No hard edges on my speakers (oval shaped as seen from above) is the first helping hand. Less diffraction in early stages due to smooth transitions everywhere. A corner on an array is still a parallel offender, something in common to most drivers as seen from the listener. Let the array average out reflections, but be very aware of any parallel plane, ridge, surface etc.
The 5/6 cycle FDW helps us view the early wavefront. But it still has to be clean enough! I absorb all mayor reflections in the first 20 ms, but that took me only 3 panels. I wanted more, but did not get permission
.
The wall behind the speakers did not need any treatment at all. Believe me I've measured with and without. I've spend quite some time chasing every wiggle in my measurement, armed with damping panels. So I learned how "to read" my room and what it looked like in the measurements (IR and Filtered IR tabs mostly, assisted by the spectogram, wavelet and waterfall plots, you need all those tabs to find out what's really going on).
Gate your measurements, look at all graphs to see the results of those gated measurements. Slowly increase the time to see the wave front develop, observe phase frequency and learn what does what in the other graphs.
It takes time, but after that the room will have less secrets as viewed from the listening area.
Lots of measurements (all along the [listening] area of interest) helps to confirm what you see. It tells you what to fix and what you shouldn't fix.
I've only treated the large offenders. But managed to get a pretty clean IR...
(and I can point out what's left and why)
It's all a big secret, one that I shared in this thread.
The 5/6 cycle FDW helps us view the early wavefront. But it still has to be clean enough! I absorb all mayor reflections in the first 20 ms, but that took me only 3 panels. I wanted more, but did not get permission
.
The wall behind the speakers did not need any treatment at all. Believe me I've measured with and without. I've spend quite some time chasing every wiggle in my measurement, armed with damping panels. So I learned how "to read" my room and what it looked like in the measurements (IR and Filtered IR tabs mostly, assisted by the spectogram, wavelet and waterfall plots, you need all those tabs to find out what's really going on).
Gate your measurements, look at all graphs to see the results of those gated measurements. Slowly increase the time to see the wave front develop, observe phase frequency and learn what does what in the other graphs.
It takes time, but after that the room will have less secrets as viewed from the listening area.
Lots of measurements (all along the [listening] area of interest) helps to confirm what you see. It tells you what to fix and what you shouldn't fix.
I've only treated the large offenders. But managed to get a pretty clean IR...
(and I can point out what's left and why)
It's all a big secret, one that I shared in this thread
.
Last edited:
...It's all a big secret, one that I shared in this thread
Sherlock & Watson whisper look for DR.....
That said any chance we get a APL sweep for that lovely remodeled new kitchen
I played around with this thing:
amroc - THE Room Mode Calculator
One thing I would like to see you try, is the low 72Hz dip. Want to see what happens if you do a narrow, full phase rotation convolution at the left tower centered at 72.14Hz (create a huge peak?)
Warnings enough about room EQ. But with such an advanced system as you have, adding multiple other sound sources in the mix, who wouldn't want to see what happens when done carefully... For one, your speakers seem to do good things with the higher frequency modes.
amroc - THE Room Mode Calculator
One thing I would like to see you try, is the low 72Hz dip. Want to see what happens if you do a narrow, full phase rotation convolution at the left tower centered at 72.14Hz (create a huge peak?)
Warnings enough about room EQ. But with such an advanced system as you have, adding multiple other sound sources in the mix, who wouldn't want to see what happens when done carefully... For one, your speakers seem to do good things with the higher frequency modes.
Last edited:
I've done my share to see what I could do with the dip, believe me.
Among those tricks were phase manipulations. However this shifting of energy was the most successful tweak to fix that hole. Hopefully soon to be fixed by the left sub.
Room EQ can work fine, but one should fix as much as possible with other (passive) means and you can only fix something that what's in common over your entire listening area of interest.
Fixing things with as short a window as feasible (Frequency dependent) has my preference though.
A lot happens once the wave front is out in the room long enough. Lots of bad stories about EQ-ing still survive. Yet with care it can be the icing on the cake.
I'm sure you are aware of this, but for passing viewers, I can't mention this often enough. EQ isn't evil, as long as you realize what it is you're EQ-ing, in other words don't use long non gated measurements as a base. For low Q tonal tweaks it might work to some extend, I do use longer gated but smoothed graphs for tonal balance myself.
Even there I prefer frequency dependent gates.
Among those tricks were phase manipulations. However this shifting of energy was the most successful tweak to fix that hole. Hopefully soon to be fixed by the left sub.
Room EQ can work fine, but one should fix as much as possible with other (passive) means and you can only fix something that what's in common over your entire listening area of interest.
Fixing things with as short a window as feasible (Frequency dependent) has my preference though.
A lot happens once the wave front is out in the room long enough. Lots of bad stories about EQ-ing still survive. Yet with care it can be the icing on the cake.
I'm sure you are aware of this, but for passing viewers, I can't mention this often enough. EQ isn't evil, as long as you realize what it is you're EQ-ing, in other words don't use long non gated measurements as a base. For low Q tonal tweaks it might work to some extend, I do use longer gated but smoothed graphs for tonal balance myself.
Even there I prefer frequency dependent gates.
I believe that going with a sealed design and a setup that takes ideas from a commercial design, using simulations... it can’t go that bad.... and even so using great ScanSpeak drivers that proved their way in other diy designs... I don’t think the risk is that large .... I reckon the question is how far-out we can risk going on our designs
@haraldo I think the Loudspeaker database is a fun place to look for possible suspects if you want to go ahead with your plans.
Here's a selection of 10"drivers:
Loudspeaker Database
It shows graphs if they have them and shows all important specs from lots of woofers. I see a lot of so called Car woofers in there too. But I'm pretty sure you'll find something interesting there. I can't promise they only feature current drivers, some may be outdated, no longer in production etc.Plus you can select on lots more variables like x-max or Fs. (Though not on mms
)
A comparison of the Scan-Speak 30w and 32w,
First the 30W:
Scan-Speak 30W/4558T00 review, Hobby HiFi 04-2011
Next the 32W:
Scan-Speak 32W/4878T00 review, Hobby HiFi 05-2013
The 32W has more displacement, but all in all the 30W isn't bad at all in this comparison. Look at the distortion figures measured under similar/same conditions. People always lust for the bigger, more expensive driver. I say give those little jewels a shot too.
In subjective reviews, which driver do you think will win? Looking at the data may paint another picture. This is what I found way back when I started looking at subwoofers.
For those interested in the smaller 10" 26W, here's some info on that one:
Scan-Speak 26W/4558T00
If you ask me, back when Scan-Speak and Vifa/Peerless parted ways Scan-Speak needed to complete their line up of drivers again. That's how we got the great 10F, some great ring radiator tweeters in the Discovery line and the 30W that needed to replace the Peerless XXLS. I'm guessing they made sure that they were at least as well performing as what they replaced. Probably the same team of people was involved in making the replacements. At least they had a good starting point.
The Peerless/Vifa brand used to be a guinea-pig for new things which later on in life got expensive sister drivers. At least that's how I interpret it. The Discovery line sure has some interesting gems if you ask me.
First the 30W:
Scan-Speak 30W/4558T00 review, Hobby HiFi 04-2011
Next the 32W:
Scan-Speak 32W/4878T00 review, Hobby HiFi 05-2013
The 32W has more displacement, but all in all the 30W isn't bad at all in this comparison. Look at the distortion figures measured under similar/same conditions. People always lust for the bigger, more expensive driver. I say give those little jewels a shot too
.In subjective reviews, which driver do you think will win? Looking at the data may paint another picture. This is what I found way back when I started looking at subwoofers.
For those interested in the smaller 10" 26W, here's some info on that one:
Scan-Speak 26W/4558T00
If you ask me, back when Scan-Speak and Vifa/Peerless parted ways Scan-Speak needed to complete their line up of drivers again. That's how we got the great 10F, some great ring radiator tweeters in the Discovery line and the 30W that needed to replace the Peerless XXLS. I'm guessing they made sure that they were at least as well performing as what they replaced. Probably the same team of people was involved in making the replacements. At least they had a good starting point.
The Peerless/Vifa brand used to be a guinea-pig for new things which later on in life got expensive sister drivers. At least that's how I interpret it. The Discovery line sure has some interesting gems if you ask me.
Last edited:
It pretty much looks like Scan Speak is in a leage of it's own now.... and also reading about and listening to people who tried the 32W, they all been blown away, no compare to the smaller 26W or Peerless XXLS ...
The 32W is seemingly significantly better than previous generation Scan Speak woofers and subwoofers....
I am wondering for sealed subs if it's these two drivers that would be on top of the shelf
Scan Speak 28w/4878t01 Looks like a somewhat scaled down version of the 32W, looks the same, same sandwhich cone construction and very similar in parameters
Scan Speak 32W/4878T11 is a brand new one with much lighter weight cone than the 32W/4878T00 and has lower Qts, and I wonder if it will behave better in a sealed cabinet of limited size ....
It looks like my choices would be towards one of those above, maybe
How much room gain did you cater for when you were modelling your subs?, there is not that much info out there that I can find, and it all depends on room size and geometry and so on ...
The 32W is seemingly significantly better than previous generation Scan Speak woofers and subwoofers....
I am wondering for sealed subs if it's these two drivers that would be on top of the shelf
Scan Speak 28w/4878t01 Looks like a somewhat scaled down version of the 32W, looks the same, same sandwhich cone construction and very similar in parameters
Scan Speak 32W/4878T11 is a brand new one with much lighter weight cone than the 32W/4878T00 and has lower Qts, and I wonder if it will behave better in a sealed cabinet of limited size ....
It looks like my choices would be towards one of those above, maybe
How much room gain did you cater for when you were modelling your subs?, there is not that much info out there that I can find, and it all depends on room size and geometry and so on ...
Last edited:
Vituix has a good library of scanspeak woofers in its data base, including the 32W/4878T11. It only took a minute or to to trot out the enclosure tool and see it needs only 35L for QTC=.7. and 26L for Qtc=.9 which I would say is doing pretty good in a limited size enclosure.
Room and boundary gain are strongly house construction dependent. Unless your walls and floor are block or concrete, they are likely to become transparent to the <30 Hz bass, where you really need the support. Best thing to do is build a single driver test box and compare indoor to outdoor measurements.
Yes, I know, easy thing for me to say from Florida at the onset of winter.
Room and boundary gain are strongly house construction dependent. Unless your walls and floor are block or concrete, they are likely to become transparent to the <30 Hz bass, where you really need the support. Best thing to do is build a single driver test box and compare indoor to outdoor measurements.
Yes, I know, easy thing for me to say from Florida at the onset of winter.
Not my own measurements. Don't know which article I got it from but here's one for the 28W 4578T00
Test Bench: Scan-Speak 28W/4578T00 10" Subwoofer | audioXpress
The black lines are sealed configuration at 2.83v and 33v.
Just wanted to show that frequency response is not a static thing, and listening voltage (+Linkwitz transform) should be considered when choosing box size.
I don't know if this changes when using current amplification, I'm not an expert.
Volume dependant EQ could be considered too.
This is result I got from the "Embla" simulator putting 6 Scan Speak 32W/4878-T11 in a 140 liter cabinet, including room gain with corner loading and a 7 db boost at 17 Hz. The linkwitz trabsform with this program is limited to around 7dB max but some more gain would give flat in room response to below 18 Hz. (Please ignore all the vented box figures)
This simulator states 100 dB in room at 20 Hz with less than 2 mm cone travel, it's got to be very far off
I have no idea how accurate this is, and as of the room gain it is highly dependant on room. Thinking of checking these simulatins and see how they fit to what winisd gives.
This simulator states 100 dB in room at 20 Hz with less than 2 mm cone travel, it's got to be very far off
I have no idea how accurate this is, and as of the room gain it is highly dependant on room. Thinking of checking these simulatins and see how they fit to what winisd gives.
Last edited:
Not my own measurements. Don't know which article I got it from but here's one for the 28W 4578T00
Test Bench: Scan-Speak 28W/4578T00 10" Subwoofer | audioXpress
The black lines are sealed configuration at 2.83v and 33v.
Just wanted to show that frequency response is not a static thing, and listening voltage (+Linkwitz transform) should be considered when choosing box size.
I don't know if this changes when using current amplification, I'm not an expert.
Volume dependant EQ could be considered too.
Anyone.... what is is that can be the root cause of that 36Hz peak at 110dB
Well then, exactly what I meant when I said:
And why? Two reasons here...
- one: You're moving from a 10 inch cone to a 13 Inch, that's going to do something!
- two: I've you just paid about twice as much for a driver, what else would you expect but to be blown away...
But in reality, did you take a good look at those two distortion graphs in the pdf files I posted? Both were playing at 90 dB, which one was cleanest?
Thx, not yet .... gonna read tomorrow and try to get through all constructive input
And I had no idea that Qtc changes at different SPL levels
Those are just simulations, who cares... that's not an indication of anything.
If you want to know what happens, one should take the real driver and measure it at those values. That plot is pointless.