Nice mic that goes to 40kHz!
What are measurement conditions? Voltage drive level, distance, and are SPL
Levels calibrated absolute?
None of these measurements are on horns. Both brytts and mine are naked drivers as per pics. Didn't want any confounding factors!
Think its mic difference also because had read elsewhere that UMIK-1 noise floor kicks in at HF area at distortion tests and in mine measurements it looks like difference starts somewhere in 1,5-3kHz area. Think UMIK-1 is fine mic as seen in below FR comparison at its locked native 48kHz verse EMX-7150 at 96kHz, black trace is a textbook stopband BW2 20kHz. My UMIK-1 is not pro calibrated as the one xrk971 have from Cross Spectrum Labs, mine is ab fabric probably mass production calibrated, but the new EMX-7150 is pro calibrated at iSEMcon and their own brand. Biggest reason got the new EMX-7150 is because its so flat that can run in setups that don't offer to implement correction file and because its analog can sync I/O clock in REW to get time of flight in sync.
Attachments
I do know that when a resonance that can be corrected via eq is fixed that the time domain issues seen in a csd are also corrected in a fixed gate plot. Is this also the case when the same resonance (whose detail is obscured by the fdw as done in rew) is fixed using that fdw plot?
As far as your 3khz example if it's coming from the loudspeaker and it's not a room reflection then that's something that I want to be aware of.
Maybe what we're really discussing is REW's implementation of the fdw rather than the fdw itself.
I agree that in theory the fdw would be great but like you I don't think it's quite getting it done in rew. Afaik it does use a sliding window like you describe so I wonder what's different between what rew does and what drc does like wesayo's comparison showed.
Of course gating has limitations.....like anything else. With the fdw lower in freq if you do have issues then you need to determine if it's the speaker or the room.
For now I'll keep dragging my speakers outside [emoji6]
We are discussing exactly that, and that's why I posted my comparison to warn for REW's overly smoothed results. My original proposal to use REW's FDW was to export a minimum phase IR from REW, re-import that IR and use the FDW to base EQ decisions on that minimum phase version of the original IR.
After seeing how course/smoothed the FDW in REW really is I do not find its current implementation is good enough to do that job. Which is why I posted it here (and on my own thread). But that doesn't mean a better FDW implementation wont be implemented in time.
But you still need the step of exporting the minimum phase version of the IR and re-import that IR to do it in REW, once they implement a better version. The FDW I showed from DRC is a minimum phase "slice" out of the IR.
The FDW in DRC is a step less process. I doubt the current implementation in REW is done in the same manner. See the graph I posted. REW keeps evolving, in time we might see a more useful implementation. As I do not use REW like that, but promoted using it in the above manner, I figured to post about it, right after I had seen evidence in a measurement that showed the differences are just too big. I'm using the measurement part of REW, but DRC is used for EQ in my case. My own research in REW was done prior to the addition of the frequency dependant windows.
Gmad had done experimentations and reported on that here which is what triggered me to post that info/method on this thread. That's all really. I'm still very fond of using FDW's, just not REW's current version.
Last edited:
Seems a nice grade mic that is not too costly for diy/hobby use.
For measurement conditions original request was : )
Then made as in below first picture SB65 mounting interface that was placed on a second microhone stand elevated to middle of room from all boundaries. A Behringer set of DEQ2496/ECM8000 dialed in SPL at 90dB at 30 cm distance, pink noise or whatever it is REW send in "Check levels" menus. Did not measure voltage drive level at the time, so just did it. My non fluke DMM claim true RMS but limited to 45-500Hz and when connected to 90dB SPL REW pink noise it fluctuates around 1,9-2,4 volts. Then started REW "Generator" at 400Hz sinus and that reads 2,329 volt and DATS device read driver is 3,869 ohm at 400Hz, calculator at www.sengpielaudio.com then outputs:
Attachments
Thanks for the clarification Byrtt. You did not need to go through all the calcs and extra work - but Iappreciate you giving so many details. I am glad I got a Fluke multimeter as my cheap ones never had enough resolution to say 2.83v. Fluke actually says 2.830v (3 digits after decimal, whereas previous DMM's had 1 digit to the right).
You guys are spoiled by technology. Anybody old enough to remember what "VTVM" stands for or to have owned one?
My first volt meter was all analog with a galvanometer needle (with a mirror strip to reduce parallax error). It did not have vacuum tubes though. My first computer had 4K of RAM and used audio cassettes to store data. 🙂
Request and work is okay, its probably good value to know the info before real use of the data : )
Fluke is HQ and is a wish but i trust also this more than €100 DMM, it has also 3 digits after decimal and inside specs at 400Hz it was 100% steady at 2,329 volt.
Agree 100% 😀
I remember only being able to afford a VOM but lusting after a VTVM. I eventually got a used probably WW2 surplus VTVM and like yours, it had a beautiful mechanism, long since obsolete - not just the meter but the rotary switch and discrete resistor ladder.
Back in the day, I designed computers like the one you described and built some of the prototypes in my basement, starting with the Intel 8080.
Back in the day, I designed computers like the one you described and built some of the prototypes in my basement, starting with the Intel 8080.
Z-80 all the way.
Those are still used as the micro controllers on a bunch of Stanford Research Systems instruments (lock in amp, picosecond delay generators, etc).
Those are still used as the micro controllers on a bunch of Stanford Research Systems instruments (lock in amp, picosecond delay generators, etc).
Would this work with a jbl 2432 compression driver in a 2384 horn?
What would you recommend for the mid woofer? Would Dual 8in be sufficient ? How do you calculate the location of the woofer injection ports? As near as possible to the throat?
Sent from my GT-I9505 using Tapatalk
What would you recommend for the mid woofer? Would Dual 8in be sufficient ? How do you calculate the location of the woofer injection ports? As near as possible to the throat?
Sent from my GT-I9505 using Tapatalk
I only have a UMIK-1 with factory calibration, so it's nice to know what it's limits are.
I loved the z80. If you want one today, you can get the IP and implement it an FPGA. Its only 5K gates or so.
My first assembler was on a Z80. I had real trouble getting my head around 6502 after that.
The chap I work with is a Z80 nut. I have to say I have no idea how he codes the thing in ASM but directly via Hex.
Nuts.
I can use it PIC devices, but my C is not great (nor my ASM tbh)
I have some Z8 uC about, but I haven't even scratched , surface and looked for a compiler/programmer yet.
Once I've got my head around VB6 then I'll have another cracking at a serious application on a uC of some kind.
Nuts.
I can use it PIC devices, but my C is not great (nor my ASM tbh)
I have some Z8 uC about, but I haven't even scratched , surface and looked for a compiler/programmer yet.
Once I've got my head around VB6 then I'll have another cracking at a serious application on a uC of some kind.
When I started my degree, we did a module on PIC. It was pretty cool, and my slightly naff code still did 90% of the job haha
Arduino was just gaining momentum in Uni then, filtering down from the Masters thesis classes etc.
Tried a Pi. Didn't get very far. It doesn't help that there's a bewildering choice of languages. Plus the other difficulty I had was the small i/o count.
Oddly however, I seem to be doing respectably with VB.
Arduino was just gaining momentum in Uni then, filtering down from the Masters thesis classes etc.
Tried a Pi. Didn't get very far. It doesn't help that there's a bewildering choice of languages. Plus the other difficulty I had was the small i/o count.
Oddly however, I seem to be doing respectably with VB.