Hi John, thanks for the charts. I expected to see more of a difference.
The problem with a f/r chart is it needs to be compared with distortion figures. A rise of 3 dB for example, may be due to a resonance of some sort, and not a huge issue to most.
It could also be due to the harmonics of a sub multiple of that frequency. A trap with software testing.
Real results need a manual sweep and a tuned filter, not just measuring the wideband output from a swept sine wave or pink noise.
That takes more gear and time.
Geoff.
The problem with a f/r chart is it needs to be compared with distortion figures. A rise of 3 dB for example, may be due to a resonance of some sort, and not a huge issue to most.
It could also be due to the harmonics of a sub multiple of that frequency. A trap with software testing.
Real results need a manual sweep and a tuned filter, not just measuring the wideband output from a swept sine wave or pink noise.
That takes more gear and time.
Geoff.
Hi Geoff.
The whole point of the exercise was to show the audible difference between materials. Using the same drive unit in each box takes away the only possible variable.
To be honest, I expected more of a difference also, especially between the two higher grade materials (BB and MDF), and spruce sheathing.
I tried to be as scientific and consistant as possible with these tests. To my way of thinking, that means using an accepted measurement method, and removing as many variables as possible. The program uses an MLS signal at a high sample rate, repeated 7 times then averaged. I can't see how the method you descibed would be more accurate.
I took the time to run these test to satisfy myself as well as to post here. If anyone else knows of a more detailed study, or if someone with a better test procerdure and equipment would like to run the tested, thats fine.
But for me at least, the question is answered - there's no signicant difference in the sound quality of a box built with either baltic birch or MDF.
The whole point of the exercise was to show the audible difference between materials. Using the same drive unit in each box takes away the only possible variable.
To be honest, I expected more of a difference also, especially between the two higher grade materials (BB and MDF), and spruce sheathing.
I tried to be as scientific and consistant as possible with these tests. To my way of thinking, that means using an accepted measurement method, and removing as many variables as possible. The program uses an MLS signal at a high sample rate, repeated 7 times then averaged. I can't see how the method you descibed would be more accurate.
I took the time to run these test to satisfy myself as well as to post here. If anyone else knows of a more detailed study, or if someone with a better test procerdure and equipment would like to run the tested, thats fine.
But for me at least, the question is answered - there's no signicant difference in the sound quality of a box built with either baltic birch or MDF.
MJL21193 said:
Did you listen to the different boxes? Or are you going to let the graphs tell you what they sound like?
I don't think your tests are conclusive.
I have a whole crowd of people who heard a difference between 2 otherwise identical pairs of speakers made with MDF & plywood -- they couldn't agree which they liked better, but what they heard was consistent.
dave
planet10 said:
Without looking into my crystal ball, I could see this coming.😉
How much more could I do? The boxes were not for listening. I heard the MLS signal, but honestly, listening is too subjective. It's coloured by any number of variables, such as mood, time of day, stress level, preconceptions. It's not a reliable, consistant measure of a materials quality.
How seriously would you take me if my results were based on my 41 year old ears?
MJL21193 said:
I don't have to -- you are the only one who matters for your speakers. I already have my own results. Besides individual sessions, a group test with some 20-odd people and at least a couple sets of really well trained ear/brain systems. A post test scrutiney showed that what people heard was generally consistent.
Listening tests are fraught with potential problems, but no more than a mic, SW & a sound-card. And in the end it isn't the computer listening to the speakers.
"In all things audio, the ear is the final arbiter."
Harry Olson
dave
MJL... its good work you have done ... but I also think that music is much more complex than a test signal .... maybe that does make a difference somehow
tinitus said:
Thanks Tinitus.
Music is complex, so why is it that a favourite track is not used for acoustic measurements? The following is a quote from PureBits.com:
"MLS in a nutshell...
MLS is an abbreviation for Maximum Length Sequence. It is basically a pseudo-random sequence of pulses.
Nowadays Maximum Length Sequence measurements are quite standard in many different application fields. One of them is acoustics.
Using MLS techniques, it is possible to perform quasi-anechoic measurements of a loudspeaker without having to place it inside an anechoic chamber (a room free from echoes and reverberations). The impulse response can be easily windowed in the time domain, in order to analyze the signal and reject the reflections from the walls of the room. Moreover the room impulse response itself (and all the related parameters such as reverberation time) can be measured.
The MLS method can also be used to analyze and obtain information about the impedance or the absorption coefficient of a surface. "
Fairly straight forward - a signal tailor made for testing acoustics.
planet10 said:
I have sold speakers, traded them for other things, and even given them away. I've not heard any complaints.
Earlier in this thread (i think) we discussed the validity of the listening test detailed above, and my opinion still stand.
The whole reason for this thread in the first place was your advice not to use MDF - use anything except MDF, even solid wood (and softwood at that). Yet in the face of all of the evidence presented here (and also by ShinObiwan, on another thread), you refuse to concede.
What more can I say?
What do you use to simulate a speakers response? Pen and paper? An abacus? Trial and error? Pretty sure you use a computer for this complex analysis. I do, and I generally trust the results. Same goes for measurement.
In the end the ear is the final judge, but IMO a million other factors are more important to good sound quality than the kind of wood used.
MJL21193 said:
Which when analysed further supported my contention...
As i've said before, you don't have to prove anything to me, i've done my own tests and the proof is in the product. Not that that is going to stop me from getting even better at it.
The analysis only gets you into the ballpark. There are many things that the computer doesn't tell you, and one has to rely on experience & actually building the things & tweaking from there.
dave
"The analysis only gets you into the ballpark. There are many things that the computer doesn't tell you, and one has to rely on experience & actually building the things & tweaking from there."
That can work the other way too. If I can hear a possible problem, I may use a computer to identify the problem in order to find a fix.
What we are trying to establish is what may be a better material to build with. One thing I know for sure is MDF is worse than particle board.
BB Ply is out of the equation here in Oz. If you can find it, it's very expensive - some went on e-bay a while back at well over $100 a sheet, plus freight.
John, your testing has been interesting. It means we need to keep looking, who knows, we may find something better than what we are using now, if we are not satisfied. The swept sine wave I described earlier is only useful for frequency response, and as I demonstrated in another thread, only shows a small part in how a speaker sounds.
Regards,
Geoff.
That can work the other way too. If I can hear a possible problem, I may use a computer to identify the problem in order to find a fix.
What we are trying to establish is what may be a better material to build with. One thing I know for sure is MDF is worse than particle board.
BB Ply is out of the equation here in Oz. If you can find it, it's very expensive - some went on e-bay a while back at well over $100 a sheet, plus freight.
John, your testing has been interesting. It means we need to keep looking, who knows, we may find something better than what we are using now, if we are not satisfied. The swept sine wave I described earlier is only useful for frequency response, and as I demonstrated in another thread, only shows a small part in how a speaker sounds.
Regards,
Geoff.
Dave,planet10 said:
It was never my intention to prove it to you. You prefer BB plywood, that's fine. I just wanted to put forth some solid evidence that MDF is not an inferior choice. It's as good as BB plywood for speakers.
It's good to know there is an alternative to a hard to come by, expensive product, for building speakers of the highest quality.
Geoff H said:
Geoff, you would pick a wood product that I didn't test!😉
I seriously doubt particle board is better than MDF, as they have a very similar structure.
In the end, despite the evidence to the contrary, each person has thier own personal bias, which colours thier opinion on things. Also, the tendancy to accept without question the unsubstantiated opinion of someone "in the know" at face value. I prefer to form my knowledge base on a more substantial footing. Show me the proof, the science not the heresay, or the subjective opinions.
The testing has been interesting, and I plan to continue. Next is a run at different damping methods.
I also would like to get my hands on a piece of Corian to build a test box from. Anyone out there with some 1/2 inch thick scraps kicking around? I'll pay the shipping.
MJL21193 said:
There was at least 1 AES study/paper that confirmed this. MDF was at the bottom of the heap, with particle board a couple higher.
dave
MJL21193 said:
Keep looking for a way to do CSD -- that will give much more meaningful & interpretable results.
dave
planet10 said:
Can you post a copy of this study? I'd be interested in reading it. In what way is PB better than MDF? As a speaker box construction material?
Is the difference detectable to the un-biased human ear? Can it be measured by objective equipment?
Your answers here are too brief, and lack vital information.
planet10 said:
Hi Dave,
I also would like to know which AES paper are you referring to.
The only AES PREPRINT I know is by Backman, 1996, and this is the abstract which actually says, in my understanding, that all the tested woods are almost equal, BUT MDF and Plywood have some damping ALSO at higher modes.
Preprint Number: 4395 Convention: 101 (October 1996)
Author: Backman, Juha
The paper reports the results of an experimental study where nine enclosures of similar dimensions and material thickness but with different materials were measured. The parameters measured included vibration, far-field sound radiation, and near-field sound radiation. The internal standing waves of the enclosure were damped as much as possible. The enclosures consisted of flat panels glued to each other with no additional support structures. The materials included particle board, plywood, MDF, and plywood with different vibration-damping sheets made of different viscoelastic materials and lead. Among the common enclosure materials (MDF, plywood, particle board), there were rather small differences between the enclosures at the lowest resonances, although both plywood and MDF exhibited small damping also at higher modes. The lowest mode is usually the only mode that needs serious attention, since it is both the most difficult to control using the internal damping and radiates more efficiently than the higher modes because all the points of a given surface vibrate in phase. Only the application of an effective damping material produced so significant a reduction in the resonances that the coloration due to the resonance was removed. As expected, the vibration level well below the lowest resonance is almost unaffected by the choice of enclosure material, with the exception of the enclosure where a lead sheet was used which had a mass significantly larger than the others.
MJL21193 said:
I guess so... but also the only easily made objective measurement that will tell us anything useful.
FR doesn't tell us anything but the gross levels and we wouldn't expect much differences, and the impluse response is too hard to interpret (ie a CSD interprets the impulse in a manner we can understand)
dave
claudio said:
I'm only going to be able to get you that info if i stumble on to it. I'll have the hardcopy somewhere, but it could be in any number of places. It would likely date from the late 70's tor early 80s.
The tests were completely objective (i don't recall that any listening was done)
I can tell you that at the time i was building out of HDF and was quite disturbed by how poorly MDF did (much as John was, i imagine, when i declared my current position on use of MDF)
dave
planet10 said:
Claudio: Thanks for the post on that study. Exactly what I'm looking for.🙂
Dave: All of my measurements are useful. If there is a disturbing sound coming from the wood or the cone, the mic will "hear" it.
Honestly, you act as if I fudged the results to suit my views.
You say you didn't expect much from FR, but I think you did - it's a revealing factor. Would you buy a new driver without looking at it's published FR?