measuring new subwoofer build

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
so im playing with my new subwoofers, and using REW/umik1 to try and get them to sing in my room. i have a problematic room, and ive tried most of the acceptable locations and im not getting great results so far.

anyway, a major part of this issue comes from me being a noob with REW and not being very good at interpreting measurements. ive done measurements from the 3 main seating positions and run them through Multisub optimiser, but for now, id just like some feedback on my methodology and result.

attached are screengrabs showing response (all in room so far)
a) with mic under downfiring sub,
b) just inside sub port mouth (side firing ports)
c) 15cm from sub, between ports and woofer

and finally

d) response of same sub (at a higher level of output) at main listening position.

messy huh?

any advice? or are my subs, contrary to sims, rubbish?
 

Attachments

  • under woofer.jpg
    under woofer.jpg
    80.4 KB · Views: 210
  • vent mouth.jpg
    vent mouth.jpg
    78.5 KB · Views: 204
  • 15cm distant between vent and woofer.jpg
    15cm distant between vent and woofer.jpg
    87.7 KB · Views: 213
  • at MLP.jpg
    at MLP.jpg
    86.4 KB · Views: 206
I have never used rew but if i were you i would have started with a simple measurement like rta, started with 1/3 oct resolution and later increased resolution if needed. I would first measure the background noise in the room, saved that curve in the graph and then fired up the subwoofer, starting at low lovel and inrease level to a suitible level over the background noise. Adjust subwoofer response by ear while observing the response on the graph and learn what is what. Also play the system fullrange and adjust levels by ear while watching the graph. Pink noise is a fine stimuli but do check with music too, do adjustments and re-check with pink noise in a never ending process ��
 
thanks for the advice, particularly about the smoothing. i did fiddle with it, and it made my graphs look much nicer (also on my fullrangers) , but it felt like cheating! what is a typical level of smoothing used when discussing speaker response..?

i shall try with the RTA tool.

in any case, looking at my graphs, do you see a normal, functioning subwoofer, or something with poor performance?
 
ah, i also did the "subwoofer crawl" with one of my subs in the main seat. did a sweep with the mic in almost all the possible locations for a subwoofer (including a couple which are almost impossible due to cable routing) none of them seem to be very good.. see graphs..(not smoothed yet..!)

my main hope is to use the two subs together to get an even (ish) response around the room.
 

Attachments

  • 1.jpg
    1.jpg
    78.2 KB · Views: 206
  • 2.jpg
    2.jpg
    77.2 KB · Views: 51
  • 3.jpg
    3.jpg
    77.2 KB · Views: 52
  • 4.jpg
    4.jpg
    76.8 KB · Views: 50
  • 5.jpg
    5.jpg
    79.1 KB · Views: 42
  • 6.jpg
    6.jpg
    79.5 KB · Views: 43
  • 7.jpg
    7.jpg
    77.2 KB · Views: 37
  • 8.jpg
    8.jpg
    77.9 KB · Views: 37
  • 9.jpg
    9.jpg
    78.5 KB · Views: 45
Wonderful to see you proceeding so coherently and nice to see the data.

Hard to interpret without knowing anything about the sub. Is it a small box versus a big one and what crossover point and maybe something about the room? The impedance curve, while some fuss to plot, may also be revealing about how your driver is working with the actual box tuning (no need to calibrate the impedance, just get a 100-ohm resistor and lots of alligator clips and post the plot to see where the peak(s) are).

1/12 is a good compromise smoothing and necessary. A mic stand with a boom arm (or DIY creation) very helpful.

Phase is related to distance from cone and not too helpful. But it sure would be good to see THD (and maybe 2nd and 3rd) which REW produces at the click of a mouse.

Good to stabilize your research by focussing on a single chair and average of 3 readings (just above the chair backrest, left, right, and centre of the chair back, spanning say 20 inches). That operationalizes "at my ears" criterion.

That will be like the way I introduced my 17-foot TL-like box:

17 foot pipe sub 12-230 Hz ±5dB - diyAudio

Great stuff.

B.
 
Last edited:
Ok try to respond to as many questions as possible. Unfortunately i had to hard restart my pc, so will need to do all new measurements to get the thd readings, Which i cant do right now. Have to wait till im alone in the house!

I do remember thd was below noise floor, but quite close. apart from a small peak around 150 hz which went above. Both noise floor and thd were between 40 and 50db. There is a road with traffic in the distance. Subs are both small vented cabinets of 20 litres. with a 10 inch dayton driver optimised for small enclosures. Vent is tuned (in sim at least) to 28hz.

Seating position is a 3m long sofa in the centre of a room, facing longer wall. Longer dimension of room is 8m, shortest is 6m, ceiling is barrel vaulted, and 5m high. All plastered stone walls/cieling, with tiled stone floor. When room is empty it echoes like a church, now its full of stuff (plants, bookshelves, cabinets, desks, giant sofa etc). Which has helped, but since its still bare for the entire top half, including the curved cieling, intelligibility has always been a problem in this space.

I was hoping either a source/sink subwoofer setup, or some room mode manipulatuon via multisub optimiser might help a bit with the lower frequencies and some time alignment higher up might get the most i can from a poor space without fitting a false cieling or hanging matresses on the walls!
 
Nice results for a pretty small box.

THD plot sinking into REW's noise level just means at that loudness, you can't measure lower. 40-50 dB below fundamental is quite good (bit under 1% good for a cone speaker and should be noted by those who fret about .1% in their pre-amps), but your eye may have missed the region below 50 Hz.

Impedance may show details of tuning and sim.

Yes, room is large and good ceiling, but hardness is a big liability (like most of the rooms attached to this forum). Multiple subs can be a great benefit but better to make a different and larger second sub to tuck away in a corner. Directional controlled dispersion tweeters may help.

B.
 
Ha i think the thd was not normalised. Not 50db below fundamental, but 40-50db actual noise floor/ thd, so not so good ;)
THD and impulse response are often lost sight of at this forum esp. in the quest for bass from little boxes shaped like horns.

Some subtleties to THD plots. Normalized is the more sensible way to read THD. But where output sags a lot at low freq, not as meaningful.

THD plot gets better (and in a sense, is better) as speaker gets louder (because you are raising the distortion products up from the muck) up to the loudness where true speaker HD starts deteriorating. Prolly a lot of people live in places where their AM radio has better S/N than their room and 50dB better than their fancy pre-amp.

B.
 
ah, i also did the "subwoofer crawl" with one of my subs in the main seat. did a sweep with the mic in almost all the possible locations for a subwoofer (including a couple which are almost impossible due to cable routing) none of them seem to be very good.. see graphs..(not smoothed yet..!)

my main hope is to use the two subs together to get an even (ish) response around the room.

All of these measurements show the same major null around 28 hz. You said your couch is in the middle of the room facing the long wall. So 3 meters into the room facing your 8 meter wall? 3 meters from front and 3 meters from back? If so that corresponds to the 28ish hz nulls in all your measurements. I would suggest moving your couch 1 meter back to 4 meters from the front and 2 from the back and remeasure.
 
If so that corresponds to the 28ish hz nulls in all your measurements. I would suggest moving your couch 1 meter back to 4 meters from the front and 2 from the back and remeasure.
Whether or not 28 Hz relates to that listening spot when the sub is off the chair, middle is an unfavourable location.

I've never tried woofer-on-chair testing. But when I did 14 locations for my sub around the room by moving the sub, even bigger variations than OP found. And when the mains are included, even the polarity needs to be varied to see which way works best, no kidding.

B.
 
I have had good experience by moving subwoofers away from boundarys, according to some articles good locations are roughly at 1/4 length between opposite walls and floor/ceiling. The same applies for the listening location, roughly, near the center of the room gives no bass and near a boundary gives excessive bass and mostly so in tri corners
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.