..I should have also mentioned that I want the capability to run them all (if *desired) at the same time.
*though I don't know why I'd want to run the DATS VM at the same time as the others..
It still looks "off".
I was expecting something more linear (..like Tim's measurements).
Ironically the "dip" at 2.2-2.6 doesn't bother me that much, but rather that it doesn't "recover" to nearly the *same spl by 3.3 kHz. I'm also expecting that the response much below 900 Hz is less accurate, so I discount that result.
To give more context to how I view it: it looks like the target average usable spl is what you see at 3 kHz on this graph (more than -2db than what I was expecting).
IF it's correct, then we will have to look to more "robust" traditional tweeter solutions OR loose more efficiency.
*narrow band dips are often perceptually inconsequential (and can even be beneficial in some circumstances), but the result you have hear is to wide-band and will easily be audible unless eq.ed or designed with a different sort of accompanying tweeter.
Maybe you will laugh, but probably not...I really feel like an idiot on this one
Noticed that I just remembered that I had the Equalizer APO when I did the test the first time, well guess that I forgot I also had it while I was doing that Calibration!
I don't know if that will help much the Fostex, but it's certainly will get a more accurate response
Really I need to make my own checklist to verify everything, I can't continue on doing a quick test and forgetting about what I have to do.
The calibration match my equalize correction
How goes the saying again ? Practice makes perfect
One of things I want is related to measuring loudspeakers.
I specifically want a few "fixed" (not updated) OS's with software.
1 VM (Windows): DATS (pass-through to that USB).
1 VM (Linux): REW (pass-through to another USB to another audio adapter).
1 VM (Windows): SoundEasy (pass-through to my PCI-based Echo Layla 3G), along with some other modeling software (probably VituixCAD)
"pass-through" in this instance means that only that particular VM "sees" the device it has assigned to it.
I've already got the hardware for this, it just a matter of remastering (the way I want it) a Linux VM server and then configuring each VM. Unfortunately I've not spent enough time, (even having started and stopped this multiple times within the past 2 years
I understand you about the unexpected changed settings that can really mess up the work you do (and time loss), especially after my misadventure with that equalizer I had, having to redo everything again because I am not isolated from my usual do all on it computer.
-Have you verified your sound devices drivers are supported on the Linux distro and for pass through on the KVM ? (I know I had often the issue with drivers)
-What is less flexible is having to pass files between VMs but it's a detail that can quickly be worked around by using a common shared space.
-I have noticed that once DATS is connected on my Windows computer the default sound device become the DATS
Congrats on the chairs, now the dipole subwoofer ?
Clean that garage at Father Day it's ok, it's only once a year
How goes the saying again ? Practice makes perfect
..it's a learning experience.
..and remember, it's never going to be "perfect": it's just a matter of striving for something as close to perfect as you can get it to "push" to modeling. ..just wait until you start trying to get good measurements from an actual loudspeaker.And yes, that's what I'd expect from Fostex's datasheet. (..more-so than Tim's measurement - which looks a little too good, but doesn't have the efficiency I'd expect.)
A lot of people would look at this and say "pass" because of the dip - but really it's in the right area and if you use a tweeter w/ a complementary "bump" then it can even result in a more linear response (depending on design and tweeter).
Last edited:
I understand you about the unexpected changed settings that can really mess up the work you do (and time loss), especially after my misadventure with that equalizer I had, having to redo everything again because I am not isolated from my usual do all on it computer.
-Have you verified your sound devices drivers are supported on the Linux distro and for pass through on the KVM ? (I know I had often the issue with drivers)
-What is less flexible is having to pass files between VMs but it's a detail that can quickly be worked around by using a common shared space.
-I have noticed that once DATS is connected on my Windows computer the default sound device become the DATS
Congrats on the chairs, now the dipole subwoofer ?
Clean that garage at Father Day it's ok, it's only once a year
-yeah, that "static" system just setup for a particular software/hardware combination can be a pretty enviable setup. It can often make the difference between completing a project and just giving-up.
-yes, I've got drivers for the Echo.. and it (auto) shows-up with my VM Server (base) OS (..but it will be on a Windows (SoundEasy) VM, and I've got drivers for that.) On pass-through it isn't neccesary that the VM server has drivers, it's just a matter of assigning the pci slot to the VM, but the VM DOES require the drivers for that device.
I'll have OpenMedia Vault as another VM (w/ SnapRaid).. I *could* integrate it with the VM Server, but I'd rather not. I've used UnRaid before: which effectively combines the two, but there are some things I don't like about it ..and it's not "Live" OS (..I want everything running in System Memory). (..basically really "light" systems with very little in the way of additional programs running - averaging about 500 MB for Linux OS's, though the file server OpenMedia Vault will be a LOT more). Still, I've got 64 gig of ECC at my "disposal".
Yes, DATS effectively "takes over". And sometimes when you remove it from the system your previous defaults aren't correct.
Really, best practice is to use it on an entirely different system (..or the VM with pass-through).I'm not even thinking about the dipole sub, my guess is that it and the area-floor will be a few months away for me. (..I've got the carpet tiles on order to fill-out the sample I already purchased, haven't purchased any wood yet, and still haven't gotten the low-pass "chunky" inductor I purchased.)
To give you an idea of my choices, here are the chairs I have and the carpet I ordered. (..I'm not going for much more surface area for the carpet than what the chairs require when in-place):
https://www.samsclub.com/sams/mm-pe...-recliner/prod21320849.ip?xid=plp_product_1_4
https://images.homedepot-static.com...lo-berber-carpet-tiles-4010275107-31_1000.jpg
Last edited:
..it's a learning experience.
And yes, that's what I'd expect from Fostex's datasheet. (..more-so than Tim's measurement - which looks a little too good, but doesn't have the efficiency I'd expect.)
A lot of people would look at this and say "pass" because of the dip - but really it's in the right area and if you use a tweeter w/ a complementary "bump" then it can even result in a more linear response (depending on design and tweeter).
I just want to produce relatively reliable that I (we) can use to make informed decisions on the design, looking at the curve and trying to see where they would join better is one thing but I actually listened a bit to them between the measurements I have done and what I "feel" about them is the following :
The Fostex is smooth and wide, female voices are in the right spot, but to me the top end of the Hf spectrum feel like a kind of muted and a bit dull, especially on instruments likes cymbals and triangles
The Gradient without considering the strong efficiency they have over the Fostex, they surprise me in the amount of detail they can reproduce on a very wide band, clear and warm at the same time, I wouldn't be against to use them with a tweeter that can go a bit low in a 2way, I can't know the low part without an appropriate cabinet for them.
About the Radian
I have written to TLHP about having a dipole version they wrote me they are not available in small quantity for general public
I am still waiting on Thomann to confirm if they can have them in Dipole, but they can make a special order already.
-yeah, that "static" system just setup for a particular software/hardware combination can be a pretty enviable setup. It can often make the difference between completing a project and just giving-up.
-yes, I've got drivers for the Echo.. and it (auto) shows-up with my VM Server (base) OS (..but it will be on a Windows (SoundEasy) VM, and I've got drivers for that.) On pass-through it isn't neccesary that the VM server has drivers, it's just a matter of assigning the pci slot to the VM, but the VM DOES require the drivers for that device.
I'll have OpenMedia Vault as another VM (w/ SnapRaid).. I *could* integrate it with the VM Server, but I'd rather not. I've used UnRaid before: which effectively combines the two, but there are some things I don't like about it ..and it's not "Live" OS (..I want everything running in System Memory). (..basically really "light" systems with very little in the way of additional programs running - averaging about 500 MB for Linux OS's, though the file server OpenMedia Vault will be a LOT more). Still, I've got 64 gig of ECC at my "disposal".
Yes, DATS effectively "takes over". And sometimes when you remove it from the system your previous defaults aren't correct.
I'm not even thinking about the dipole sub, my guess is that it and the area-floor will be a few months away for me. (..I've got the carpet tiles on order to fill-out the sample I already purchased, haven't purchased any wood yet, and still haven't gotten the low-pass "chunky" inductor I purchased.)
To give you an idea of my choices, here are the chairs I have and the carpet I ordered. (..I'm not going for much more surface area for the carpet than what the chairs require when in-place):
Member'''s Mark Pegasus Power Theater Recliner - Sam'''s Club
https://images.homedepot-static.com...lo-berber-carpet-tiles-4010275107-31_1000.jpg
Nice I didn't knew that you could do that with sound-cards in virtual layer, are they shared when you used them like that or you need one for each VM ?
All the VM I work with does not even have sound on them (ESXi servers)
You are well equipped computer-wise it seem
The Chairs look so comfy and with a nice design too, so is it actual animal leather at the price ? The computer chair that I use at home with a bit of leather on them is not even comfortable but they cost the same price, I'm jealous man.
That carpet is a good choice too, I like the change of pattern and tone.
Nice I didn't knew that you could do that with sound-cards in virtual layer, are they shared when you used them like that or you need one for each VM ?
All the VM I work with does not even have sound on them (ESXi servers)
You are well equipped computer-wise it seem
The Chairs look so comfy and with a nice design too, so is it actual animal leather at the price ? The computer chair that I use at home with a bit of leather on them is not even comfortable but they cost the same price, I'm jealous man.
That carpet is a good choice too, I like the change of pattern and tone.
-it just passes-through to one VM. I also have a few network cards on that system that have SRI-OV that can do that (multiple VM's), but that's the only thing (..though I can transfer from VM to VM on that machine at incredible speed with a 10 gig network card that has SRI-OV.)
The chairs are just a cheap bonded leather (like woodchips for particle board).. but with more polyurethane (30/70). It's actually a good pliable material that breaths a little bit and wears-well except for strong abrasion (like a "gouge"). It's a cheap solution that works well for at least 3 hours.
It's weird though that it doesn't feel like polyurethane or leather (..even bonded leather). I'd have greatly preferred a really good group of full-grain leather chairs (that met the HT requirements), but pricing for each starts around what I payed for all 5.The carpet reminds me of a good airport lounge floor. I've had the samples up against the chair and it's a good match. The reason I went with carpet tiles is because they are tiles and it won't be a problem using them given the need to create a multi-layer surface and the need to pass-through wires.
Last edited:
I think we will have to call that one the "Money Shot".
-yeah, it looks very good!
I'm not surprised that the Fostex doesn't provide quite the treble-response you are looking for.. it's good for what it is, and for many people it's "good enough", but not for everyone (including me).
I have just learned a new expression in English
Btw, I can notice my serious handicap in math in this hobby, If you had the courage to instruct me on a particular subject, in the document "Texas Instrument Audio Characterization Primer" you shared me :
I do not understand it, I have tried read "D. B. Keele_LF_Near-field_Measurement" but the formulae are even worst
Piston Radius : Is the diameter of the cone diaphragm with half of the suspension area divided by two, indicated here as "r" AND "a" but what do I do with that value? "Ka" What do I have to do with that? k+a, k-a, etc ?
2 x pi : is 6.283185
f/c : What does that mean ? I see the speed of sound indicated ok but why?
Very frustrating
Then i read "How to Achieve Accurate In-Room Quasi-Anechoic Free-Field Frequency Response Measurements Down to 10Hz By Jeff Bagby" but after that Is is still unclear how to know the best distance to measure from and at what frequency I should merge the two curves.
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwisw_rBu-viAhXCYVAKHdMxAi4QFjAAegQIARAC&url=http%3A%2F%2Faudio.claub.net%2Fsoftware%2FFRD_Blender%2FWhite%2520Paper%2520-%2520Accurate%2520In-Room%2520Frequency%2520Response%2520to%252010Hz.pdf&usg=AOvVaw3M67cyu_TYnjmUQECJa-5F
Last edited:
Yeah data transfer between VM in the virtual layer is great, but do you care? it's only small files no ?-it just passes-through to one VM. I also have a few network cards on that system that have SRI-OV that can do that (multiple VM's), but that's the only thing (..though I can transfer from VM to VM on that machine at incredible speed with a 10 gig network card that has SRI-OV.)
What is the advantage to run REW in Linux instead of Windows?
The chairs are just a cheap bonded leather (like woodchips for particle board).. but with more polyurethane (30/70). It's actually a good pliable material that breaths a little bit and wears-well except for strong abrasion (like a "gouge"). It's a cheap solution that works well for at least 3 hours.
The carpet reminds me of a good airport lounge floor. I've had the samples up against the chair and it's a good match. The reason I went with carpet tiles is because they are tiles and it won't be a problem using them given the need to create a multi-layer surface and the need to pass-through wires.
Oh that is a bit misleading that they can call that leather, they should use pleather instead, but you are probably a bit safer with that considering the potential spill of cream cheese and salsa
Did you made the hollow floor with removable tiles also to run the cables ?
Yeah data transfer between VM in the virtual layer is great, but do you care? it's only small files no ?
What is the advantage to run REW in Linux instead of Windows?
Oh that is a bit misleading that they can call that leather, they should use pleather instead, but you are probably a bit safer with that considering the potential spill of cream cheese and salsa
Did you made the hollow floor with removable tiles also to run the cables ?
This is not the only use for the server: in fact the primary use is file-server (OpenMediaVault + Snapraid) and more particularly a router (with some unusual characteristics where VM to VM high-speed transfer can keep things moving along with low-latency more than high/(bit-rate)-speed transfer). Other uses include loudspeaker design and even a back-up gaming machine (..if modest: games with less demanding video-processing with only a 1060 video card). Of course for the most part none of those other uses will be "up" while the file server and router are active. Basically a multi-use modular system (..though with the negative that some of those processes won't be able to support full simultaneous operation: like multiple Windows OS's - even though remastered/"gutted" to a very small (for Windows) size.
Linux can be made a LOT smaller (than any Windows OS). When running from system memory that's key!
One Windows system for Soundeasy and one system for REW simultaneously (while at least allowing for the router-system being "up").A LOT of people commented on Sam's add as being misleading (and it is), but even at Sam's "retail" price what would a reasonable consumer expect: full grain Grade "A" leather with multiple hardwood supports and dense foam + wool etc..?
Sam's Club did update the page to show the percentage (..but predictably it's not particularly "apparent"). Probably would still get punished in the EU though.. And yeah, sugary soft-drinks, buttery popcorn, chips and salsa, etc.. Though I haven't built anything yet, the floor/riser will have plenty of hollow-space for reasonably thin wiring for bass-shakers, sub-woofer, and the power to the chairs (for the power-recline). I'm not expecting to remove any of the carpet tiles for replacement (they have industrial glue on them), rather I wanted the "joint" between tiles to serve as a pass-through for the wiring (rather than cutting holes into a large piece of carpet and risking the structural integrity of the carpet). Like the chairs, the whole platform needs to "float"/be-isolated from the floor so that the bass-shakers do their thing without sinking energy to the floor, and that should leave me some room for wiring beneath it.
Last edited:
Honestly my eyes glaze over when looking at math.
I just keep to the prose:
"Near-field measurements are taken where the distance between the driver and the microphone is as small as safely possible, that is, almost touching the dust cap of the driver diaphragm.."
Of course this also depends on excursion/"stroke" for a driver relative to the test signal and measurement set-up. (..where a subwoofer, though near-field (..and lower resulting excursion), might result in excursion that would come very close to or touch the microphone - which is obviously BAD).
Last edited:
Remember, no guarantees from me!
Of course I hope that it will work-out.
-and yes, the waveguide is a good idea, not just for the loading, but also the "off-set" (..phase relationship between it and the fullrange depending on the crossover).
This is not the only use for the server: in fact the primary use is file-server (OpenMediaVault + Snapraid) and more particularly a router (with some unusual characteristics where VM to VM high-speed transfer can keep things moving along with low-latency more than high/(bit-rate)-speed transfer). Other uses include loudspeaker design and even a back-up gaming machine (..if modest: games with less demanding video-processing with only a 1060 video card). Of course for the most part none of those other uses will be "up" while the file server and router are active. Basically a multi-use modular system (..though with the negative that some of those processes won't be able to support full simultaneous operation: like multiple Windows OS's - even though remastered/"gutted" to a very small (for Windows) size.
Linux can be made a LOT smaller (than any Windows OS). When running from system memory that's key!
Well that seem to me like a smart but very complicated setup, I sure hope you don't have an hardware failure with all that roles on that server
No wonder you are running after time with all that
Ha yeah true that footprint is a lot better to get everything in memory indeed
I glimpsed at that Soundeasy, I often read stuff on It's all about sound and audio systems | Audiojudgement.com, he explain things in a way my brain does not shutdown immediately, there was a few article about that software. It look very complex, an interface from another age with as many button squeezed together as possible, i'm sure it's powerful but the name uses that software does not seem to fit
True in EU it's more regulated about advertisement, probably wouldn't pass here, If I had seen a Seat at that price in a shop in EU I wouldn't be gullible enough to believe it's top quality leather, but in US many things seems a lot cheaper with a better service than it is here, VAT 21% on most goods does not helpA LOT of people commented on Sam's add as being misleading (and it is), but even at Sam's "retail" price what would a reasonable consumer expect: full grain Grade "A" leather with multiple hardwood supports and dense foam + wool etc..?
I was interested in Emotiva Power Amplifier at some point, so I go see the website and I see the A700 (listed in EUR) at 522.11, I had a look at the listed distributor website here in Belgium and it was 899, I contacted Emotiva and they put the distributor in contact with me, they made a breakdown of all the taxes, duties, shipping and in the end they ended up with 10% gain for them.
I suppose they also have a better price than the listed price on Emotiva, but even then it's awful.
Ho ok I believe I see what you mean, so you reserve a technical space between some tiles to pass cables with some king of sheath.
That floating platform, not sure I see how that is working, with some dampening materials in between the floor and the platform ?
Honestly my eyes glaze over when looking at math.
I just keep to the prose:
"Near-field measurements are taken where the distance between the driver and the microphone is as small as safely possible, that is, almost touching the dust cap of the driver diaphragm.."
Of course this also depends on excursion/"stroke" for a driver relative to the test signal and measurement set-up. (..where a subwoofer, though near-field (..and lower resulting excursion), might result in excursion that would come very close to or touch the microphone - which is obviously BAD).
Somewhat glad I am not alone in that situation with math
But then how do you know when you where to merge the near-field to the far-field?
I have seen in a document of Kimmosto that you subtract around 20db from near-field to adjust for gain of the proximity in relation to the 1m but that only help to get them around same level.
How do you select you merging point frequency ?
Remember, no guarantees from me!
Of course I hope that it will work-out.
-and yes, the waveguide is a good idea, not just for the loading, but also the "off-set" (..phase relationship between it and the fullrange depending on the crossover).
Too late Scott I have already contacted my lawyers
Anyway I have spend some time around to evaluate all the planar that could work as bipolar and didn't seen that many in that price range.
I sure hope we can make it work, in bipolar configuration the curve should go down a bit, how many dB I don't know yet, the 95dB datasheet specs are with the Wave-guide, recommended 3khz@12dB/Oct, If I need I can still attenuate a bit but will it be enough for that Fostex in between that I'm doubting
I have planned a few days to start that IEC baffle but the structure to hold it is still unclear, I know I will use some insert bolts for wood so I can dismantle the whole thing when not in use, but don't know yet the size of the wood I will use to build the structure and the exact shape to hold all in place, would be more easy if I had a space to leave it there.
I'll probably limit the size to the 4x4 and then think how I'll integrate the rest to enlarge it if need be, Nearfield measurement should help me getting something usable (Hope so)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.