That’s a bit easier to read with 1/12 smoothing
Printable View
That’s a bit easier to read with 1/12 smoothing
Have done time alignment with impulse responses now. Distance measurements weren't that far off.
Had to add some delay to get a deep null in the sub/midbass xover and then flipped polarity. Let the TA for the sub at 0.
I think the installer wired the sub backwards, therefor the polarity flip. If i added 7.9ms to the delay on midbass/tweeters i had good summing with no polarity flip anywhere.
But for whatever reason i cannot get the midbass/tweeters to sum up fully, but as someone mentioned it might be hard to get a full summation in that area.
See for yourself.
Attachment 11303
And this is the total response left+right. Good summing in bass, but not so good in midrange/tweeters.
Attachment 11304
Center image is strong. When listening there is nothing wrong tonally.
you may also which to try looking with 1/48 octave as you may see more detail about dips and peaks, smoothing makes a response look artificially better, so what you think is the level may actually be 2db off for example...
if you smooth a 1/48 octave response to 1/3 it can look like a straight line, taking it to an extreme if you have a wave like pattern of measured response with equal peaks and dips of 3db each you could smooth it to flat, now if the opposite side response had the exact opposite peaks and dips and you smoothed that to 1/3 octave and it looked flat if you then played the two together you would then get zero summation by looking with 1/3 octave smoothing, but with more detail in 1/48th octave you would see that dips cancelled peaks and peaks cancelled dips
sonetimes we have to look a bit deeper to find what’s actually happening, I will always measure in 1/48 octave and apply smoothing post measurement, that way you always get the bigger picture if you see what I mean
you will often get peaks and dips that are in differing frequencys on opposite sides of the car so it’s very reasonable to dummies this may be why you don’t get 6db of summation
i hope that makes sense 👍🏼
Yes, it makes sense, but, i'm talking about summation on left side midbass and left side tweeter crossover, and right side midbass and right side tweeter crossover, not summation of left and right together.
I checked polarity and couldn't find anything wrong with polarity (checked with speaker pop app on my iphone).
Here in 1/48 smoothing.
Right side:
Attachment 11305
and left side:
Attachment 11306
If possible, it would be awesome if you could share your process on how to measure phase in SMAART. I am COMPLETELY new to it but would love to learn more and compare to what I’m seeing in REW. Any tutorials you found helpful?
Thanks!
Sent from my iPhone using Tapatalk
Ok, I think i got the reasons for subpar summing between midbass and tweeters.
1. i have a sharp dip of about 4-5db on both midbass drivers around 2500 Hz (and have my crossover set there).
2. below the dip in the midbass i have dips on the tweeter, mainly between 2000 and 2500 Hz of about 2-3db.
3. something is wrong with polarity.
Regarding point nr. 3: before my last remeasure i had to flip polarity on the subwoofer to get good summing from sub to midbass. Without that i have a sharp dip right at the crossover.
When i flip the sub back to normal polarity and flip the midbass instead i again have good summing from sub to midbass and better summing from midbass to tweeter as compared to normal polarity for midbass and tweeter.
Anyway... the dips are not responding to EQ at all, so i guess I have to live with that for now until i decide to upgrade the fronts from 2-way to 3-way with a dedicated midrange and higher crossover between mids and tweeters.
Only thing that puzzles me is the polarity issue. I measured polarity with a speaker pop before time aligning and eqing the drivers and got a positive on all speakers, but apparently this is wrong.
Or is there any other explanation for that? Delay settings are in the range of the measured distances plus minus a few centimeters.
nathan lively... sound design... look him up on you tube, everything I know I e picked up from him and his videos, he covers everything from interfaces to first measurements to simple stuff, some formulas and a few simple demonstrations, he’s a cool kid... didn’t even mind when I dropped him an email about one of his formulas being incorrect, we had a bit of a chinwag
the nice thing about it is you can see phase issues and then dial in crossovers to avoid them as much as possible, it also shows up reasons why summation doesn’t happen fully or at all... see my previously posted screen shot as a demonstration
i tend to use 1/12 smoothing in phase plots and 1/12 on magnitude plots also
once you get the hang of it it really is a game changer for getting sub to midbass to midrange all in phase and timed to perfection (I use a static mic for below 300hz and timing above that frequency)
it’s a very nice setup, I use an m-audio usb mobile pre and a Dayton mic which needs a 48v phantom power preamp the model escapes me, but it’s relatively cheap and the audio interface was £35 second hand incase I didn’t get on with it
The actual phase of each driver isn’t as important as sometimes you will need to be reversing the phase on some drivers, for example in my pic above I have my midbass out of phase and the time alignment at some odd settings, there are always at least two ways to align drivers, and one definitely doesn’t follow a tape measure
the nathan lively video about satlive I think it is explains about the delay finder and how it sometimes works out a polarity inversion and a different delay can make the summation better than a simple distance measurement... if I can lay hands on it I will post it up, it won’t nesc be relevant if you don’t have the facility to measure phase but it illustrates how we don’t always end up with settings that are immediately obvious and they work also
That’s the series of videos...
https://www.youtube.com/playlist?lis...vaN7W4sl2DGTCl
And this is the one where he has two options for a polarity inversion and more delay or less delay without a phase inversion I was referring to... I would recommend watching the whole lot of videos as a learning tool, it is easier if you can play and test stuff, but you will get the idea about this powerful tool and what info it gives
https://www.youtube.com/playlist?lis...vaN7W4sl2DGTCl