Jawen,
I hope you are having a pain free day.
I am not a REW user, so my answer may be of the mark.
I believe you need to have the start window just in front of the impulse you can move it around in small increments between no signal and the max amplitude and see how it changes you data in the high frequencies. From the posted image i see the window starting some time after the impulse as built up?
As you close the window by moving it from the 8mS you have shown down to 6, 5, or 3 mS you should, have a measurement where you have gated out the room reflections/floor bounce. You will also see that the low frequency information becomes lower resolution and you will get a roll off shown by the data declining below approx. 400 Hz depending on how small the overall window time is between 3 to 8mS. say.
I am sure you have seen YouTube videos where people get the loudspeakers outside on a step ladder to give them an overall longer flight before the lessened floor bounce. This is due to the added height causing the floor bounce to be moved further out in time, which in turn means you can measure further down in frequency without the bounce signal messing up the frequency response. Please don't try this with your current health.
Have a look on YouTube I think there is a VituixCad cad tutorial that explains about splicing the nearfield frequency response of the woofer to that of the midrange. depending exactly on your wiggles that may help smooth out the lower data (350- 400 Hz approx.) when used with a short window that cuts out early reflections or floor bounce.
I hope you are having a pain free day.
I am not a REW user, so my answer may be of the mark.
I believe you need to have the start window just in front of the impulse you can move it around in small increments between no signal and the max amplitude and see how it changes you data in the high frequencies. From the posted image i see the window starting some time after the impulse as built up?
As you close the window by moving it from the 8mS you have shown down to 6, 5, or 3 mS you should, have a measurement where you have gated out the room reflections/floor bounce. You will also see that the low frequency information becomes lower resolution and you will get a roll off shown by the data declining below approx. 400 Hz depending on how small the overall window time is between 3 to 8mS. say.
I am sure you have seen YouTube videos where people get the loudspeakers outside on a step ladder to give them an overall longer flight before the lessened floor bounce. This is due to the added height causing the floor bounce to be moved further out in time, which in turn means you can measure further down in frequency without the bounce signal messing up the frequency response. Please don't try this with your current health.
Have a look on YouTube I think there is a VituixCad cad tutorial that explains about splicing the nearfield frequency response of the woofer to that of the midrange. depending exactly on your wiggles that may help smooth out the lower data (350- 400 Hz approx.) when used with a short window that cuts out early reflections or floor bounce.
I've been away & celebrated midsummer, Sweden's biggest holiday/ festival/ feast
An approx. 600-1200-year-old feastday, & it is now that the bright nights are turning so the daylight is getting shorter and shorter. In the days- everything that grew in nature at that time was believed to have magical powers
By braiding a flower wreath, the girls could save it to have the magical powers the rest of the year.
& the dance around the midsummer pole, was for the children to get to know each other & to strengthen the bonds between people.
And if a girl picked 7 different flowers & put under the pillow when she was going to sleep, she would dream of the man she would later marry.
Some still wear Swedish folk costumes, and music is played with fiddles & the very old Swedish instrument "key harp"
We eat and drink togeter and sing snaps-songs ( snaps= 3-6 cl 40% spiced liquor, but every snaps need a song first)
Here is two short videos from two small places i Sweden celebrating Midsummer, for you how like so se what it´s about.
And i will come back in a separate post about the diy speaker.
Best regards John
An approx. 600-1200-year-old feastday, & it is now that the bright nights are turning so the daylight is getting shorter and shorter. In the days- everything that grew in nature at that time was believed to have magical powers
By braiding a flower wreath, the girls could save it to have the magical powers the rest of the year.
& the dance around the midsummer pole, was for the children to get to know each other & to strengthen the bonds between people.
And if a girl picked 7 different flowers & put under the pillow when she was going to sleep, she would dream of the man she would later marry.
Some still wear Swedish folk costumes, and music is played with fiddles & the very old Swedish instrument "key harp"
We eat and drink togeter and sing snaps-songs ( snaps= 3-6 cl 40% spiced liquor, but every snaps need a song first)
Here is two short videos from two small places i Sweden celebrating Midsummer, for you how like so se what it´s about.
And i will come back in a separate post about the diy speaker.
Best regards John
It´s so many (for me unknown) "special terms" used in different areas.
Mesaurements have a lot, drivers also and if we talk diy a complete speaker it´s 100:ed´s.
20 years back i studdy computer security for 2 years with sometimes thick books with 988 pages in "computer-english"
It was the same problem for me because every special niche, has it´s own "words" (words that are not normally used)
So exactly what is "raw impulse response"?
regards John
No need to gate.
Just need to get rid of the phase wraps from constant delay first.
Easy way is to use FDW in IR windows. Reduce number of cycles till phase smooths out.
Other way is to get impulse delay right. Try REW's estimate impulse start etc.
Why do you have 2000ms Timing offset in measurement? And why play with window size yet...until you past the point of phase making sense sense?
Walk before run, and good luck
Just need to get rid of the phase wraps from constant delay first.
Easy way is to use FDW in IR windows. Reduce number of cycles till phase smooths out.
Other way is to get impulse delay right. Try REW's estimate impulse start etc.
Why do you have 2000ms Timing offset in measurement? And why play with window size yet...until you past the point of phase making sense sense?
Walk before run, and good luck
Hello dear Ray!
Never have pain-free days, and the "thing" is to make it through the days and try to cope it.
In a scale from 1-10, so I can manage 3-4 & a few hours 5, but when the pain goes up into long 5's or 6-7-8 it's hard to cope.
And i try to take as little Oxycodine as possible, because i have seen what it does to people, so maximum 2-3 oxy 5mg in total in the weekends, to get a slightly higher quality of life. (but oxy and morfine just take "the pain-tops away some, to cope the worst hours)
I have things & movements as well as body positions, which sometimes suppress it, and to stand upp and walk some is the best, but can't do that all day because my my strength is running out.....recovery for the body is a constant struggle, because it tryes to heal it self 24/7 so hard for me to get "energy" but train some to get my head more "possetive"
About mesauring and all measurement techniques it´s real hard to learn and understand.
Sadly Youtube havent any good videos about later versions of REW, and the one i find is not particularly education for the spectator.
I had took one of the mesaurement from last week, and try "to play" with it in REW, but without understanding of what you see its impossible to strive for improvements or see disadvantages in a DIY design, even if i allways ends up with real good sound in my diy´s ( my opinion of course, and some friend)
Here i have 3 ms and have draw it to minus-impulse, and there is at least some movement on the scale
regards John
Attachments
I dunno. REW keeps changing.
Try inputting the 3 time windows in #33.
Or open a new REW and make new meas.
edit...I'd really make some new measurements. The 2000ms time offset in the originals has to be backed out before any phase sense can be made.
It's often easiest/best for me to restart fresh, when nothing makes sense
Try inputting the 3 time windows in #33.
Or open a new REW and make new meas.
edit...I'd really make some new measurements. The 2000ms time offset in the originals has to be backed out before any phase sense can be made.
It's often easiest/best for me to restart fresh, when nothing makes sense
Last edited:
I had a look at some other older REW mesaurements from unother diy speaker i build, and it also have minus"time offset"
Don´t knom why, but i cant find any in my computer that has "zero time offset"
Here is one from last year ( MTM with Audax hm210z10 and Mundorf AMT) named like this 3
and one from last week aka 3 way SS build, so 2 different diy... named "like this 4.
As you se both have IR window settings "default" Left window 125 ms and Right window 500 ms
But Ref time ms is minus- 2,57 at one and minus - 8,25 at the other (dont know why)
And i can select "Add FDW" without getting "Red warning" on both of them, and also play around with different "cycles".
But if i change " Ref time ms" on these mesaurements, the program/REW red warning says "Windows Ref time is not at peak,the peak is at -2,570 ms (and -8,250 ms)
Everything is complex
Attachments
Hi Jawen thanks for the video links which I will watch after this.
The update on your current health and the meds, very sad to hear about that, but glad to hear you managed to do some traditional things for the solstice and singing and schnaps are known to be good for you, but only in moderation or you will have a sore throat as well.
Lots of stuff here and some further advice on REW, which is good but having to re do measurements looks like it wouldn't do your health any good?
Re your post 28 could you possibly bring the tail further to the left around the -1.0 mark just before the little bump and also -2.6 basically at the bottom of the pulse decay. What do they look like?
If you have to reset REW as described maybe do some real simple measurements on a small speaker and post those results so that other can comment if REW is configured correctly ?
The update on your current health and the meds, very sad to hear about that, but glad to hear you managed to do some traditional things for the solstice and singing and schnaps are known to be good for you, but only in moderation or you will have a sore throat as well.
Lots of stuff here and some further advice on REW, which is good but having to re do measurements looks like it wouldn't do your health any good?
Re your post 28 could you possibly bring the tail further to the left around the -1.0 mark just before the little bump and also -2.6 basically at the bottom of the pulse decay. What do they look like?
If you have to reset REW as described maybe do some real simple measurements on a small speaker and post those results so that other can comment if REW is configured correctly ?
Ref time can be moved in REW by pointing with mouse, then pressing Shift and dragging. Reference goes to highest amplitude (peak) automatically and when left window is long (as by default) - it might be just some clonk, touching mic etc. Windowing settings before taking a measurement are not permanent, they can be altered this way. This is one of caveats of using usb-mic one-channel measurement without fixed timing reference.
This was causing me troubles when measuring dipole speaker indoors... I had to move 90deg measurement's Ref 8,6ms earlier because delayed reflection had higher amplitude than direct sound.
This was causing me troubles when measuring dipole speaker indoors... I had to move 90deg measurement's Ref 8,6ms earlier because delayed reflection had higher amplitude than direct sound.
Last edited:
