Obtaining my White Whale (25 Driver IDS-25 Line Array Build)

Waterfall:
Left center couch
Right center couch
 

Attachments

  • 2019_04_23 Waterfall_New Position_Left_Center Couch.jpg
    2019_04_23 Waterfall_New Position_Left_Center Couch.jpg
    156 KB · Views: 347
  • 2019_04_23 Waterfall_New Position_Right_Center Couch.jpg
    2019_04_23 Waterfall_New Position_Right_Center Couch.jpg
    162.6 KB · Views: 339
Dropping pretty hard above 6kHz.
Unless that's your preferred listening FR curve.

Don't worry, I'm getting there! I made some EQ tweeks last night that sound very promising. I won't be able to get measurements though till Thursday.

I'm still trying to learn how to EQ them using peq filters. The eq adjustments I have made so far are simple parametric EQ.
 
% Impulse:
Center of couch left
Center of couch right
All left measurements
All right measurements

I'm still reading along :).

If I look at this view of the multiple impulses (in % mode), it seems they are not aligned at the "0" point. We really should fix that before making an average of those IR's.
I know you're going to make a new set of measurements, but it can't harm to practice some in the mean time.

When looking at an impulse in the "%" mode, the first biggest peak should be exactly centered at zero (0).
Go trough your measurements one by one and align them.
Usually that just means to click controls (upper right corner and push the "Estimate IR delay" button and accept what it needs to shift to put that peak at zero. However, sometimes it's hard to figure out automatically and that would mean you have to shift the peak yourself, you can try and read the graph by zooming in and see where the peak is. Next you can use the t=0 offset (ms) box to shift the peak left (positive numbers) of right (negative numbers).

After aligning all (using the first big peak in each measurement) peaks to "0", then we can create a more decent vector average.

Once we have the vector average we can try and EQ that to the desired shape. I'd first use broad strokes (meaning PEQ's with low number Q's, Low would be 3 or less. Q=0.5 for instance works on a large wide area, Q=3 is working on a relative small area. You can also use REW's automatic EQ. For that you'd need to set a target and REW will try to EQ output to that shape.
Use the filters, something like 1/12 smoothing if needed to get a clearer graph. We're not going into too much detail yet.

Hope this all makes a little sense to you, the more you work on it, the more it will fall into place.

To get an idea of a target, you could use something like this for a start:
attachment.php

See the blue line

The numbers on the right should give you a similar shape to start with.
The only one you should not copy is the Target level (dB). The target level should be set to be about mid way of your measured result. This will mean you cut some energy at ~200 Hz and will need boost for the low end to follow the target graph, maybe some boost on the top end as well.

The target shown here is a crude approximation of the "Trained listeners" graph from Harmon:
attachment.php


If you haven't got anything else to go by, this is quite a decent starting point. So the trick is to get the left average and right average to follow that blue target as best you can. The more left and right start to look similar, the better your imaging will become.
With any luck, the usage of the averaging and setting overall balance to follow that graph will get you quite a decent sound, tonally.

Later on you can deviate from it to your hearts desire, to suit your taste. We first need some kind of reference to get to.

This is turning out into one of my infamous long posts already, so I'll quit now. Just practice with your current measurements to get the hang of it.
The more you do it, the easier it will become.

Deal? :)

P.S. Don't go too wild with the EQ, don't try to boost any narrow dips. You can get away with quite high boost numbers with the arrays, especially when it's close to a corner, but that would need a side wall damping panel to work. As you have them further from a corner, don't try to follow the curve all the way down to 20 HZ by forcing it... Ideally we'd look at the distortion by using an EQ-ed measurement to see if we get it too far.
I use boost up to 14 dB at the low end, but fewer PEQ's placed close together add up quickly.

By placing your target midway between the high and low values you prevent needing too much boost. I cut about 10 dB of output at ~200 Hz.
 

Attachments

  • target.jpg
    target.jpg
    197.9 KB · Views: 430
  • toolecurve.jpg
    toolecurve.jpg
    52.5 KB · Views: 537
Last edited:

TNT

Member
Joined 2003
Paid Member
wesayso, I never saw your figures for how much EQ you do on your lines. Figures in ballpark to mine.

Nvphotos - my lines sit all the way into the corners. This is work in progress....

EQ: (PEQ no, P(eq), Freq, Q, Gain)

16 P 200 0,7 -2
17 P 250 1,5 -7
18 P 1800 1,5 5
19 P 150 1,5 3
20 P 60 0,6 16
21 P 7k 0,7 6
22 P 14k 0,7 6
23 P 14k 0,7 6
24 P 7.7k 8 -5
25 P 12,4k 7 -5,5
26 P 11,6k 2 -4
27 P 777 0,1 -3
28 P 100 8 -6 (questionable)
29 P 15k 0,6 3

Result in listening prosition: (+ no eq)
 

Attachments

  • LI.png
    LI.png
    100 KB · Views: 56
  • 2.jpg
    2.jpg
    154.1 KB · Views: 83
  • Untitled 4.png
    Untitled 4.png
    144.9 KB · Views: 73
Last edited:

TNT

Member
Joined 2003
Paid Member
Yes - there is lots more to do inc FIR for sure - and I will. But I fear that with the driver I used, it will never be really good in the top i.e. the peaks att 7,7 and 11,5. These are audible I'm afraid. And my cabinet is to small.

Still, I'm very pleased with them - but/and not finished !! :)

//
 
Life has taken over lately so I've mostly just been enjoying them. They are definitely the best sounding speakers I have owned.

I've played with their positioning more but keep coming back to the same spots away from the walls. I really want to find a spot near the walls/corners but I really like what I'm hearing in their pulled out position.

Wesayso, I though I had the gain structure figured out until I started typing this just now and my music started distorting like crazy. Ironic!
 
WOW!!!!!!!!

My whole family is still in awe! Finished the eq in REW and then finished calibrating it in Dirac. Best, most realistic, sweet sound I have ever heard. Currently going through my wife and I's favorite music. Will have to post graphs at some point... Until then, happy listening!!!
 
Are Subwoofers Needed With an IDS-25 Line Array?

Are subwoofers needed with line arrays? I feel like their bass performance with a 125 watt amp is pretty decent with music at a reasonable volume but I wouldn't want to push it and go into clipping. With a 200-250 watt amp, I would have no concerns with clipping and feel like you could really unleash their potential. Personally, I think a 250 watt amp would make a better upgrade purchase for the arrays than a subwoofer

I'd love to hear others thoughts on this topic.
 
Good to read you're having so much fun with these arrays. I will never forget all the fun I had at the early stages. It is most impressive what these little drivers can do together.

I guess it depends for a large part on the size of the room. If the array gets enough support from a nearby wall or corner in a relatively small room, they can sound pretty powerful. Even better if it isn't fully symetrical and you can balance the bass between left and right.
Bass can be quite impressive. For music this may be all you need.

The difference between a 125 watt amp and a 250 watt amp is only 3 dB. But this little difference might be enough to prevent some clipping here and there. I have found the boundaries with my arrays a couple of times. For the longest time I only had a 100 watt amp. The arrays were not the real culprit. Often it was digital clipping which is quite obvious if it happens. Balancing the left and right channels to prevent that clipping made for a balanced setup that could take almost anything I threw at it. Save for the occasional song that had a lot of energy below 30 Hz.
I moved up to a powerhouse, probably somewhere around 350 watt into 8 ohm. But it's also quite different in topology compared to my old amp. It is more than impressive what this combo can do together, the arrays and this amp. I do feel I have a bit more wiggle room and haven't run into (digital) clipping anymore.

I will still go for subs, they are almost finished. But it isn't because I really felt I needed it. I'm doing it for two reasons primarily. One is headroom and the other curiosity. The curiosity part being: can I improve the sound even further with the added subs?
 
I'm still reading along :).

The target shown here is a crude approximation of the "Trained listeners" graph from Harmon:
752032d1556137426-obtaining-white-whale-25-driver-ids-25-line-array-build-toolecurve-jpg


If you haven't got anything else to go by, this is quite a decent starting point. So the trick is to get the left average and right average to follow that blue target as best you can. The more left and right start to look similar, the better your imaging will become.
With any luck, the usage of the averaging and setting overall balance to follow that graph will get you quite a decent sound, tonally.

Later on you can deviate from it to your hearts desire, to suit your taste. We first need some kind of reference to get to.

I created a configuration of each of these target curves in my miniDSP and can select them each on the remote.

It is very enlightening to scroll through them. Overall, I fit into the "all listeners" category but "trained listeners only" is growing on me. For some reason, the "all listeners" target sounds more holographic than the other 3.
 
Last edited: