Yes seems right but wouldn't write of TG9FD-08 in long run lets see. TC9FD think is much stable in Z-curve measurement and measured impulse responses under different conditions from free air to good or bad boxes it is served it just stays stable in those measurements and sound seems staying neutral/transparent signature too, admit had only 10F for a week but it seems very sensitive to above and mayby its too alive to be real neutral or transparent. This post is just subjective tech talkbut i sence TC9FD a bit more neutral/transparent than 10F so far but 10F a lot more entertaining like turning on a digital enhancement filter in JRiver, it could be the looking for "eyebrows" and the real right stuff but it could also be a coloring to original track, think needs more investigation the three between therefor TG9FD-08 is on order especially because it has a much better cost than 10F.
Thanks link the pdf papers.
Either am i if we talk 50 times 10F's TG9FD-08 is a lot more eatable.
I won't write off the TC9, don't worry
. You might be absolutely right if you're saying the TC9 is the more natural sounding one. 50 times the 10F might really be over the top!I'm sure one could tame the 10F somewhat in the right enclosure. xrk971's enclosure seems to do that job very well.
I think it boils down to a preferred sound of materials, paper or glass fibre.
If we'd compare it to image enhancement tools the TC9 could be the softening filter while 10F would be a sharpening tool. You don't want to sharpen an already sharp picture
. But the same can be said about a soft picture... don't soften it up any further. I'll just keep using my FIR sharpening tool for that seems to work quite well on 50 drivers.In my Car I tried a Aramid cone mid bass and a Paper one. The Paper one stayed. But the Aramid coned driver was a lot of fun! But it was way more picky on types of music than the Paper one. In the end I got the Paper coned driver just where I liked it. I spend at least an equal time on taming the Aramid cone. Never succeeded entirely.
That stuck in my head when I was deciding between the TC9 and TG9 4 years ago.
The question I asked myself: "Do you want analytic sound or entertaining" I chose the last one. But the 10F might just have all the right stuff to be a bit of both.
It does everything the TC9 does, but it does it a little better. Still it's entertaining that at 95 dB / 1 meter the TC9 was better in distortion graph's than the 10F at 200 Hz.
I did not measure it on it's own at that time. As I had it in my car door. My aramid woofer was a relative of the Pioneer PRS line. Though my acual speaker was the cheaper sister of that one.
These measurement swayed my choice to a paper cone: Car Audio | DiyMobileAudio.com | Car Stereo Forum
The PRS is in there as well, my Pioneer was the cheaper sister of that from the TS-E170Ci set. I went with the Herz Mille mid basses (soon after this they changed them, cheaper made and worse than the 2008 model) and sourced high end € 180,00 Herz tweeters (I got them cheaper though) to match with it.
I ended up replacing the high dollar tweeters with Vifa XT25 SC90 (€ 20,00) and placed those in my own design 3D printed tweeter cups and still run them to great satisfaction.
The tweeter does have sort of an impedance correction passive circuit on it even though I run the whole setup active.
That whole journey can be read here: http://www.diymobileaudio.com/forum/build-logs-project-install-gallery/86635-1982-porsche-911-first-build-simple-system.html
These measurement swayed my choice to a paper cone: Car Audio | DiyMobileAudio.com | Car Stereo Forum
The PRS is in there as well, my Pioneer was the cheaper sister of that from the TS-E170Ci set. I went with the Herz Mille mid basses (soon after this they changed them, cheaper made and worse than the 2008 model) and sourced high end € 180,00 Herz tweeters (I got them cheaper though) to match with it.
I ended up replacing the high dollar tweeters with Vifa XT25 SC90 (€ 20,00) and placed those in my own design 3D printed tweeter cups and still run them to great satisfaction.
The tweeter does have sort of an impedance correction passive circuit on it even though I run the whole setup active.
That whole journey can be read here: http://www.diymobileaudio.com/forum/build-logs-project-install-gallery/86635-1982-porsche-911-first-build-simple-system.html
Heil AMT then go home .....
No a joke but can only speak subjective so far but three times tried it and was audio able improvement. I actual sit writing PM to wesayso regarding the trail network that soon should ship to him so we can get more critical ears on than mine and make documentation and share if it works in Netherlands too. If you serious i know you a lot faster
look forward feedback.
Regarding the networks you could also lean back participate in debate and enjoy so far not having the cost.
In meantime if missing more speaker tasks setup the the two winners from round one and two against HEIL AMT so they can learn it. Measurements and sound clips would not be bad to look into those three between.
In meantime if missing more speaker tasks setup the the two winners from round one and two against HEIL AMT so they can learn it. Measurements and sound clips would not be bad to look into those three between
.
The TC9FD vs 10F vs Heil AMT all with RS225-8 serving bass duties would be an interesting showdown for sure. In this case XO at 700Hz so all 3 can have same XO point. TC9FD and 10F and even TG9FD all have same bolt hole and cutout diameter so should be easy. Or keep XO for cone drivers at 350Hz as set presently (600Hz acoustic) and set Heil at 600Hz acoustic.
Pick new tracks to test. This could be really fun. Although not sure how fair it is to put a $12, $22, and $100 driver vs a world class $350 driver.
Still want to do the impedance flattening though.
Pick new tracks to test. This could be really fun. Although not sure how fair it is to put a $12, $22, and $100 driver vs a world class $350 driver.
Still want to do the impedance flattening though.
Last edited:
Hey xrk971 and BYRTT, great work on the build! I'm enjoying following along.
X, I like how you experiement with materials and formats and post data and sound clips. All very helpful!
Thought I'd provide some feedback (with silly quasiaudiophool descriptions) regarding what I'm hearing since you go to such effort being transparent and sharing info.
I listened to sound clips on two systems (stereo two-way cardidod desktop build using Tang BAnd W3/Acoustic Elegance TD6H (still undergoing development), stereo two-way econowavesque B&C DE250/12"PA woofer build) and headphones (AKG Q701 driven by an Audio-GD dac/headphone amp).
Re: my listening preferences, I tend to prefer a more taylored 'hammocked' slope cutting frequencies from 120 - 500Hz (similar to BYRTT's last plot in post #150) and rolling off the highs.
From sounc clips - post #119:
XO A: Sounds warm and alluring. The way her voice hangs is alluring. However, the upper midrange and treble (say 125 - 2000Hz) sounds diffused and incoherent in parts. High hats don't sound right.
XO B: Her voice sounds the smoothest/most natural. Diffused upper midrange and treble. High hats don't sound right.
XO C The drums sound the most natural (midbass waack on what sounds like a congo and a brush on a tom). Diffused upper midrange and treble. High hats don't sound right.
Re: 3 way with Heil AMT sound clip post #153:
Sounds like there is too much midrange, lower-treble energy. Lacking low end.
Someone mentioned her voice sounding hard and I agree.
Also, for all four sound clips, lower midrange seems overly emphasised. I suspect the enclousre is having an effect but it was most noticeable in the 3 way.
For me, the previous two-way examples sound smoother, but the midrange and treble needs.
Test tracks (that I own and am familiar with):
Norah Jones - come away with me
Dire Straits - Sultans of Swing
Couple of queries:
What source are you using (eg PC, DAC)?
How would you go about winding your own coils?
Do you/can you post polar and off-axis responses (sorry if I missed it)?
Again, thank you two for sharing your work and DIY adventures.
X, I like how you experiement with materials and formats and post data and sound clips. All very helpful!
Thought I'd provide some feedback (with silly quasiaudiophool descriptions) regarding what I'm hearing since you go to such effort being transparent and sharing info.
I listened to sound clips on two systems (stereo two-way cardidod desktop build using Tang BAnd W3/Acoustic Elegance TD6H (still undergoing development), stereo two-way econowavesque B&C DE250/12"PA woofer build) and headphones (AKG Q701 driven by an Audio-GD dac/headphone amp).
Re: my listening preferences, I tend to prefer a more taylored 'hammocked' slope cutting frequencies from 120 - 500Hz (similar to BYRTT's last plot in post #150) and rolling off the highs.
From sounc clips - post #119:
XO A: Sounds warm and alluring. The way her voice hangs is alluring. However, the upper midrange and treble (say 125 - 2000Hz) sounds diffused and incoherent in parts. High hats don't sound right.
XO B: Her voice sounds the smoothest/most natural. Diffused upper midrange and treble. High hats don't sound right.
XO C The drums sound the most natural (midbass waack on what sounds like a congo and a brush on a tom). Diffused upper midrange and treble. High hats don't sound right.
Re: 3 way with Heil AMT sound clip post #153:
Sounds like there is too much midrange, lower-treble energy. Lacking low end.
Someone mentioned her voice sounding hard and I agree.
Also, for all four sound clips, lower midrange seems overly emphasised. I suspect the enclousre is having an effect but it was most noticeable in the 3 way.
For me, the previous two-way examples sound smoother, but the midrange and treble needs.
Test tracks (that I own and am familiar with):
Norah Jones - come away with me
Dire Straits - Sultans of Swing
Couple of queries:
What source are you using (eg PC, DAC)?
How would you go about winding your own coils?
Do you/can you post polar and off-axis responses (sorry if I missed it)?
Again, thank you two for sharing your work and DIY adventures.
Last edited:
Mrk7,
Thanks for the feedback. I have not had any time to retrace my steps to figure out which XO is which for those blind sound clips. I have been working on the time alignment and xo to get the high hats to sound better. The speaker sounds better now than it did on those previous clips. I have Norah Jones Come away with Me and Sultan of Swings. I can record those for you. I use combination of PC playing high res wav or flac files and CD player as source. NJ is wav file and Dire Straits is CD. The stuff going on at 2kHz may be an over emphasized bass leaking through due to the 1st order XO. Probably will sound cleaner as second order or even 4th order LR. But then nice clean Step Response won't be as good. It's still a work in progress. I still don't have my good class D for the woofer yet and using my old car amp (LM3xxx based) driven by a 19v laptop brick.
I still need to apply the oscon cap mod and bootstrap snubber mod to the amp and that should raise it up a notch or two. I guess I could borrow the nicely modded amps from my reference system and give that a try?
Thanks for the feedback. I have not had any time to retrace my steps to figure out which XO is which for those blind sound clips. I have been working on the time alignment and xo to get the high hats to sound better. The speaker sounds better now than it did on those previous clips. I have Norah Jones Come away with Me and Sultan of Swings. I can record those for you. I use combination of PC playing high res wav or flac files and CD player as source. NJ is wav file and Dire Straits is CD. The stuff going on at 2kHz may be an over emphasized bass leaking through due to the 1st order XO. Probably will sound cleaner as second order or even 4th order LR. But then nice clean Step Response won't be as good. It's still a work in progress. I still don't have my good class D for the woofer yet and using my old car amp (LM3xxx based) driven by a 19v laptop brick.
I still need to apply the oscon cap mod and bootstrap snubber mod to the amp and that should raise it up a notch or two. I guess I could borrow the nicely modded amps from my reference system and give that a try?
xrk971, am looking forward to hearing revisions (particularly if accompanied with design revision notes). Again, I can only provide basic feedback though.
Seeing you are set-up to measure and listen, investigating the impact of a higher order slope and compromised step-response would be a good investigation.
The impact of amplification would be another easy test. I say go for it! Generally I've found speakers (particularly woofers) like headroom.
Regarding coils, every time I use an inductor calculator to make a 'large' air core (ie 8mH) I seem to get crazy numbers. If ever you make a large value homebrewed coil please share!
Seeing you are set-up to measure and listen, investigating the impact of a higher order slope and compromised step-response would be a good investigation.
The impact of amplification would be another easy test. I say go for it! Generally I've found speakers (particularly woofers) like headroom.
Regarding coils, every time I use an inductor calculator to make a 'large' air core (ie 8mH) I seem to get crazy numbers. If ever you make a large value homebrewed coil please share!
I was playing with 24dB/oct Linkwitz Riley last night, I think there may be a cleaner sound from standpoint of less of a confused sound on complex orchestral music, but then when I used it for small venue live mic'd performances, the impact of percussion was not as good - it lost the immediacy and coherency of the mid bass and percussive leading edge. It is very subtle, and if you never heard what the perfect step response sounded like, you would think it sounds pretty good. I think most speakers sound like this, but when you don't know what you don't know, you are happy. Ignorance is bliss
However, if you have heard what the speaker sounds like with flat linear phase and perfect step response, you know it's missing. It is a tough tradeoff - there may be tricks to play with and I will keep exploring and when I find one that sounds great, I will report back.Re rolling your own coils: for big (anything > 2mH) forget about air core - get some iron or ferrite cores from old transformers, a bundle of nails, etc. They make a huge difference and regardless of what people say about cored inductors not sounding as good as air core, you aren't going to hear a difference at the lower currents and for bass frequencies they are fine. I say this because class D amps use ferrite core inductors to block the 400kHz carrier frequency to let the audio band through and they sound fine unless you really drive so much current that they heat up and saturate.
Hi xrk971,
Sorry if you have already mentioned elsewhere, but what would speakers are you referring to which have these attributes?
By trade-off do you mean that such a speaker would not fare as well on complex orchestral music?
Also curious, have you played any metal on these ref monitors?
The subject of this thread, my 10F/RS225 FAST Ref Monitors
is one. Wesayso provided a link to a Stereophile review of Step Response measurement earlier in this thread (Measuring Loudspeakers, Part Two Page 2 | Stereophile.com). The reviewer said that he has perhaps only tested 10 speakers in his career that achieve a step response that looks similar to what I have been able to get here. Examples cited include: Quad, Thiel, Dunlavy, Spica, and Vandersteen.Here is the "perfect" example cited by Stereophile:
Here is my latest Step Response that gives the speaker the magical sound for percussion and immediate impact:
After having built and measured many of my speakers, I now realize that this is perhaps the most difficult measurement for a speaker and sets a great speaker apart from a good speaker. You can get flat frequency response, clean looking impulse response, and even decent looking absolute phase, but they may not produce a SR like this unless you work it hard. In fact, unless you specifically target this property and massage all the little variables/tips/tricks up your sleeve - it is very elusive. Try measuring some speakers, even very good ones and look at the SR.
Actually, it sounds VERY good - I have listened to several pieces and this speaker shines for violin concertos and flute/recorder concertos as the 10F perfectly plays the ranges of those instruments with super clear fluidity. Piano concertos also sound excellent as the step response allows the percussive strikes to be realistically played back. Chopin piano concertos are very good. I think on huge grand symphonies is where it may not do as well (but then what speaker does except a good 3 or 4-way with sub).
Also curious, have you played any metal on these ref monitors?
Yes, AC/DC and Metallica - the guitars on Metallica's "Fade to Black" are mesmerizing...
Attachments
Last edited:
Thanks for the explanation xrk971.
Opening guitars on "Fade to Black" - that's acoustic man... What about the rest of the song or some "Call of Ktulu"?
Wesayso,
He, he... you keep on enjoying the music - what's on the playlist today?
p.s. Wonder what "War Ensemble" would sound like on your Two Towers
Opening guitars on "Fade to Black" - that's acoustic man...
What about the rest of the song or some "Call of Ktulu"? Wesayso,
He, he... you keep on enjoying the music - what's on the playlist today?
p.s. Wonder what "War Ensemble" would sound like on your Two Towers
Last edited:
I started reading your thread because I was interested in impressions/measurements of the the TG9FD10. BTW, is yours the 4 or 8 ohm versions?
Anyways... I've read through this whole thread and I think you have a very nice speaker there. Those are some high performing drivers so it's good to see that you are continuing to experiment and measure because they should be phenomenal when you get it nailed down.
If you don't mind a little help then I have a few ideas for you to help get it refined further (Some of this was suggested by 5th element earlier and you really should take his advice as he knows what he's talking about)
First of all, you really should try to hit your specific acoustic slopes. The minidsp is just dealing with the input/output signal. It's up to you to deal with the drivers actual response. The areas that deviate from the standard slope are messing with your frequency response and they should be dealt with on driver side eq, not global. Also, your xo should be exactly 6db down at the xo. No less, no more unless you're using butterworth.
I can't remember how to do it off the top of my head, but you can import standard slopes, ie. LR12 etc. into REW and then adjust your xo/driver eq and measure until you hit your desired slopes. That's how xo's are typically designed and how I do it, both digital and passive. If you have the interest and time, I recommend download passive crossover designer or something similar and spend some time playing with it. There is a learning curve and it'll seem complicated at first, but it'll help you understand what's going on. BTW, PCD has an active section too so you can model your dsp setting to hit slopes there too.
Secondly, you're using the RS225 too high. A 750 xo with 1st or 2nd order is just not going to work that great (2nd order maybe if you deal with the breakup) . Regardless of what your distortion measurements are telling you (you need to raise the level as stated before because at .71V you're not really getting any useful information in you distortion sweeps, you're mostly just seeing the noise floor)
I know you think it sounds great, but that's mainly because you likely haven't heard it sound the way it should. That's not a critisism, I used to think all my designs sounded great when I first set them up too. When people are saying that the female vocals sounds hard, that is likely due to the woofer's breakup. In the graph above, the breakup is only about 20db down. You're hearing that, and believe me, you don't want to hear the breakup on those dayton drivers. It should be at least 40db down, preferably 50.
I'd move the xo back down to around your baffle step frequency (450Hz?) and HIT YOUR SLOPES so that the crossover is actually where you expect it to be and the individual driver responses are 6db down where they cross. If you need to, use the peq function to add a notch or two at the woofers breakup so that it's at least 40db below the summed. And I would stick with 2nd order or higher. With 1st order slopes you are going to be introducing distortion at both ends (low on the 10F, high on the RS225)
Third, you have some phase issues going on. Anywhere where the individual driver's response is above the summed response you have phase cancellation. Looks like pretty much from 70-5K there are phase issues which are going to affect the frequency response and imaging. The drivers response should not go above the summed and you should have more gain around the crossover in the summed response than your measurements show.
Reverse the polarity of the 10F and measure. you should see a dip in the summed response at the xo. My guess is that you will see a minor dip at the xo and the response actually raising from 70-350 and around 2k. That's not what you want.
To make it right, with the 10F reversed and both drivers connected, measure and adjust the delay of the 10F only until you get a suckout of at least 20db or so. A second order LR xo should have a wide and deep notch, 4th order a narrow deep notch. When you get a deep notch, switch the polarity back to where you had it and measure. Hopefully the summed response is no longer below the individual responses.
Now, you'll likely see a hump at the xo because your current xo is not 6db down like it should be. To fix this either hit the correct target slopes as outlined above (recommended) or move your crossover points apart, ie 350Hz on the woofer and 450Hz on the mid.
Lastly, if it was me I would build a good solid enclosure out of wood or another solid material. The box contributes more than you think to the sound and will smear the image if it's resonating. Taking an impedance sweep should show you any problem areas.
This is easy and cheap to make: Impedance Measurement
I would also probably make the "dagger" enclosure extend to the back of the box and then vent outside with the point not quite closed. Stuff this lightly to get an aperiodic vent. This should give you a more open sound and will damp the 10F resonant peak if it's stuffed right.
Good luck!