'Common'? Frankly, on many forums or even looking at consumer sales, it seems the exact opposite to me. I don't understand how so many folks world-wide can tolerate the high levels of distortion I hear in many consumer and local DIY speakers.
The only place where it's obviously 'common' is in high end consumer or serious DIY HT builds. Anyway, many rooms don't have a lot of room gain nor tonally correct LF signals on many recordings, so some of us boost the lowest lows to compensate.
Regardless, drivers that sound a bit 'thin' through the mids can benefit from a slightly boosted mid-bass and down low where our hearing acuity has fallen off, boosting the bottom octave can give the entire spectrum a fuller, more extended presentation even though we're programmed to mentally fill it in somewhat when completely missing.
GM
The only place where it's obviously 'common' is in high end consumer or serious DIY HT builds. Anyway, many rooms don't have a lot of room gain nor tonally correct LF signals on many recordings, so some of us boost the lowest lows to compensate.
Regardless, drivers that sound a bit 'thin' through the mids can benefit from a slightly boosted mid-bass and down low where our hearing acuity has fallen off, boosting the bottom octave can give the entire spectrum a fuller, more extended presentation even though we're programmed to mentally fill it in somewhat when completely missing.
GM
Just a little observation - the drivers from the video where not Fostex Fe103En, they where Fe103e, slightly different specs, but almost identical, the main difference is that they are flat in the frequency response, no "valley" between 50 and 100hz. The Xmax is terribly small - 0.35mm.
http://www.madisound.com/pdf/fostexdrivers/fe103erev.pdf
Update: My Listening Room - YouTube - at 0:23
My Other Listening Room - YouTube - he said (at 3:50) they are 'en's but they probably aren't since hexibase said he used the same drivers.
Oh and one more observation, the lowest sound I found in the recording of what the Fostexes where playing (my other listening room video) was at 41hz, loud and clear (at 3:07-3:08) and very hi-fi (I compared it with the original track), while a 37hz tone was much weaker, so the critical rolloff point is around 40hz (in that room, and listening position).
http://www.madisound.com/pdf/fostexdrivers/fe103erev.pdf
Update: My Listening Room - YouTube - at 0:23
My Other Listening Room - YouTube - he said (at 3:50) they are 'en's but they probably aren't since hexibase said he used the same drivers.
Oh and one more observation, the lowest sound I found in the recording of what the Fostexes where playing (my other listening room video) was at 41hz, loud and clear (at 3:07-3:08) and very hi-fi (I compared it with the original track), while a 37hz tone was much weaker, so the critical rolloff point is around 40hz (in that room, and listening position).
Last edited:
2 days ago Hexibase confirmed on the youtube comments section for the video (My Other Listening Room - YouTube) that the on axis -3db range is "~38Hz-21.1Khz on-axis".
However I would like to add that the 38hz extension is related to the placement of the enclosures in room and listening position.
However I would like to add that the 38hz extension is related to the placement of the enclosures in room and listening position.
That's remarkable. Given that the FE103En deviates rather more than that in Fostex's own published FR (taken in an anechoic chamber, on an IEC standard baffle, with something in the order of £20,000 of test equipment, and a modest degree of smoothing applied), we must be witnessing a cabinet that re-writes the laws of physics. 
Last edited:
I'm saying that Fostex's own published FR graph shows that the FE103En is ~10dB down at 20KHz relative to where it was at 15KHz, and about -6dB relative to the nominal SPL figure. So I can't quite see how this speaker can be -3dB at 21.1KHz, unless the driver is peaking above 20KHz, or it has some Eq applied.
Re the bottom end, while you can certainly drop the 103 into a pipe tuned to give you -3dB at 38Hz, I wouldn't care to run it like that myself. The 103 is a nice unit for what it is, but it will have about zero headroom when tuned that low. YMMV of course, and if you like the results, that's what counts.
Re the bottom end, while you can certainly drop the 103 into a pipe tuned to give you -3dB at 38Hz, I wouldn't care to run it like that myself. The 103 is a nice unit for what it is, but it will have about zero headroom when tuned that low. YMMV of course, and if you like the results, that's what counts.
Last edited:
Krakatoa, Thanks for the these measurements. I will attempt to build a pair of enclosure like Jameslc did. Just out of curiosity, how do you think Hexibase came to a cross section area of ~123cm2 for a Sd of .005m?
Re
A pipe with a section of 2.5Sd is not something out of the ordinary; here is how I believe it was done: hexibase wanted to tune as low as possible, so he made an assessment for the room, driver and position of the enclosure for what works best (hexibase always tries multiple designs, he said that it almost ended something completely different) and he came up with ~40hz extension. The corresponding enclosure being a simple pipe with constant S was determined in relation to 40hz (probably tuned slightly higher) - the length. The section was increased until it was big enough not to compromise the strength of the pipe resonance, and not strong enough to stand out. I came up with approximately the same numbers when I tried to tune that low for the given Fostex (in Hornresp) so the design is not out of the ordinary in any way, it is simply a straight pipe (more suitable for a low Qts) tuned really low for that particular driver. The only trick I don't know for sure what it means is the mouth of the pipe up there, maybe it was done to avoid booming...
In other words anybody who tunes that low (for this driver) ends up with the same design.
But! if you look at this build, also designed by hexibase for the same Fostex - Ok, I lied. Here are a few pics of my “almost” most recent build | MiltonBilt - you get a sense of... how low is that tuned, 20hz? that enclosure is larger that the one in the video even though it's a contracting geometry which are always smaller that the ones with constant or expanding geometry, I don't know the secrets of that enclosure but there are some unexplained things about it. I made an experiment recently to try and explain the bigger is better principle, and it may actually be true, in the sense that simulations look worse than they sound for unnecessary large enclosures - as you make them bigger they sound bigger and nulls in the frequency response become less of an issue. What I'm saying is that there is a slight chance that the enclosure in the video is by coincidence the standard deal, the design method being more complicated.
A pipe with a section of 2.5Sd is not something out of the ordinary; here is how I believe it was done: hexibase wanted to tune as low as possible, so he made an assessment for the room, driver and position of the enclosure for what works best (hexibase always tries multiple designs, he said that it almost ended something completely different) and he came up with ~40hz extension. The corresponding enclosure being a simple pipe with constant S was determined in relation to 40hz (probably tuned slightly higher) - the length. The section was increased until it was big enough not to compromise the strength of the pipe resonance, and not strong enough to stand out. I came up with approximately the same numbers when I tried to tune that low for the given Fostex (in Hornresp) so the design is not out of the ordinary in any way, it is simply a straight pipe (more suitable for a low Qts) tuned really low for that particular driver. The only trick I don't know for sure what it means is the mouth of the pipe up there, maybe it was done to avoid booming...
In other words anybody who tunes that low (for this driver) ends up with the same design.
But! if you look at this build, also designed by hexibase for the same Fostex - Ok, I lied. Here are a few pics of my “almost” most recent build | MiltonBilt - you get a sense of... how low is that tuned, 20hz? that enclosure is larger that the one in the video even though it's a contracting geometry which are always smaller that the ones with constant or expanding geometry, I don't know the secrets of that enclosure but there are some unexplained things about it. I made an experiment recently to try and explain the bigger is better principle, and it may actually be true, in the sense that simulations look worse than they sound for unnecessary large enclosures - as you make them bigger they sound bigger and nulls in the frequency response become less of an issue. What I'm saying is that there is a slight chance that the enclosure in the video is by coincidence the standard deal, the design method being more complicated.
And one last thing before you or anybody else builds one of these towers - if you are looking for the deep sound try another driver with a Qts higher than 0.37, at least; the MarkAudio CHR70 is going to make the Fostex seem like it's fasting (and training for the olympics of fasting). Please look at the simulations I posted at post 52 (page 6) - the Fe103en is not exactly the ideal candidate for deep bass. Look for something with a Qts of 0.4-0.55, you will get much more bass output...
Hope you post some pics with your towers too.
Hope you post some pics with your towers too.
FE103 family has always hit a sweet spot in the mids, but definitely isn't a bass monster - CSS EL70 would be another candidate for ~3" driver on steroids or at least HGH.
Different drivers
Some simulation I've done with 9 different drivers in Hornresp - the nominal SPL is the one calculated by the software not the one from the specs; all are done for the same enclosure. The general rules are these: the best drivers are those with Qts of 0.4-0.55 and Fs bellow 80hz; Vas gives you the volume of the enclosure (how big the section of the pipe must be). Large Vas can be a problem, for example TangBand W4 930SF needs an extra 10 liters to be ok because of the higher Vas.
MarkAudio CHR70 has the strongest bass and overall best behavior off all these drivers - needs a small enclosure, is cheap and good bass/Xmax, good reviews in general.
I would be curious to see how CHR70 looks in MJKws simulation for this enclosure.
Some simulation I've done with 9 different drivers in Hornresp - the nominal SPL is the one calculated by the software not the one from the specs; all are done for the same enclosure. The general rules are these: the best drivers are those with Qts of 0.4-0.55 and Fs bellow 80hz; Vas gives you the volume of the enclosure (how big the section of the pipe must be). Large Vas can be a problem, for example TangBand W4 930SF needs an extra 10 liters to be ok because of the higher Vas.
MarkAudio CHR70 has the strongest bass and overall best behavior off all these drivers - needs a small enclosure, is cheap and good bass/Xmax, good reviews in general.
I would be curious to see how CHR70 looks in MJKws simulation for this enclosure.
Attachments
Re
I forgot to say that the enclosure I used for those sims needs some adaptation for each driver, the purpose was to show how big the difference can be between one driver and another and that Fe103en is perhaps the worst choice among drivers of it's category (price, size, quality). And another thing - the tremendous nuance of the sound in the video may be related to room modes around 50hz or some king of distortion (working under fs?) because there seem to be other sound in there too (just reverberation?), but the bass is stronger around 50hz than in the original recording.
Fe103en - Fs=83, Qts=0.33, Vas=5.95, Xmax=0.6
FF105wk - Fs=75, Qts=0.41, Vas=4.8, Xmax=1.7
Fe103e - Fs=79.5, Qts=0.36, Vas=6.9, Xmax=0.35
CHR70 - Fs=65.4, Qts=0.55, Vas=5.17, Xmax=4
Alpair7 (gen3) - Fs=70.9, Qts=0.54, Vas=4.58, Xmax=4
EL70 - Fs=64, Qts=0.55, Vas=6.1, Xmax=4.5
W4 1320SJ - Fs=75, Qts=0.35, Vas=6.41, Xmax=3
W4 1337SD - Fs=70, Qts=0.37, Vas=4.86, Xmax=3
W4 1052SD - Fs=70, Qts=0.44, Vas=7.58, Xmax=3.4
I forgot to say that the enclosure I used for those sims needs some adaptation for each driver, the purpose was to show how big the difference can be between one driver and another and that Fe103en is perhaps the worst choice among drivers of it's category (price, size, quality). And another thing - the tremendous nuance of the sound in the video may be related to room modes around 50hz or some king of distortion (working under fs?) because there seem to be other sound in there too (just reverberation?), but the bass is stronger around 50hz than in the original recording.
Fe103en - Fs=83, Qts=0.33, Vas=5.95, Xmax=0.6
FF105wk - Fs=75, Qts=0.41, Vas=4.8, Xmax=1.7
Fe103e - Fs=79.5, Qts=0.36, Vas=6.9, Xmax=0.35
CHR70 - Fs=65.4, Qts=0.55, Vas=5.17, Xmax=4
Alpair7 (gen3) - Fs=70.9, Qts=0.54, Vas=4.58, Xmax=4
EL70 - Fs=64, Qts=0.55, Vas=6.1, Xmax=4.5
W4 1320SJ - Fs=75, Qts=0.35, Vas=6.41, Xmax=3
W4 1337SD - Fs=70, Qts=0.37, Vas=4.86, Xmax=3
W4 1052SD - Fs=70, Qts=0.44, Vas=7.58, Xmax=3.4
Hi all good to see that people are still thinking about this design after reading through Krakatoa and others comment i think i might also upgrade the driver to a EL70 because of the higher xmax then the fostex to see if i can increase the bass with the same size driver and enclosure but i have just found out that i am going to need a bigger hole in the baffle to fit the el70 and that means a bigger T line size
Last edited:
Sim. observations of one posted scree. (of 9).
You are saying that Xmax will prevent reaching that SPL? I don't think it's that bad - Xmax is the maximum excursion available before you start having compression of the dynamics, in other words they will start to become more compressed, less powerful, but I don't think it's much off an issue, Fostexes have been known to be tolerant with this; xmech is much larger. It's not the best design in the world, you have horns for low xmax drivers, and these Fostexes aren't good at all for this design; however you can take a look at the video that started the thread and hear them in action - My Other Listening Room - YouTube
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.