thumb.jpg.5d31ed18ecd1bb9c4a04995c2caa1aa5.jpg

 

This is a curve taken some years ago with the DEQX. 1m from the horn. The difference is that the passive crossover network in the horn is in place

 

5ad3519a30a3a_horn-after(1).thumb.jpg.138d514417115b8e5bd4db9199b26372.jpg

 

These are the native frequency response curves of each driver. Yes, the top end of the woofer (in red) is really jagged, but that's because it has been pushed way beyond its breakup frequency. Pay attention to the green curve - this is the midrange horn, but with the passive crossover bypassed

 

It appears that the passive crossover network applies a -6dB/oct high pass to flatten out the response of the horn. Yes, I know the scales are different so you have to read it carefully. 

 

horn-corrected.thumb.jpg.8b047f767798fffbcc3f0a683f49fb96.jpg

 

The green curve is the correction filter for the horn which essentially reverses the slope. In red we have the measured corrected frequency response for the horn. 

 

I need to do a better job trying to get the system to conform to the target curve. I also need to do some distortion measurements in Acourate. I wasn't sure how to do it, but I have since been pointed in the right direction by the helpful members of Acourate forum. 

 

I'm really loving Acourate. I have made no hardware changes for a year, but I have been making incremental refinements in the way I use the software. All I need to do is load up an old set of filters from a year ago and compare what I can do now, compared to back then. The difference is pretty astounding. 

 

I'll just need to keep chipping away at it. 

Share this post


Link to post
Share on other sites
On 26/03/2018 at 4:02 PM, roh008 said:

 

I meant something like REW built in (as part of the package). It can load and apply a file created in another software package.

Yes I believe so, in the Convolution section of DSP Studio.

Share this post


Link to post
Share on other sites

By the way, I came across some interesting resources for those who are interested in DSP. 

 

The first is Floyd Toole's 2015 paper on what is and is not possible with DSP. It is quite eye opening: 

http://www.aes.org/journal/online/JAES_V63/7_8/#paper1

 

The second is JJ Johnston's presentation in 2008 about acoustic and psychoacoustic issues in room correction. Taking a measurement will give you lots of data, but what should you aim to correct? Obviously, the first thing you need to correct is for what you can perceive, but what can you perceive? Traditionally, research into psychoacoustics has been lacking. This presentation goes into some detail about what is audible, what isn't, and what contributes to spatial cues, and a whole bunch of other things. 

 

http://www.aes-media.org/sections/pnw/pnwrecaps/2008/jj_jan08/

 

When I first went down the DSP line, I had pretty simple aims. All I wanted to do was driver correction, alignment, and room correction. Nowadays I am overwhelmed with information such as what is contained in these two papers which go above and beyond what is possible in most "normal" systems. 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.

×