Can't tell you how many times I've changed speaker position. When we're done doing the calculations I'll go and revisit speaker placement. For quite a while I had blue tape all over the floor with notes attached. Wife's very understanding.
Yes, useful to be able to mark the position, I have cork tile floor which makes it easy since I move them out of the way at times. Just a thought, perhaps you'd like to PM Colin http://www.diyaudio.com/forums/members/colin.html from Jordan, to let him know what you're up to
@wesayso
Thanks advises for measurements, from beginning its not worlds smoothest and maybe not a wish in this SLOB system, guess that is up to owners how much time offered to smooth it out in precision style of Barleywater, intend focus to get XO region acceptable for this exercise, below red trace is manufactures response times OB simulation in Edge and orange is what mkane77g had pages back.
I get that, however if one would EQ out a reflection at the measured spot, the result would never be right at the listening spot. The peaks/dips we see in the measurement look like floor bounce to me. If this is floor bounce at close measurement distance, it will hit somewhere else at the listening position. All that EQ for crossover perfection would be invalid.
Look at the IR, the Filtered IR, the wavelet and first find out what's happening. Solve that (even if it's only temporary) so you can make sure the speaker hands over to the next without the room or other boundaries being involved. Anechoic like measurement conditions for crossovers is why we measure up close. Raise the speaker if needed...
True that reflections make it difficult. But just adjusting timing on the fly can get strange results too, which is why I'd keep looking at all graphs. One can adjust the timing of the tweeter (I'll continue calling the high frequency driver that) and have it "in phase" while in fact one is a full cycle off in timing. The sum would look nice enough, but you'd have to know if both drivers are actually time aligned.
While this is more a problem of higher frequency crossovers, it is way too easy to look at each graph to get a visual confirmation of what we are looking for. So use those tabs! Use the overlays, use the filters and see what is happening. They are there for a reason.
Not directed at anyone in particular, just a reminder there is more than one tab in REW that gives you hints to the information you require. Why not use them...
The graphs are all related, so you can find more info about that unforeseen dip etc...
Yes if can trust we only measured direct sound and there no reflections from boundaries involved woofer should be corrected and if it has electric LP @125Hz LR4 in miniDSP then maybe try lower it to LP @80-100Hz LR2 or BW2. Is it possible you add target curves to published plots because it would be faster than each time have to scan curves here in the other end : )
Also if you have a cloud drive can you share sessions REW mdat-file and share a link public or via PM.
Not sure understand right maybe lack of native language, are you listenening to it now and it sound is close to goosebumbs or you need a new target for house curve : )
No matter how it sounds we ain't there even it should happen sound good, we need to have the two pass band right and get their SPL level in line plus found the deep such out when one driver have reversed polartity while delay for tweeter is upped in small steps.
Attachments
mkane77g shared a sketch pages back size numbers and BS plus diffraction is below, many curves had been looked at or manipulated down the road but pretty shure that red curve is nothing but a traced curve of manufactures website response that is commanded times exported file from Edge. Have never worked with OB and could see when ticked on in Edge diffraction was worse and roll off steeper than if driver had a say 3,5 liter sealed cavity behind same baffle, manufacture response without any baffle step loss looked close to cover a 80Hz BS2 roll off but live measurements and that simulation showed we had to up point of XO frq.
Attachments
Last edited:
Low freq LR2 in-room should look something like this (8/60ms right window)
Mkane77g, you don't get summation at all, relative delay is almost 180 deg wrong. Please remember that with OBs very often the reflected sound from front wall is louder than direct sound, and REW picks up the loudest signal...
Mkane77g, you don't get summation at all, relative delay is almost 180 deg wrong. Please remember that with OBs very often the reflected sound from front wall is louder than direct sound, and REW picks up the loudest signal...
Attachments
We try to set it acoustic LP/HP LR4 at first because that can be gathered time aligned when a deep deep null is happening with one reverse polarity driver, then thereafter if in electric filter we EQ a transfer function from LR4 slopes to BW4 and BS2 and acoustic slopes will also change to same and only need delay numbers for mid-tweeter added to where it ended in LR4 setup, admit haven't try allign this way before but hope it works and reason is because mkane77g is new to diy a DSP speaker.