or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › CEDIA 2010 and Distortion
New Posts  All Forums:Forum Nav:

CEDIA 2010 and Distortion - Page 3

post #61 of 68
Depending on how good your maths is, have a read of Keith Howard's article here which explains how AddDistortion works and its limitations.

I have spent an altogether unhealthy amount of time investigating distortion measurement in my spare time so I thought I would share the benefits of any insight that I have gained along the way. The Aalborg University thesis paper you linked to earlier is a good read and replicating some of that work has been quite educational.

Like you I am somewhat puzzled that this subject doesn't provoke more interest but it seems to me that folks see a harmonic distortion spectrum plotted for a single tone and think that tells them all they need to know about how that system would behave with a music signal. The reason I introduced the IMD point was to try and make clearer that IMD is a natural consequence of non-linearity just like HD and the two tend to go hand in hand. Anyway, I for one will be interested to see where your efforts take you.
post #62 of 68
Thanks for the link, it will really help out.

I understand IMD ("I am aware of," might be a better way to put it. .) I also now know why you brought it up at that point. To be honest, at first I thought you were trying to be, shall we say "non-helpful." I thought it might be an attempt at muddying the waters so to speak. A lot of that goes on around here ya know.



On another note, your name seems very familiar. Seems like I remember it from a few years ago, on a European forum maybe?
post #63 of 68
Quote:
Originally Posted by soho54 View Post

On another note, your name seems very familiar. Seems like I remember it from a few years ago, on a European forum maybe?

Very possibly, though I tend to be fairly selective on where I post these days. My chief claim to fame is being behind the AVTalk subwoofer tests but I haven't had time to do any in the last 12 months due to other commitments.
post #64 of 68
Quote:
Originally Posted by cjwhitehouse View Post

My statement was indeed a generalisation and it is always possible to cite exceptions, but as a rule, most subwoofer distortion behaviour is dominated by the excursion issue down low and this is what soho54 was trying to emulate.

I thought the idea was to try and prove/disprove whether the CEA2010 distortion thresholds are a good measure of distortion audibility and I seem to recall that it was your request to do the test with a distorted music signal rather than sine tones.

I was just trying to generally comment that excursion isn't the only form of distortion in a subwoofer, and the reason I asked about the goals is to suggest that a more complicated model is necessary if you really want to capture the kinds of distortions that a subwoofer makes. However, Klippel already has a nonlinear speaker model that they use for their auralization, which is the direction I see all this going. The problem then becomes the specific nonlinear model chosen won't be representative of the wide range of subwoofer choices available.

For instance, I find that inductance modulation is more noticeable than CMS and BL nonlinearities. I believe some of the Klippel auralization testing shows that to be true too. I think it kinda makes sense though because the changing inductance creates a frequency shifting component, and it happens at higher frequencies. Excursion nonlinearities behave more like a soft-knee limiter at the lower frequencies.
post #65 of 68
Quote:
Originally Posted by MBentz View Post

I was just trying to generally comment that excursion isn't the only form of distortion in a subwoofer, and the reason I asked about the goals is to suggest that a more complicated model is necessary if you really want to capture the kinds of distortions that a subwoofer makes. However, Klippel already has a nonlinear speaker model that they use for their auralization, which is the direction I see all this going. The problem then becomes the specific nonlinear model chosen won't be representative of the wide range of subwoofer choices available.

For instance, I find that inductance modulation is more noticeable than CMS and BL nonlinearities. I believe some of the Klippel auralization testing shows that to be true too. I think it kinda makes sense though because the changing inductance creates a frequency shifting component, and it happens at higher frequencies. Excursion nonlinearities behave more like a soft-knee limiter at the lower frequencies.

If you insist on 100% accuracy and happen to have access to a working copy of the Klippel model and a suitable set of typical state variables, I would be happy to use it here. But in the absence of that we are left with trying to do a 90% job using relatively simple tools that are available.

Can a test based on single tones (or shaped tonebursts) fully assess the large signal behaviour of a subwoofer? Clearly not, as it fails to excite all relevant non-linear behaviour. As a minimum, Klippel would insist that testing with a fixed and moving two-tone test signal is required.

Which leaves us with whether the CEA2010 test is worthwhile in assessing "clean" behaviour of a subwoofer. Ignoring the fact that this test is intended to be performed with shaped tonebursts, the stepped distortion weighting scheme is surely intended to model in some way the human ear's sensitivity to different orders of distortion, which is, I think, what the OP is trying to prove/disprove. If you follow the methodology outlined in the Aalborg University paper that soho45 linked to above and try and correlate perceived distortion levels for a music signal with various metrics, you will find that correlation of the THD% measure of r=0.45 improves only to about r=0.60 with the CEA2010-weighted equivalent, which is hardly very impressive.
post #66 of 68
cjwhitehouse,

I haven't forgotten about this yet. I just hit a wall. I did do everything we talked about here, but there is no way to get it to work satisfactory this way.

I played with writing a new program that would vary the HD with frequency, but it was no fun, so I let it go. I have been trying to find a way to quickly change from one distortion set to another, without any luck. I can't figure out a way to do it without creating my own musical track. This would require a lot of work, as everything would have to be done 5x over at least. Each separate note would have to be run through the distortion program to set the correct level, and then you would have to do the same thing for any other distortion profiles.

I compromised and tried sine waves mixed into midi, and got the lead and bass parts to Tool's Stinkfist down, but the drum section was giving me a fit (guitarist here) so I let it go. The work load was to high, so I set things aside again, until last night...

I remembered an old/free/simple looping program called HammerHead Rhythm Station. It is a simple to use drum sequencer with a quick changeable six sample front bank. All I would have to do is come up with a few sound banks with different distortion profiles, and people could play them however they wanted to.

Want do you think?
post #67 of 68
I'm not sure the old grey matter quite follows what you are proposing to do. If you distort each note separately, then the result will not be entirely realistic as no IMD will be created between the different notes of chords. Are you proposing different profiles of distortion in terms of the orders present for different frequency bands or just varying the relative amplitudes? Would be interested in how your "variable HD distortion by frequency" program works and why it ended up "no fun"?!
post #68 of 68
I just read through the many Klippel papers on different types of distortion in driver motors. Very interesting! And now two years later, I think it's great that Ricci's data-bass.com reviews relate the distortion test results to the design and parameters of the subwoofer drivers. It's educational.

I am wondering if Ricci or others have new thoughts about the CEA2010 20% distortion levels?

Personally, I have no problems with them even if the level of distortion is audible. Like passing a local building code home inspection, which doesn't mean that the work was done exceptionally well but rather that it was done good enough to pass like getting a C or D grade.

And as the state of the art for subwoofers advances, there can always be a CEA 2020 standard smile.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs
AVS › AVS Forum › Audio › DIY Speakers and Subs › CEDIA 2010 and Distortion