can anyone help explain what I am hearing?

[mondogenerator]

Im experimenting with a small driver and ribbon. I have made good progress tuning the filter by ear so far, and im happy with the sound.

Whilst listening and AB swapping between new project and main system i have noticed what seems to be either an over emphasis or merely better definition of some vocal sounds, predominantly 'ack' sounds.

The trouble im having is that it just sounds 'right'. Of course it could be colouration, but it feels clear, intelligible and natural.

Can anyone help me check my sanity?

 

[DavidL]

Hopefully NOT this type of "ack" [IMGDEAD]http://freethoughtblogs.com/bluecollaratheist/files/2011/12/bill-the-cat.jpg[/IMGDEAD]

Sorry but couldn't resist. Bill the Cat from Bloomcounty
Have you any means to do measurements? Measurements are your friend.

 

[Mooly]

You have to trust what you hear and if you like the outcome then where is the problem

I have owned two amplifiers that could be described as "perfect". One a diy job and one bought. Neither was musically satisfying.

Should I

a) keep and use those two amps because they are "technically correct" even though I have heard far better.

b) Use the amps that I really like listening too.

How do you know your diy speakers aren't in fact interacting "perfectly" with your listening environment to give you the sound you want.

 

[chris661]

Ribbons often have rather a lot of distortion.

You might well be hearing that, and perceiving it as a bit of lift in the upper midrange and treble.

 

[DavidL]

Mooly and Chris661 posts bring up subjective vs objective. You'll have to decide which camp you are in, I prefer to be in both but of course more in the objective side. I know a lot of people on here will take great pains to design a speaker by using available software but then they never measure the actual results to see if what they modeled meets real life. That's only going part way in my opinion. Models are only as good as the designer's ability to not only input the correct frequency, phase , acoustic offset and impedance but also knowing what components in the crossover to vary to get great results. Why so few bother to measure the end results is beyond my understanding. I hope this encourages you to do so.

 

[mondogenerator]

Thanks all
Mooly, thanks. That thought crossed my mind too, and it doesnt seem coloured or quacky as Id expect if it were a colouration.

Chris, I understand your opinion. Since the mid is an alpair5, i tried with LP and ribbon removed. Same result. Maybe a slight emphasis at 500 hz to 1k at a guess.

Im inclined to think the presence and clarity (i know, subjective buzz words) the mid-HF have is merely a level of low level detail i hadnt given the alpair credit for.

Sibilant sounds Ive found are tough to reproduce naturally, 'p' 't' 'ack'and 'd', they all sound natural.

Above 5k the ribbons take over, and sound everybit as clean as the domes in my mains, visaton g20sc, perhaps even cleaner. Personally im amazed how they sound, and pick up then cease transients.

David, you are correct of course, but i have yet to get measurement setup sorted out.

 

[CanAm Man]

Have you tried installing a 12dB/octave butterworth "ACK" filter?

 

[mondogenerator]

lol I guessed some clever Richard would try humour, eventually.

 

[Andersonix]

Above 5k the ribbons take over, and sound everybit as clean as the domes in my mains, visaton g20sc, perhaps even cleaner. Personally im amazed how they sound, and pick up then cease transients.

See? You knew the answer all along...

 

[AllenB]

Im inclined to think the presence and clarity (i know, subjective buzz words) the mid-HF have is merely a level of low level detail i hadnt given the alpair credit for.

Sibilant sounds Ive found are tough to reproduce naturally, 'p' 't' 'ack'and 'd', they all sound natural.

These sounds span quite a wide band of frequencies making it harder to get right. Most speakers don't seem do these sounds well, sometimes due to a response peak ruining them or a valley taking the impact out of them.

From what you describe in your first post you may have a gradually rising response, although it may be relatively free of distinct peaking. The ribbon's dispersion might be contributing.

 

[mondogenerator]

Thanks Allen, that sounds plausible, in fact it would make a great deal of sense. Since ive padded the tweeter by ear, as well as experimenting with small resistances in the XO, it could be that in tweaking the XO to sound flat through the transition, i have caused a small rise in the final octave. Thanks again.

 

[dumptruck]

I would (as you already did) guess you're hearing an emphasis 500-1000Hz, based on the "ack" description.

 

[dumptruck]

P.S. is your baffle about 20 cm wide?

 

[weltersys]

I would (as you already did) guess you're hearing an emphasis 500-1000Hz, based on the "ack" description.

I see more 2.25 to 3.5K "ack" but maybe I pronounced it wrong for my test .

 

[mondogenerator]

lol you guys... The baffle is maybe 20cm wide, but the box is hardly finished, the baffle is recessed and unglued!

I ran the alpairs OB on both this baffle and a 4ft square one, the 'ack' character is much the same.

To be honest, my gut says its just detail that the alpair renders better than the typical midwoof which is pressed into midrange duties these days. I just lack the confidence in my assertions to proclaim my very basic setup as natural as it feels!

AllenBs point made sense as ive played with resistance to soften the 2nd order HP to tweeter @ about 5.5k (0.33mH and 3uF).

I know my alpair5 have edge resonance ~400hz which is around my lower guesstimate, and 2k is to low (or high lol) to be the xo transition...

Hmmm, maybe its just clarity of 'ack' im hearing

 

[dumptruck]

I'm guessing you're hearing a response that shelves down around 500k or so. Not that I actually think this method of funny words and no measurements is effective , but you might as well try a simple inductor/resistor BSC circuit in front of your low pass temporarily, mess with the resistor, and see what you think.

I see more 2.25 to 3.5K "ack" but maybe I pronounced it wrong for my test .

Ha, that's the spectral content of you saying "ack"? Brilliant. The strong upper mid content is what defines the consonants, and I was interpreting the hard AAAAA sound as the key, as in "back rattle blat clatter crack". I have no idea if the spectral content of people saying "AAAAA" (not to be confused with "AHHHH") looks like a lot of 500-1k, I just know speakers sound a bit clatter-ey to me with that.

 

[weltersys]

Hmmm, maybe its just clarity of 'ack' im hearing

Or "ack" back reflections.

Try modeling in Akaback .

 

[mondogenerator]

sheee... Yeah, reflection could be it....I didnt have the OB and boxed in AB or anything. I guess its just the character of the alpair 5. I prefer it as a mid though, fullrange its ok.

Seriously though, BSC is likely the thing as ive not wrangled the woofer and rest of the xo into tests yet. The aim being to fit the mid and tweeter first, then bring in the BSC with the woofer. Allen probably still has a valid point too and with no BSC I reckon its easy for me to contine that trend, albeit a gentler slope, in the tweeter section.