Page 3 of 13 FirstFirst 12345 ... LastLast
Results 31 to 45 of 181

Thread: Experiments with M2 horns.

  1. #31
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    I understand the idea, that I want to isolate the initial response and separate it from the reflections that arrive later. This doesn't seem to be available in the software that I'm looking at though. Maybe I'm having a simple language issue, I searched the help file for "gated" and found nothing useful. What does REW call this, and what settings work best?

  2. #32
    Senior Member
    Join Date
    Nov 2008
    Location
    Gothenburg, Sweden
    Posts
    761
    Here is a link to guidelines how to do gated measurements in REW:
    http://www.minidsp.com/applications/...r-measurements

  3. #33
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,720
    A few thoughts on the measurement topic.

    Gated measurements (sometimes also called quasi-anechoic) are definitely the way to go at higher frequencies, but unless you can suspend the speaker 10 or more feet from the floor and the other room surfaces including the ceiling are at least that far away from the speaker (meaning a greater than 20' high room) your measurements will only be accurate into the midrange.

    For in-room measurements and especially those where you want to compensate for room modes etc. taking multiple measurements in a cloud is really the best technique. I use 8 measurements in a 24" envelope around the sweet spot.


    Widget

  4. #34
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    Today I went through the activities I had planned yesterday. I made some new sweeps, with all EQ curves bypassed to start with. I have the old curves saved as a preset so I can get them back any time I like. One sweep was made from right next to my chair, and two more only a foot forward and back from there. I averaged those and generated EQ filters from the result. REW made 20 filters for me, and I put those into the BSS. This did a good job of flattening out the midrange, but REW didn't even try to flatten out the CD slope down at the top or fix the bass at the bottom. I don't know why. Maybe 20 filters is all it can make. But I made a few high and low shelf filters manually and got it pretty flat. I listened to this for a while and it's not a big change from how it was before I started, which is to say, it's excellent.
    I tried making some gated measurements as well. Thanks for linking to the method, Bubbleboy. I had the mic at ear level, and ended up raising it a bit to get the first reflection over 5 ms. I generated filters for that too. The problem I ran into, was that you can't average two or more sweeps to get this filter set, it needs to work from a single sweep. The result doesn't sounds as good to me as what I had before starting, So I reverted. I'll fool with that method more next week. Here's what It looked like with a single sweep next to the chair when I reverted and called it a day. Still a lot of treble there, that can easily be corrected to taste.

    Name:  26thfinal.jpg
Views: 1473
Size:  51.8 KB

  5. #35
    Senior Member Lee in Montreal's Avatar
    Join Date
    Sep 2010
    Location
    Montréal
    Posts
    2,487
    Since the horns is what can take the most work, then can they be moved outside to avoid reflecftions? Preferably on a grass area. ;-)

    Lee

  6. #36
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    Quote Originally Posted by Lee in Montreal View Post
    Since the horns is what can take the most work, then can they be moved outside to avoid reflecftions? Preferably on a grass area. ;-)

    Lee
    This is impractical right now, as I need an audio cable from the PC to the processor, and all of the rest of the system has to be hooked up too. So everything would have to go outside. It would take an hour to get the whole system set up in the back yard, and another to get it back in place in the house afterwards. The Lexicon amps are heavy. Maybe someday in the future. Anyway, my yard is not large, and is surrounded by a tall board fence. Is anyone else even thinking of using 2452s? Seems like it's just me.

  7. #37
    Junior Member
    Join Date
    Jan 2013
    Location
    Rochester, MN
    Posts
    14
    If you're actually getting 5ms of clean data that should have sufficient resolution to work out everything from the xo up. For a wg like this I'd think a set of polar measurements are needed and then eq is based on consideration of the entire response. I shoot for flat on the listening axis as long as there aren't any other nasties in the front 40-50deg.

    Yeah, I think 20 filters is the max on REW. I do my work on the gated data separate from the eq I do at the lp which I restrict to below 300-400hz.

    Did you save the gated measurement? It'd be interesting to see the fr and impulse. What was the mic distance?

  8. #38
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    I didn't post the gated curve because I thought I had probably done something wrong. I could post it. My measurement was made with the microphone about 8 feet away from the speaker. Would closer be better for gated measurements? How close? Pointed right at the speaker?

  9. #39
    Junior Member
    Join Date
    Jan 2013
    Location
    Rochester, MN
    Posts
    14
    Unless you've got your speaker and mic 7' off the ground and have that distance clear in all directions (including up!) you're not actually getting 5ms of clean data. Just a guess here, but if you have the speaker and mic at 40" off the floor you're only getting 2ms of good data at a distance of 8'! That's a freq resolution of 500hz which is not very good. Here's an article that explains it much better than I can: http://www.hifizine.com/2011/03/refi...r-minidsp-2x4/ You can skip down a bit to the part that starts with "A Little Theory"

    In my room with 8' ceilings I get the mic and design axis of the speaker at 4' and the best I can do is about 45" from the speaker to the mic. This only gives me a little over 3ms of data. That's ok for 1khz+, but anything below that is heavily smoothed due to the decreased freq resolution. When I want to get serious I drag everything outside, hang the speaker off the back of the deck, and start annoying the neighbors.

    I'm curious to see the raw fr and impulse of the M2 wg.

  10. #40
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    OK, since you asked, here are the gated sweeps from yesterday. Keep in mind that I'm just now learning how to do this, and this probably shouldn't be considered as solid data. I was probably misinterpreting the impulse data, the 6ms bounce may be the back wall, not the floor. This is with all filters in the BSS bypassed. Nobody else has an opinion as to how far away the mic should be placed from the speaker when taking this kind of measurement?

    Impulse response:

    Name:  impulse.jpg
Views: 1393
Size:  38.0 KB

    3Ms gated response:

    Name:  3msgate.jpg
Views: 1400
Size:  48.2 KB

    5Ms Gated response.

    Name:  5msgate.jpg
Views: 1293
Size:  48.0 KB

  11. #41
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,720
    Quote Originally Posted by Don C View Post
    Nobody else has an opinion as to how far away the mic should be placed from the speaker when taking this kind of measurement?
    I always use 2m on tweeter axis from a speaker system. If I'm measuring just a single driver I will place the mic 1m from the driver or horn mouth to reduce the impact of floor or ceiling reflections.


    Widget

  12. #42
    Senior Member Don C's Avatar
    Join Date
    Apr 2003
    Location
    Santa Rosa CA
    Posts
    1,722
    Changing topic for a while, I was dissapointed to hear the BSS at first, not the audio output, but the noise that it's fans make. So I decided to look into disconnecting the fans, as others have apparently done. First, I looked inside. It seems that the brains of this BSS are an Altera Stratix 2 FPGA and two Analog Devices Sharc DSP chips.
    The FPGA is from a family introduced over ten years ago, a 70nm process platform. But I guess it's maker must still regard it as a good chip, it seems to still be in production, and is available at Mouser for $460.00 Ouch.
    The fans are four Sunon KDE0504pkv3s. I looked them up, and they are specced as some of the quietest 40mm fans available, at 18Dba each. So replacing them with something quieter probably wouldn't help. That's a lot of fans and It doesn't seem to me that the engineers would have put them there if they weren't needed.
    I wanted to see where the heat was coming from inside the unit, and used the low tech method of putting my finger on all of the chips and cans. There are air flow dividers inside, the the left two fans cool the left side only, and the right side air is directed to the power supply and main board. The whole left side, where the input and output cards live, seems to be nice and cool. It doesn't make sense to me that those cards have two fans cooling them. So I disconnected those two fans right off the bat, and that works great, half the noise. My theory is that if you had ordered the unit with digital cards over there, you would probably still need those fans.
    On the right side of the unit, the DSP chips were very warm to the touch, and the FPGA was hot, so hot that I couldn't leave my finger on top. Ouch again.
    I decided not to just disable those two fans on that side. For some reason, these chips don't have heat sinks. Reverting to my old overclocking habits I decided to add some to those three chips. I cut an older Foxconn P3 CPU heatsink into four parts, with the bandsaw. Then I glued them in place on the FPGA and DSP chips. I used Arctic Alumina heat sink epoxy adhesive. It conducts heat well. And those heatsinks are there to stay.
    I then disconnected one of the remaining two fans on the right side of the box.
    There is a lot less noise with this single fan setup, it's hardly noticeable if you walk away from the rack. Now it's about the same noise level as the satellite DVR that lives in the same rack. I'm looking at adding a foam filter to the remaining fan to further silence it. And now you guys may be wondering if I'm OCD or something, and why can't I just leave anything alone. Well it was bugging me, and now it's not, so to me it was worth the time.

    PS, For any fellow OCDs, to open the box you should have a Pozidrive PZ2 screwdriver available, and a 5/32 allen wrench.
    (@#$% Pozidrive!)

    Name:  hs2.JPG
Views: 1742
Size:  170.6 KB

  13. #43
    Senior Member ivica's Avatar
    Join Date
    Jan 2008
    Location
    serbia
    Posts
    1,703
    Quote Originally Posted by Don C View Post
    I didn't post the gated curve because I thought I had probably done something wrong. I could post it. My measurement was made with the microphone about 8 feet away from the speaker. Would closer be better for gated measurements? How close? Pointed right at the speaker?

    Hi Don C,

    Why not move your speaker with its back-side laying on the floor, and mount your microphone from the ceiling about 1m over the speaker baffle, (if the speaker depth is about 0.5m, then the mic would be at the height of about 1.5 m from the floor) so, for the higher frequency, most of the reflections would come from the ceiling, so delayed about 2m (about 6ms) or more depending on the room ceiling height. Under such situation, a measurement from, say, 400~500Hz and up would be correct, and usable. Then you can move the mic left and right of the horn axis, so off axis measurements can be get. I would suggest you to equalize the response for the 15~20 degs off-axis response as I would believe that would be very near the averaged HF system energy response, or can be a good starting point for work.

    regards
    ivica

  14. #44
    Senior Member grumpy's Avatar
    Join Date
    Oct 2003
    Location
    SoCal
    Posts
    5,742
    "And now you guys may be wondering if I'm OCD or something, and why can't I just leave anything alone. Well it was bugging me, and now it's not, so to me it was worth the time."

    Nope. It's just funny as I was thinking of doing exactly the same thing (well, maybe thermal grease vs epoxy, but yeah...).
    Thanks for being the guinea pig ... and documenting it !

  15. #45
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,720
    Quote Originally Posted by grumpy View Post
    I was thinking of doing exactly the same thing (well, maybe thermal grease vs epoxy, but yeah...).
    I know what you mean... I also have commitment issues.


    Widget

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. Dampening of ringing in horns, baked horns :-)
    By sebackman in forum Lansing Product DIY Forum
    Replies: 3
    Last Post: 12-30-2013, 12:14 PM
  2. Early Cinema Sound Experiments on DVD
    By Ducatista47 in forum General Audio Discussion
    Replies: 0
    Last Post: 03-06-2009, 10:36 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •