I did not find anything suspicious from Time align function of Auxiliary calculator. So distance from mic to acoustic center of woofer seems to be at least 10 mm longer than from mic to tweeter's acoustic center.
There are some disagreement between REW and VituixCAD about minimum phase response:
SB woofer, gray=minimum phase response by REW, lime=MP by VituixCAD.
An externally hosted image should be here but it was not working when we last tested it.
XT25BG
An externally hosted image should be here but it was not working when we last tested it.
This causes the following:
If you trust minimum phase calculation of VituixCAD and check both MP checkboxes, the result is 29 us=10.1 mm
An externally hosted image should be here but it was not working when we last tested it.
If you trust minimum phase calculation of REW and uncheck both MP checkboxes, the result is 48 us=16.7 mm
An externally hosted image should be here but it was not working when we last tested it.
Note that frequency range should not be too wide. I have set 630...6300 Hz to reduce weighting of top octave because it's irrelevant for the task i.e. phase matching around XO frequency.
I have not tested this with PCD. Better to keep my hands off it and avoid speculation without personal experience.
Summary: This is crap method due to unknown and subjective part of minimum phase calculation, slowness and weaknesses related to off-axis measurements. Therefore using of single channel measurements, minimum phase extraction and solving of difference between acoustic centers is not recommended in VituixCAD documentation.
My current measurement setup is PC soundcard optical out to DAC to amp and record with Umik and REW
I have not been able to figure out how to take dual channel measurements with this set.
Or more correctly, I have not been able to figure out how to modify this setup or what I need to add to take dual channel measurements.
I have ARTA, but it is not very noob friendly compared to REW. I don't know if it will do dual channel measurements with optical out to DAC and USB mic in. I have not seen how to add a second channel to that configuration.
I have no problem picking up the Dayton Audio Omnimic V2 software/mic combo if it would get me better results. I think the learning curve would be smaller than with ARTA.
Thank you,
David.
Stuff as this is over my head and don't like to argue about subject rather help with info and listen what the schooled have to say : ) can it help say it looks whenever REW work with a real IR wheter its imported one or from own sweeps then it calculates phase shall cross at point zero at half of the sample rate and therefor its phase at HF is north of VituixCAD. So in REW the higher the sample rate the closer to texbook, what i mean here is for example a 20kHz 2nd order low pass have textbook phase at exactly -90º point @20kHz but at 48kHz sample rate REW will only be around -15º because it shall reach up north to that zero crossing point at 24kHz, and if we ask REW about minimum phase then it show about -13º, then if that REW minimum phase version is exported as txt files and imported over to VituixCAD and ask for minimum phase we get phase point to be around -53º @20kHz.
Last edited:
^Response of XT25BG extends up to 40k by datasheet which means that MP response at 20k is probably much closer to 0 deg than -180 deg. So automatic slope detection of VituixCAD looks much better in this case. Some other case could be different. Measurement sweep probably ends at 20k and sample rate could be 48k so not much changes to get decent result without luck or additional manual parameters for MP extraction.
Fortunately dual channel measurement gear and mode are available.
Fortunately dual channel measurement gear and mode are available.
Sorry, it's not possible due to Umik-1 which does not have input for reference channel.
I don't agree. ARTA is recommended with VituixCAD for some good reasons. For example easiness and speed. But all benefits are not available with single channel gear.
That's also single channel USB gear i.e. not recommended.
--
Anyway, you can continue with existing single channel gear. Just remember that you cannot produce reliable measurement data for full space simulation with power and DI approximation. 0-60 degrees off-axis could be okay but not much more. Distance between acoustic centers can be double-checked later with some other method such as reverse null. Initially you can select trusted method and tool with a coin
Alternately, get a non-USB mic such as a Dayton EMM-6 and a sound card, hook them up as described in the REW help, and you will have no trouble doing 2 channel measurements. You can still go optical out to the DAC with both channels. Then use left channel out of the DAC to drive the speaker and loop the right channel DAC output back to sound card line in for a timing reference. MIC then goes to left channel input of sound card, which will need to include a preamp and bias supply.
I could not find the two channel setup instructions in REW.
Any suggestions on a mic preamp?
I am assuming I can use the Dayton EMM-6 and said preamp with ARTA as well?
Connection and measurement instructions for ARTA: Measurement Preparations.pdf. Only few settings may need change before polar measurement sequence. Time-windowing and exporting to txt/frd is done with VituixCAD which makes operating of measurement program easy and fast, and timing of exported responses stays stable.
That is not guaranteed with REW even in dual channel mode if all reposes are not exported at the same time.
That is not guaranteed with REW even in dual channel mode if all reposes are not exported at the same time.
Unfortunately this was not happened to me so I don't know all facts in details and REW revision. I understood that timing/phase was jumped with few measurements when user exported dual channel measurements individually with Export measurement as text. But everything worked fine when he exported with Export all measurements as text. The last one is proper way to export multiple off-axis measurements of the same sequence anyway so this should not be so big deal. Hopefully this helps.
^I was also quite surprised though I've tested only with Export all measurements as text. One possibility is that some timing setting of individual measurement was modified manually before Export measurement as text, and window/timing settings were (re-)applied to all measurements just before Export all measurements as text command. That would be clear human error, though some applications apply common settings for all measurements of exported group/sequence. For example CLIO with Auto save function, ARTA with Export IR and Convert IR to FR of VituixCAD. That is safe approach for off-axis sequences.
search REW help for "loopback". there isn't much but its all you need
REW only uses one soundcard channel to capture the output of your SPL meter or mic preamp, the Input Channel control tells REW which channel you have connected to. The default is the Right channel. If Use loopback as timing reference has been selected in the Analysis Preferences the other channel will be used a reference to eliminate time delays within the computer and soundcard, this requires a loopback connection on the reference channel. If the soundcard (or something else in the input chain) inverts its input select the Invert checkbox to restore correct polarity. If the input has a DC offset check the High Pass box to have REW automatically apply a 2 Hz high pass filter
re' preamp suggestion
I picked the following "usb audio interface" as typical of what you might want. I don't recommend it over others of its ilk as I've not used but its specs are what I would be looking for if I were looking. A lower end model from Scarlett is recommended on the REW web site. You may want one with more analog outputs if developing active speakers. As to the preamp included in the audio interface, you need at least one and it should have "phantom power". You don't want a separate/stand alone preamp; that is an expensive way to go.
Focusrite Scarlett 2i2 USB Audio Interface | Sweetwater
Yes the same loopback connection will work in Arta
re' use of MP extraction for time offset calculation
In the example, REW and Vituix came up with a 6 mm difference in offset for what looks like a crossover at 1 khz. That potential error is a small fraction of a wavelength at that frequency and thus probably good enough. If doing an active XO, one could easily tweak it right on during optimization. If doing passive XO, one would be happy to be that close to time aligned.
Its nice to have this method for a quick check or to get close prior to aligning IRs via the preferred two channel channel measurements.
In the example, REW and Vituix came up with a 6 mm difference in offset for what looks like a crossover at 1 khz. That potential error is a small fraction of a wavelength at that frequency and thus probably good enough. If doing an active XO, one could easily tweak it right on during optimization. If doing passive XO, one would be happy to be that close to time aligned.
Its nice to have this method for a quick check or to get close prior to aligning IRs via the preferred two channel channel measurements.
I won't be doing active anytime soon. I have an EMM-6 on it's way, I just need to pick a preamp. I am looking at a Focusrite Scarlett-Solo Gen2 USB Audio Interface. Thoughts?
Thank you,
David.
looks like it has all you need (and no more) and is well reviewed on the sweetwater site
At 24bits at 192 Khz its AtoD and DtoA should be excellent.
Personally I'm tired of buying these things, I'm on my 4th already and I do need more outputs for my active speaker plans. You don't want to underbuy feature wise and you want something reliable with good driver support. Check the reviews. Download the manuals. Check out the supplied software....Perhaps someone with good experience with the brand will speak up
At 24bits at 192 Khz its AtoD and DtoA should be excellent.
Personally I'm tired of buying these things, I'm on my 4th already and I do need more outputs for my active speaker plans. You don't want to underbuy feature wise and you want something reliable with good driver support. Check the reviews. Download the manuals. Check out the supplied software....Perhaps someone with good experience with the brand will speak up
You should check if it is a duplex card. The Scarlett 2i2 is probably what you want (for dual ch operation).
Just a quick note here.
Remember that the phase and calculated offset will be impacted by the HF and LF angles chosen for the tails. If using the same parameters, the phases should be similar with a very small difference based upon the method of calculation. That being said, there is no substitute for REAL measurement. The only reason to calculate and use MP is to estimate power responses off axis. Rather than worrying about the offsets, it takes little additional effort to actually take off axis measurements and then use this in VituixCAD for design of both power and axial responses simultaneously.
Remember that the phase and calculated offset will be impacted by the HF and LF angles chosen for the tails. If using the same parameters, the phases should be similar with a very small difference based upon the method of calculation. That being said, there is no substitute for REAL measurement. The only reason to calculate and use MP is to estimate power responses off axis. Rather than worrying about the offsets, it takes little additional effort to actually take off axis measurements and then use this in VituixCAD for design of both power and axial responses simultaneously.
I have seen the diagram for a generic sound card, how is the Scarlet 2i2 wired?
Left out from my DAC to my amplifer, mic to input #1 on Scarlet,
Right out from my DAC to input #2 on Scarlet?
Then input USB from Scarlet both channels?