formica Posted March 30, 2005 Share Posted March 30, 2005 ---------------- On 3/29/2005 8:42:12 AM dragonfyr wrote: But several primary tools for quantifying and qualifying phenomena in the time domain, and subsequently the frequency domain (and via Hilbert and associated transforms other domains...). .... The holy grail of these techniques is TEF. Other mechanisms, with various strengths and weaknesses exist as well. MLS (Maximum Length Sequence)/MLSSA, SMAART, SysID, and others. There also exist several quasi-tools that I will refrain from commenting on, as I am (and many others are) NOT convinced as to their validity. ---------------- Following an extended post by dragonfyr in the bass trapping above k horn corners thread in the "Architectural" forum... I decided to post a follow up here, hopping on getting a little more technical info and examples of using TEF analysis versus using the more typical MLS. What does it measure and how is the data presented? Thanks... Rob Quote Link to comment Share on other sites More sharing options...
sfogg Posted March 30, 2005 Share Posted March 30, 2005 http://www.gold-line.com/tef/tef.htm Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted March 31, 2005 Share Posted March 31, 2005 I just got in (far too late!) and discovered the post! Great! Please Pardon the delay in posting, and I will try to get on this tomorrow. One quick note....The TEF (Time energy frequency) analyzer is capable of performing a great many measurements, including time delay spectroscopy (TDS) pioneered by Dick Heyser (and TDS is often, if erroneously, referred to as TEF) as well as Maximum length sequence (MLS) analysis as well, adding to its robust array of options. And one more quick note. An exhaustive presentation of TDS is a bit beyond the scope of this thread. It is a bit too complex for that, and the AES publication on Time Domain Spectrosopy (an anthology of Richard Heyser's work... mentioned previously) presents quite an exhaustive selection of papers regardinbg this! But I will try to present an overview of the basic operational characteristics and the capabilities it offers to us. So... please pardon me while I run to bed for a while! Quote Link to comment Share on other sites More sharing options...
Champagne taste beer budget Posted March 31, 2005 Share Posted March 31, 2005 Hey dragon, just wanted to let you know that I, for one, am glad you decided to come back to the forum. Granted, some of your posts make me feel like I'm reading something about thereoretical astrophysics, but it does cause me to read carefully, try and absorb and comprehend everything I can, and drives me to search out additional information in areas of sound and it's measurement that I previously never knew existed. Thanks. Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted March 31, 2005 Share Posted March 31, 2005 Sorry 'bout that! I must admit that I find myself at a bit of a loss regarding just how to explain some very complex subjects (for example TDS!) in just a few sentences. As if one errs on being a bit general and making oversimplified statements for the sake of brvity, too many prefer to attack you for being an "idiot" for making generalizations, all the while others who are not intimate with the subject are frustrated that its too advanced! There has to be a balance - or at least a civilzed discussion, perhaps via private asides or additional threads, for delving into the more esoteric aspects of the theory as well as exploring some of the very practical applied apsects! In this way I think that all levels of interest and proficiencies can be addressed, especially as mentioning any of these technologies is based upon a prerequisite knowledge base that actually addresses 90% of all the questions that arise. After all, technologies such as TDS and MLS are simply tools to examine phenomena based upon a different perspective then is normally employed. And once you understand the new perspective (the primacy of the time domain and its relationship to all of the other domains where topologies are mapped to these other domains via transforms), the ability to examine phenomena within those realms becomes a wonderful tool. And the tools cease to be quite so exotic (that is until you choose to delve into the 'black box' internal algorithms! Then all bets are off!! But the good thing is that these technologies (TDS & MLS), while still relatively new and unfamiliar to many, are NOT fly by night concepts, and they have been thoroughly scrutinized for their accuracy, so I hope that we can avoid the emotional aspect where some claim that they are just marketing gimicks (as too many others, unfortunately, are!) ) And in this way, perhaps some can discover how these tools can be of great use in explaining aspects that have been heretofore extremely difficult at best to understand and predict... And I can say that about 80% of all the threads that routinely arise on the site (no, it won't address the political, economic and religious issues!) regarding transducer; speaker (transducer-transducer-xover combinations); and room-speaker interaction, can be illuminated with amazing clarity if the measurements are done correctly. (and no! No amount of technology will improve on operator error! So let's get that one out of the way right now!) Sound like an amazing claim!? Good! What is really neat, is that it is true! And the even more amazing fact is that when these tools are correctly utilized, they contribute significantly to a greater listening experience! 'Cause after all, even for us technoids, at some point it is nice to put the toys away and to enjoy the fruits of our labors! At least to me, that is what its all about! More to follow... Oh, and as there is probably NO way to avoid dealing with the trying to take a sip from the hydrant issue involved here, take a look at the Goldline website and the software descriptions for some of the tools. I will be the first to say that the web site is confusing at BEST!; and that the illustrations are left over from the TEF10 of 15 years ago!!! (read: "horrendous"!). Also, You might want to check out some of the resources on sites such as www.prosoundweb.com. They have archived quite a few of the SynAudCon tech sheets as well as much more that are available to the public without membership. You will also encounter some of the applied issues that come up here as well. May I suggest that you go there and do a search on "Nyquist", "Heyser", and "TEF". Any of the Pat Brown or Don Davis articles are excellent, as well as Many others! Additionally, the info on waterfall plots and general MLS info in the links to www.purebits.com are quite good. Just be aware that it is primarily a post processing tool and NOT a high resolution, noise immune DAQ system! The illustrations and plots are much more representative of current technology than the terrible Goldine TEF site plots! Hope that gives you a place to start looking until we can jump into this! Quote Link to comment Share on other sites More sharing options...
sunnysal Posted March 31, 2005 Share Posted March 31, 2005 just a quick aside, dick heyser loved the sound of klipschorns! he reivewed them once. he also noted the extreme (his words) time delay between the bass horn and the rest of the speaker´s drivers...tony Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted March 31, 2005 Share Posted March 31, 2005 "Absolutely", regarding the gross misalignment of the acoustic origins of the drivers! Two things that would immediately benefit the Khorns and LaScalas (and others as well, but I must admit to personally focusing on these!)... One is to actively align the acoustic origins of the drivers in the time domain. This would necessarily be tuned to a focal pointas movement off axis would result in a 'cosine error' where the relative distances and hence the corresponding direct propogation times would differ...(And of course, there is still the prerequisite that the acoustic center of a CD horn must be acoustically measured, as it moves into the throat area and does not exist at the diaphragm...plus, a smaller anomaly, it's location is somewhat frequency dependent! In other words it 'moves' w/ frequency!)(Re: DBKeele's research) Second would be the exploration of a coaxially oriented horn arrangement where the horns were stacked coaxially relative to the bass bin, and aligned in the time domain with appropriate delay, providing an alignment addressing off axis variation of the direct signal arrival times Ld. You see! So many considerations and tradeoffs! But that is what makes the discipline so exciting! Quote Link to comment Share on other sites More sharing options...
Guest Anonymous Posted March 31, 2005 Share Posted March 31, 2005 Dragon- Where can you get a TEF? Have a good night? George Quote Link to comment Share on other sites More sharing options...
Champagne taste beer budget Posted March 31, 2005 Share Posted March 31, 2005 Well, I was curious, so I checked it out a little bit. Seems you can get one here for about $4,000, though to get all the software and associated goodies, you'll end up spending nearly 8K. As I was chatting with Leo, he brought up another option that runs from $1800-4800, depending on how serious you want to get. I realize that's a lot of money, but then again, I was at a trade show last night looking at new scanners for the shop, range from $2500 to around 8K for the latest/greatest from Snap-On. I can see where, if a person was in that field and doing work of that type every day, buying a TEF wouldn't be out of line. Quote Link to comment Share on other sites More sharing options...
WMcD Posted March 31, 2005 Share Posted March 31, 2005 I'll point out that in making a choice between measurement systems it is important to look at what comes with it. The LMS system (a swept signal measurement system) exports data to LEAP, which is a comprehensive design program. You can do some very sophisitated designing and simulation in LEAP, then build the thing, then test it with LMS, find troubles, and crank the design again. This does not address the relative merits of the measurement systems, but it might make the choice easier. Gil Quote Link to comment Share on other sites More sharing options...
Champagne taste beer budget Posted March 31, 2005 Share Posted March 31, 2005 Whooaaa big fellow.... I ain't saying I endorse that product, or that I know anything about using it, just that the "stuff" out there isn't exorbitantly expensive if it was something you had a need for. That was just one example, I'm sure there are all kinds of goodies out there. Right now, all the acronyms mean nothing to me. I have my own set to worry about with vehicle computers. Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted March 31, 2005 Share Posted March 31, 2005 GoldLine is distributing the TEF analyzers. And yes, they are pricey, but the new TEF25 runs for approximately 1400 plus lab grade mic such as an earthworks M30 for ~500 and the additional software. It is not a casual tool and heretofore they have been purchased primarily by acousticians and speaker and microphone manufacturers. They are routinely employed for transducer design as well as for tuning both large and small acoustical spaces (as they have different criterion). Additionally, it does take training to understand how to conduct valid measurements. So they are not a 'toy' to casually run out and buy. But on the other hand, everyone doesn't need to buy one. If one desires their listening envronment to be analyzed the services of a consutltant can usually be found for the cost of a couple of hundred dollars. My analyzer is currently in Nashville on loan to a friend who is working on a studio, but if my schedule lightens up this summer I was rather fancying the notion of trying to see if I could generate a comprehensive series of measurements for the Heritage series (And if anyone has access to a Jubilee.....hint!). This data would also allow for those desiring to employ the use of an active crossover or delay to adjust for the offset of the acoustical origins of the transducers. While I hesistate to stick my neck out too far for practical concerns, I would might even be available to shoot a few rooms...but, bear with me as I don't want to over commit myself and make claims or promises that I may have difficulting keeping. But in any event, with some accurate date, there is also the possibility of acoustically modelling rooms with the above data in EASE, coupled with accurqate reflectivity data regarding the rooms, and much of the process could be automated. And while that would not be perfect, it would certainly give one an EXCELLENT map to start from. But I am getting far ahead of myself! Tonight I will be trying to gather a semi-coherant description of some of the capabilities of TDS and MLS, and the derivative data that can be generated. The capabilities are pretty easily listed, but an explanation of exactly what each part demonstrates and how it is used and how they all interact is all a bit daunting...as i am sure that few want a techno-babble description! But at the same time just making a bunch of big claims and saying "trust me" isn't sufficient either! Oh well! So bear with me as I try to assemble a brief overview of some of the capabilities. And in the process I will try to see if I can scan a copy of Don Davis's Audio article on LEDE rooms - not to sell anyone on the concept but to let you see how the analyzer is used in a simple example. Additionally, I will try to scan some info regarding the effect of taking two ideal drivers (you know, the flat from DC to gamma rays tranducers we all have laying around) and showing you the effects of adding a second driver. I will also suggest a simple experiment that will very dramatically illustrate the effects of multiple drivers unpon a signal. And if you have kids, it will make a great basis for a science fair project, as the effects are dramatic! So let me get to work...And if any of you have specific questions, as I am sure many of you are at differing levels of familiarity with all of this, free to send me private posts, asn I will either respond directly or try to incorporate the ideas into any info I can find... Gee I wish we had a blackboard here! As the majority of ideas are not easily described, but easily illustrated! Quote Link to comment Share on other sites More sharing options...
Guest Anonymous Posted March 31, 2005 Share Posted March 31, 2005 Dragon- It is my understanding that with a TEF and using the software to plot the measurements (Nyquist Points), that it will also plot on the z-axsis. Now, plotting on the z-axsis will reflect time over the course of taking measurements and therefore give a 3-d model. Is this correct? What is the benifactor of having a 3-d plot with a time reflection. Have a good night George Quote Link to comment Share on other sites More sharing options...
Champagne taste beer budget Posted March 31, 2005 Share Posted March 31, 2005 I'm hitting the sideline and watching, learning and appreciating the knowledge of those around me. Quote Link to comment Share on other sites More sharing options...
formica Posted March 31, 2005 Author Share Posted March 31, 2005 ---------------- On 3/31/2005 8:22:40 PM dragonfyr wrote: So bear with me as I try to assemble a brief overview of some of the capabilities. And in the process I will try to see if I can scan a copy of Don Davis's Audio article on LEDE rooms - not to sell anyone on the concept but to let you see how the analyzer is used in a simple example. Additionally, I will try to scan some info regarding the effect of taking two ideal drivers (you know, the flat from DC to gamma rays tranducers we all have laying around) and showing you the effects of adding a second driver ---------------- Sounds like a good proposition... I thought I'd give a little background to source of my original question. I believe we can learn a lot from measurement (and understanding the data it produces) so I've been tempted to experiment a little further in that direction. Like most of you guys (dragonfyr excluded), I'm simply a hobbiest. I asked these questions not really because TEF would be within my reach (that's just too much money for a regular hobbiest)... but I'd like to know if I'd be wasting my money on the products which are. In other words, if TDS measurement leaves your more standard MLS and MLSAA measurement in the dust... why bother with the latter? Right now, I've got quite a few audio projects lined up, including building a dedicated room (design work is under way), building an integrated InfiniteBaffle (drivers have arrived), restoring an old pair of Klipschorns (D models I've bought), and measuring and tuning the entire combination. I may have to sell a couple of those "DC to gamma rays tranducers" a have lying around to help finance these purchases... For the measurement part, I was hoping on getting by with a Behringer ECM8000 measurement microphone , a Behringer U-802 for phantom mic power, and ETF software ... the whole package for under 400$US. I have tried the ETF5 software, and it is powerful enough to go beyond my current knowledge of measurement systems. In other words, it should give me hours of hands on and learning experience... and I'll emerge a more knowledgeable person. Later.. Rob Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted April 2, 2005 Share Posted April 2, 2005 I am going to attempt to upload a bunch of documents that can present topics much better then I can... Let me try to put a few things into perspective in an attempt to anticipate a few questions... Lets agree to assume that the measurements, assuming they are structured properly and the proper perquisite knowledge is utilized, are valid. (as always, if you ask the 'wrong question' and set the parameters incorrectly, your results will follow from the flawed question and the results will likewise be flawed.) There is no sense wasting time debating incorrectly measured or interpreted procedures or results! Also, as they have withstood the scrutiny of both academic engineering environments, lets avoid debating on whether TDS or MLS techniques are valid. Is this reasonable? Also, while I will maintain that the measurements are sound. There does exist possivbilities for further evolution of models and interpretation of phenomena. It is precisely this aspect that renders acoustics a vibrant field! interpretations and the significance of relationships of some of the results have been evolving - as they should! With the advent of looking at acoustic phenomena in the time domain (that has profound impact upon the frequency domain), the results have been multi-faceted. The frequency domain consists of many observable phenomena that are not in their basic forms. Rather they are the results of a series of complex interactions, and the frequency domain does not allow us to atomistically isolate the various components. It is by adjusting these various components that allow us to rectify or optimize the resultant phenomena that we see in the frequency domain. And the interaction is rather profound. The time domain affords us a unique perspective on the world of acoustical phenomena. A quality such as intelligibility and its associated ramifications is able to be addressed in the time domain. All of us have experienced gyms, airports, etc. where the PA system has plenty of gain, yet we cannot understand a word that was said! That is a direct result of multipath signals which are not coincident in the time domain. Now before everyone jumps on this statement, I will point you to the forthcoming Davis LEDE article for an explanation of the Haas effect (referring to the limits of our ability to localize sound, much as one visually perceives depth of field. There are parameters and limits to this ability). Various other derivative tools have been developed such as STI, %ALCONS, RT60, etc.. Again, it is beyond the scope of this attempt to delve into the strengths and weaknesses of the various ratios and practical measures. Suffice it to say that what makes the advances in this area of audio so stimulating is that the time based measurements have allowed us a window into the acoustic phenomena that allows us to examine and potentially adjust the various parameters. And these tools have allowed us to begin to examine and speculate exactly what the various combinations and relationships 'mean' and just what manipulation results in a desired improvement. In other words, we are now free to speculate and postulate in a much more informed manner! And, unlike so many other dead disciplines where the fundamentals were adequately determined 100 years ago and now all we do is plug values into formulas, electro-acoustics is a currently evolving field! The best article that presents an overview of the components and the use of time based analysis for tuning rooms is D. Davis' LEDE article published in Audio. It will take you through a basic examination of a room and present a brief presentation of measurements and their practical significance while introducing you to many of the associated considerations that make them valuable. The two predominant time domain measuring techniques are TDS and MLS. A major distinction between the two is ambient noise immunity. MLS measurements will be adversely affected by ambient noise, often critically. A MAJOR advantage of TDS is its radical immunity to ambient noise. Sweeps can be made in the middle of a raucous environment without the need to eliminate ambient noise. And the presence of noise in the space is a major practical issue in many rooms, construction/ contracting sites, and venues both indoors and outdoors. Examples of major MLS packages are those from Dra Labs MLSSA and WinMLS. TDS is presently only licensed via the TEF analyzer and is a much more complex system employing tracing filters which provide significant ambient noise immunity. The best description of the very complex technique is found in both the AES Time Delay Spectrometry, the anthology of Richard Heysers works, and in Sound System Engineering by Don & Carolyn Davis. The TEF analyzer was initially designed to enable Disk Heyser's Time Delay Spectrometry (TDS) measurements. So it is common for many (including myself!) to refer to a TDS measurement as a TEF measurement. The TEF analyzer is also capable of performing Maximum Length Sequence (MLS) time based measurements. So the TEF is the 'Swiss Army knife' of analyzers. It is the most comprehensive analyzer available. In addition, it also provides significant post processing that further reinforces it as position among the most comprehensive of all of the various available tools. Other more limited tools such as SIA SMAART use time based measurement to do more focused tasks (SMAART, for example, is more oriented toward EQing a live system in a complex environment). There are also a vast array of post processing tools exist with a wide range of capabilities, ranging from MatLab and LabView , to programs such as PHd, CATT, Ulysses, and EASE/EARS that allow for the modeling of acoustic spaces complete with detailed CAD diagrams and component detail attributes such as the reflective coefficients of all of the geometrical surfaces (such as walls, curved seats, ceiling beams & trusses, etc.) Programs such as EASE also allow for the importing of time based measurements that facilitate adjusting the room parameters and auralizing the changes listening to the audible changes relative to a particular listening position in the room after adjusting room parameters. A rather amazing feat of programming! Both TDS and MLS systems can be used to provide measurements for many levels of the acoustic food chain. They are used in the manufacture of individual electronic devices, transducers and microphones, speakers (featuring the interaction of various transducers, crossovers, enclosures, etc., to the interaction of the speakers within an acoustic space as they interact both with other speakers and with the acoustic space. So there are many complex levels of interaction between all of these aspects to be considered in the real world! Additionally, with the extensive array of post processing tools for manipulating the test measurements, an array of powerful new perspectives emerge. And these tools allow us to focus in on the nature and various levels of the complex interactions, depending on how we choose to structure the measurements. Never before have we had such control and power. Suddenly everything becomes both wonderfully apparent and amazing complex! We have quite a toy box of tools at our disposal! But simply having the tools doesn't solve the challenges! Now one must take the time and make the investment to learn more about what the measurements illustrate and what the basic nature of the interactions entail. In other words, we become an active member in the chain! After all, just as giving a person an array of state of the art tools and analytical gear does not make them an expert auto mechanic/design engineer! So now that I have again expanded the possibilities rather then simplifying them (oh well!), I will try to present some resources and links to allow you to look at some of the tools and to also become more familiar with the practical considerations. After all, the point of this is not simply to overwhelm you with the wealth of new tools, but to make the complex a little more understandable and manageable. But in order to do this it is worthwhile to identify the layers of interaction. The physics principles are the same, so you neednt learn them over and over, but it does make life easier to be able to deal with one level of abstraction at a time, while at the same time recognizing that the various realms DO interact! I am still a bit perplexed as how to best present all of the info, as it is predicated upon so much other info! I fear I have stumbled into a tar baby! (Now would be a good time to suggest anyone who is interested to attend one of the week long SynAudCon seminars! It will be the best week you could spend regarding things acoustic! After uploading the various papers and resources, perhaps folks can glance through them and we can discuss basic time domain effects or anything else that is of interest. And I will give you the very simple, yet dramatic, demo that you can do at home with two Aura Tone-like speakers (twin 3cubes or, if you prefer, finally a useful purpose for the Bose cubes you have sitting around and using for paperweights!) Well here is a start, and I will work on the uploads and trying to scan a few other articles, in particular the fundamental LEDE room article that introduces the methodology in a very educational manner. Additionally I will attempt to present a few of the rather special tools that the TEF analyzer affords. I hope that this will begin to show how valuable a tool it can be for the practical solution of many complex issues as well as how valuable it can be for research functions. Sites with great educational aspects: http://www.audiosystemsgroup.com/ http://www.audiosystemsgroup.com/publish.htm http://www.prosoundweb.com/install/sac/n26_4/nyquist/nyquist.php# http://www.mcsquared.com/tefadr.htm Likewise their calculators and demos!!!!! Worth the time to explore this site! http://www.prosoundweb.com/studyhall/sac/ Check out the articles! I will post some of the same files for download as well along with a few additional including Acoustical Measurements pt.3 & 4. Products: http://www.gold-line.com/tef/t-ship.htm (Be warned that this is perhaps the worst site I have encountered in a long time! I think it is an attempt to keep the TEF analyzer a stealth tool!) And the screen captures are from an OLD TEF 10 from at least 17 years ago! Check out the anemic waterfall plot! Just DONT take it as accurate regarding the tool! The software descriptions are also understated and the illustrations, well, lousy! You will see MUCH better illustrations of many of the tools in other articles! http://www.winmls.com/ http://www.mlssa.com/ http://www.renkus-heinz.com/ease/index.html http://www.catt.se/ Be sure to explore the demonstrations! They are a tutorial in themselves! http://www.mcsquared.com/tefadr.htm Likewise their calculators and demos!!!!! Worth the time to explore this site! http://www.ifbsoft.de/ Ulysses http://www.at.oersted.dtu.dk/~odeon/ Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted April 2, 2005 Share Posted April 2, 2005 Only one attachment per post? Oh my! AcousticMsrmts1Basics.pdf Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted April 2, 2005 Share Posted April 2, 2005 AcousticMsrmts2TimeDomain.pdf Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted April 2, 2005 Share Posted April 2, 2005 AcousticMsrmts3FreqDomain.pdf Quote Link to comment Share on other sites More sharing options...
dragonfyr Posted April 2, 2005 Share Posted April 2, 2005 AcousticMsrmts4SndLevels.pdf Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.