Longer sweeps are better, but as with all good things there is a point beyond which there is no real advantage. Find that point and use it.
If the recovered impulse response will be subject to further post-processing, particularly anything looking at the later parts of the response such as low frequency waterfalls or analysis for modal resonances, then longer sweeps give higher signal-to-noise which gives the post-processing better quality data to work with (for LF analysis restricting the sweep frequency range also improves S/N within the range measured). Each doubling of sweep length adds about 3dB S/N. If you are primarily looking at the frequency response then short sweeps are perfectly adequate.
Sumaudioguy, have you actually read some of your posts? You have labelled many as worshippers, call out many as being part of the "church of impulse". You do realize you have posted more references to religion then any other poster?
You have more times asked everyone to not insult but still you throw out these little daggers from mocking the work of EVERYONE, the work of Toole, posting that the common measurement tools suck and so on. Add to that several posts mentioning how much you are idolized by others in some world of audio and how much smarter you are then anyone else.
What really do you you want us to believe and do??? You never seem post valid measurements or links to back up anything you have posted, lots of subjective opinion which seems so strange being that you are a highly skilled and very well known expert somewhere. Do you actually really think you are the only speaker designer that is doing it right?
You have more times asked everyone to not insult but still you throw out these little daggers from mocking the work of EVERYONE, the work of Toole, posting that the common measurement tools suck and so on. Add to that several posts mentioning how much you are idolized by others in some world of audio and how much smarter you are then anyone else.
What really do you you want us to believe and do??? You never seem post valid measurements or links to back up anything you have posted, lots of subjective opinion which seems so strange being that you are a highly skilled and very well known expert somewhere. Do you actually really think you are the only speaker designer that is doing it right?
Enough already. We don't agree.
The IASYS is the only test system which gives me results very close to what I get when testing using other rigourous methods. That is all there is to it. None of the other systems ever used have and that is all there is to that.
Since you do not agree how about completely forgetting about this whole subject as the conclusion is you are right and I am wrong, as Geddes always says Geddes is always right? To me Geddes is usually wrong. Is there anything else to say?
The IASYS is the only test system which gives me results very close to what I get when testing using other rigourous methods. That is all there is to it. None of the other systems ever used have and that is all there is to that.
Since you do not agree how about completely forgetting about this whole subject as the conclusion is you are right and I am wrong, as Geddes always says Geddes is always right? To me Geddes is usually wrong. Is there anything else to say?
Last edited:
Golly, doug20,
you actually took the time to qoute them? I'd think that spending time integrating ideas and thoughts would be much more interesting. Most of the better engineers in certain areas form their own logic after decades of working in a specific industry. It's not that easy to change.
I am gradually getting used to looking at impulses, and cross referecing other forms of data. I can tell say that lot's of things seen are hard to explain. It's like teaching a new doctor how to cleanout a chest wound, sometimes it's just necessary to stick those fingers in there until you dig out the right stuff with someone tutoring you along the side.
you actually took the time to qoute them? I'd think that spending time integrating ideas and thoughts would be much more interesting. Most of the better engineers in certain areas form their own logic after decades of working in a specific industry. It's not that easy to change.
I am gradually getting used to looking at impulses, and cross referecing other forms of data. I can tell say that lot's of things seen are hard to explain. It's like teaching a new doctor how to cleanout a chest wound, sometimes it's just necessary to stick those fingers in there until you dig out the right stuff with someone tutoring you along the side.
It's not about Earl or his position on any of this. The point is there is quite a bit of published work out there that you either choose to ignore of refute as not correct. Where are your published papers or others that support your position. That's all anyone is asking for.
Rob🙂
Not only published work but thousands of successful speaker designs. All based on flawed measurement tools, flawed design theory.
How do we remotely live with all these BAD designs? Maybe SUM will send me a speaker to evaluate and then I will know what Im missing????
Hell, tell me what brand of speaker to buy, it shouldnt be too hard since from a post I think SUM has built/design more speakers then JBL and Danley added together...WOW!! 😀
Many many times throughout history the were many who agreed and then there was the lone voice saying something which did not concur with popular knowledge of the day. This has happened over and over in history as anyone who has studied scientific history knows. I prefer to consider this another example of my lone voice weighed against the popular knowledge of the day.
It is possible I am wrong but so far using rigorous direct measurement methods in my lab and in public demonstrations I am right and the popular knowledge of the day is what is off.
As for Geddes I use him much for reference now as for what he says is often assumed by me as a guide to the incorrect. Leaning by exclusion. But maybe that is a bit much as some of what he writes is correct. I should not single him out though other than he likes to say I do not know what I am talking about. Well right back at you Earl.
I never said I built more stuff than JBL, I said I never saw a Harmon Product in a musical recording studio. But feel free to invent and discredit however you see fit. That is of your choosing.
Thank you all again for bringing me up to date on the popular knowledge. You really have helped a lot. As I said long ago, "The same old mistakes, the same old misconceptions, and the same mediocre results. Nothing has changed."
I will not be posting to this thread again. There is no point. However reading DIY has been very interesting and will be for some time I am sure.
It is possible I am wrong but so far using rigorous direct measurement methods in my lab and in public demonstrations I am right and the popular knowledge of the day is what is off.
As for Geddes I use him much for reference now as for what he says is often assumed by me as a guide to the incorrect. Leaning by exclusion. But maybe that is a bit much as some of what he writes is correct. I should not single him out though other than he likes to say I do not know what I am talking about. Well right back at you Earl.
I never said I built more stuff than JBL, I said I never saw a Harmon Product in a musical recording studio. But feel free to invent and discredit however you see fit. That is of your choosing.
Thank you all again for bringing me up to date on the popular knowledge. You really have helped a lot. As I said long ago, "The same old mistakes, the same old misconceptions, and the same mediocre results. Nothing has changed."
I will not be posting to this thread again. There is no point. However reading DIY has been very interesting and will be for some time I am sure.
You are not going to be a known name on the pages of audio history. 🙄 You should just get over that silly superiority complex issue you have.
Why not post the exact brand of speaker that has your expert knowledge behind it?
I will buy it and I will listen to find out what is superior about it. All that matters in the end is real products, real people and real listening experiences. You might want to educate yourself on those small facts, you might actually learn something.
Why not post the exact brand of speaker that has your expert knowledge behind it?
I will buy it and I will listen to find out what is superior about it. All that matters in the end is real products, real people and real listening experiences. You might want to educate yourself on those small facts, you might actually learn something.
Last edited:
Heh,
Whenever technology/knowledge cannot advance faster than discussion in a thread, it always ends up in a spit fight.😀 There are many groups of scientests and engineers that just have their own ways. It seems good at first, but just like marrying a close relative, the offsprings become...
If there is really a disagreement in measurement methods, the best most productive way is to take similar examples and have a head on comparison. Up to now, I have not seen anyone able to do this.😎
Whenever technology/knowledge cannot advance faster than discussion in a thread, it always ends up in a spit fight.😀 There are many groups of scientests and engineers that just have their own ways. It seems good at first, but just like marrying a close relative, the offsprings become...
If there is really a disagreement in measurement methods, the best most productive way is to take similar examples and have a head on comparison. Up to now, I have not seen anyone able to do this.😎
I have a question for all the experienced people around here.I know from experience that pestering is annoying. Let's get off the man's back. and discuss measuring. Otherwise this thread will just head down an already beaten path of flaming.
I would say that until he offers proof, stick with what you know. At least that's where I'm at.
Check out these new graphs from REW after and before stuffing the ports with cotton sequentially:
Then their respect impulse responses:
snip...
Dan
In the above plots, why would there be response before time zero? This does not happen in reality. Due to mathematical process of creating an impulse, I understand it will show up. But is it not possible to truncate the purely mathematical section and only show the part that reflects the actual response?
In the spectrogram that is an effect of how such plots are generated. A window is moved along the impulse, starting some way before the impulse peak and continuing on after it. The edges of the window start to encounter the impulse while the window centre is still at negative time, the window width causes a smearing of the time resolution - a narrower window has higher time resolution but lower frequency resolution and vice versa.
Regarding the impulse plots, the first factor is how the location of time=0 is chosen. For convenience it is often placed at the peak of the impulse, or sometimes at the first point the response rises above 20dB below the peak, but other analyses can be performed to better determine where t=0 should be located, from a separate delay measurement step before making the response measurements through to cross-correlation of the impulse with a minimum phase equivalent (which is purely causal and zero for t<0).
A second factor, with logarithmically swept sine measurements, is that harmonic distortion is separated from the linear part of the response and ends up at negative times, appear as scaled down copies of the main impulse according to the level of the particular harmonic - those appear some way back, however (depending on the sweep rate) and are not visible on the time scale of the plots you posted.
The slight impulse spreading which is a consequence of measurement bandwidth cannot simply be truncated as that would alter the resulting frequency response, which would then no longer represent the system measured, but it is very small and impulses viewed on a linear scale would not typically show it, once the correct t=0 point has been located the response should be causal - significant ringing in the response before t=0 may indicate a problem with the measurement.
Hi,
You cannot remove the impulse pre-ringing if the system is expected to have (almost) linear phase.
- Elias
You cannot remove the impulse pre-ringing if the system is expected to have (almost) linear phase.
- Elias
I also noticed if the DUT has greater bandwidth, there will be more pre-ringing. Is this pretty much expected?
Is the software even avaliable for free? I mean its kind of a pointless debate if DIYers have no access to the hardware/software used for this "alternative" measurement method.
- Status
- Not open for further replies.