DIY Waveguide loudspeaker kit

[theshade]

I have been reading on Dr. Geddes designs for weeks now and in those weeks, i have learned a lot. Thank you for imparting the information. During the years of being addicted to audio, i have gone from being a subjectivist to my current state of being an objectivist. The ideas about multiple subs and the waveguide taught me a lot. i also really enjoyed the thread on setting up the nathan wherein measurements were shown.

I have been on a quest for getting the most sound on a budget and I have learned that the weakness is on the loudspeakers. Source components have reached a point wherein you can only measure the differences but not hear them. Amplifiers when driven within their limits also do not show audible differences. Exotic cables are a waste of money and zipcord is certainly enough. Double blind tests are a must when determining sound quality. These are the beliefs that I now espose.

What I still find confusing is the differences in loudspeakers. Due to the influence of Dr. Tooles papers on me, my choices in speakers became much easier. Speakers measured and designed with the help of the NRC very much appealed to me. (PSB, Paradigm) The pinnacle of which to me is probably in the Harman brands. (Infinity, JBL, Revel) However, I also learned about Linkwitz Orions which I think is also based on objectivity along with the Bang and Olufsen beolabs which espouses better off axis measurements than others. In my quest for the best measured speaker, I also learned that YG acoustics has great measurements. Then I learned about DEQX speakers like legend loudspeakers and NHT xd. A lot of differently designed speakers with "great" measurements with different emphasis on directivity. Im confused really. Can someone enlighten me?

Since I live in a third world country, the problem with most of them is the price and the limited output except for JBL pro. I am glad I chanced upon the threads here in DIY audio and learned about the waveguide speakers which is more in my price range. I recently also read Tooles paper "Loudspeakers and Rooms for Sound Reproduction—A Scientific Review" if I understand correctly says that reflections have no negative effects and have even positive effects on audibility. Dr Geddes ideas on room design also supports this as he recommends low frequency absorption without high frequency. Hope I got this right. Welti's multiple sub paper is also first espoused and practiced by Dr. Geddes.

I also have a question about waveguides. Do these JBL's espouse the same theories?
http://www.jblpro.com/catalog/Genera...x?PId=27&MId=5

Once again thank you Dr. Geddes. I have really enjoyed your threads. In fact I am addicted cause I spend almost all my free time reading your informative posts. Hope I can save enough for my own nathan's or abbey's!!!

[gedlee]

Thanks for the kind words.

I don't really think that all those speakers that you mentioned really do have comparable measured performance. Within what they show they might, but its what they don;t show that is the key. One needs good high resolution off axis data that always seems to be sadly lacking.

I agree with 95% of what Floyd Toole writes, but I do disagree about the VERY early reflections, those under 10 ms. Floyd extrapolates the limited results for reflections to say that all reflections are OK, but I have some (admittedly limited) data that says that very early reflections are not good. After 10 ms. though we are in complete agreement. I think if Floyd did have some data on the first few reflections he might find a slightly different answer.

The JBL EOS is a form of waveguide very much like what I do, but it is a dome and domes don't feed the correct wavefront to a waveguide and this will cause undesirable effects. Compression drivers feed a much better wavefront than a dome, although it too is usually not perfect. But lets not forget about the foam plug. This has a major effect and only my designs use it.

I really think that the Summa line is just what you need as you correctly realize that the loudspeakers is pretty much everything. If its not right, nothing is right. I think that my speakers "get it right".

[theshade]

Dr. Geddes, thanks for the reply. yup, the summa line is on the top of my wishlist right now. I have to ask though cause I come from the Philippines, how much would they cost shipped the nathan's for example. And whether I can save on the kit or fully built. We have very good woodworkers here. I am just worried about the crossover if i'll get them right. Additionally, what does built and tested mean? Do you measure all fully built speakers before you send them?

I also have other questions, hope that its not too much.

I don't really think that all those speakers that you mentioned really do have comparable measured performance. Within what they show they might, but its what they don;t show that is the key. One needs good high resolution off axis data that always seems to be sadly lacking.

I do not know if these measurements suffice. They have on axis, listening window, 1st relection, directivity index sound power, directivity index 1st reflection, impulse response, etc. Are the measurements lacking to be able to know whether they sound good or not? Or are the shown measurements themselves bad?

Quoting from jbl "We all know that many loudspeakers have similar measurements but sound different. By going beyond simple on-axis frequency response measurements, JBL defines the ultimate performance specification for new systems – what it will sound like in your room. At the mix position, you hear a combination of direct sound and sound reflected from the rooms surfaces. For sound arriving at the mix position to be smooth and neutral, it is not enough for a speaker to measure “flat” on-axis - it is essential the speaker have excellent off-axisperformance. While other manufacturers use a single on-axis frequency response measurement taken at one point in space, JBL measures monitor systems over a sphere that encompasses all power radiated into the listening room – in every direction. This data reflects 1296 times the information of a single on-axis response curve. Seventy-two measurements of the direct sound field, the reflected sound field, and the reverberant field, the entire sound field heard by the listener, is correlated to optimize response at the listening position. In place of spectral smoothing used by some manufacturers, which actually conceals data, the JBL approach actually exposes flaws in systems, such as resonances, poor dispersion and other causes of off-axis coloration. The data shown below is a set of spatially measured graphs that are the heart of JBL’s philosophy."

I think of all the other speakers i have mentioned the harman brands have the most capability of making good speakers due to their resources. My worry is that even if they have the talent and the resources, the research isn't applied to the fullest in their end product especially the ones I can afford. Nowhere have I seen the amount of technical data than from the harman brands excepting of course your summa's and linkwitz orions. And what I most like about the summa line is that people interested get to have answers directly from you. This gives people peace of mind that what they are buying fully espouse the research. Finally, your designs compared to most state of the art loudspeakers cost much much less and is certainly attainable compared to Revel Ultima's or JBL Synthesis for example.

I do have a question regarding interpreting polar maps. I do not understand how to interpret them. I have been looking at the polar maps of the summa, abbey, and nathan and I just assume that the summa is better. Is it simple to interpret? Sorry, i lack technical expertise. Dr. Geddes do you have a picture of a perfect "ideal" polar response that could be compared?

Finally, regarding room design, will you be posting the additional chapters in your home theater book at your website? I only know that low frequencies need to be treated while mid to highs need minimal but I dont know how to realize them.(construction)

Once again, thank you Dr. Geddes and i hope that I havent taken too much of your time.

[theshade]

I seem to have a problem inserting files. I'll just link the measurements of the JBL.

http://www2.jblpro.com/products/reco...ingblocks.html

http://www2.jblpro.com/catalog/suppo...pe=3&docid=858

[soongsc]

Unfortunately they fail to address stored energy issues which is the main reason for different sound between drivers with very close response curves.

[lineup]

Originally posted by theshade
I seem to have a problem inserting files.

we have had this issue before
in August 2008

Error posting pic
Error posting pic

Probably same thing now.
I get this error message when try attachment.

[gedlee]

Originally posted by soongsc
Unfortunately they fail to address stored energy issues which is the main reason for different sound between drivers with very close response curves.

I'm sorry, but I don't agree with this. If you are going to post in a thread on Summa loudspeakers be prepared to support claims like this that come with no substantiation other than your personal opinion.

[gedlee]

Theshade

I would agree that Harman is the best of the companies out there in terms of R&D and product line. They supply the best data, but it is still lacking IMO. Theye are trying to get a valid but "simple" set of measurements. I think that they oversimplify. What they do is better than most, but I'd still p[refer to see the "real" data.

I assume that you understand polar response graphs shown on axis and at angle off axis. The data in a polar map is identical, just shown in a different way. What becomes obviuos in a polar map is the field of flat response. For example my polar maps are arranged so that they show color changes every 4 dB with the first color being +-2dB from the nominal level (0 dB). So from this you can immediatly see what frequencies and what polar angles have +- 2 dB capability. In my designs this is usually (excluding the LF limit for now) up to 30 degrees and 10 kHz. There is sometimes a small "hole" on axis which means that theya re flat from +-7.5 degrees from 70-10 kHz.

The second thing that become obviuos is the coverage angle becuaes the secodn line is the -6 dB line which is the coverage. This should be fairly constant.

Also obviuos are resonances as they show as bumps in all levels at a fixed frequency. Diffraction will show as a bump (a peak and or a dip) which changes frequency with angle in a constant arc. These take some experience to see.

Shipping the speakers to the Phillipines would cost about 33% of the total cost (based on experince, but this goes up all the time). The croosver is easy to make and would be hard to screw up, but there is always the issue of a bad componenet that is unknown when assembling it. This is easy to test for me, but difficult for the end user. Yes, I do test each speaker that I assemble. Good paint is the hardest thing to do in building the kits.

As to more chapters for my book - well that takes time that I don't seem to have these days.

[soongsc]

Originally posted by gedlee


I'm sorry, but I don't agree with this. If you are going to post in a thread on Summa loudspeakers be prepared to support claims like this that come with no substantiation other than your personal opinion.

Show me identical curves and CSD data, hide the listening comments. I will tell you how different they will sound and then you show the listening evaluation. If I'm way off, you can discredit me all you like.

[gedlee]

You need to prove your point, not have me do it for you!!