Loaded file in to LA and noticed M2_BaseConfig_BSS is set to run at 48KHz (vs 96).
Was there a rationale stated that I missed (or forgot)? e.g., limited processing resources
or common file that works on lowest common denominator hardware or someone thought
it sounded or measured better? Just curious at this point, as most of my hardware
remains in the 'ordered' state.
I don't know. It's possible that they set it that way so that it will run on a Blu-80, that has the slower 1X DSP. I noticed that the file name at download says Blu-800, but the file is for the Blu-80. They have the ability to change the device type when you load the file.
That's what I was thinking... setting up the LCR M2 config for 96KHz on a BLU-160 only gets the resource usage meter up to 38%
(no room or user eq)
Don - there's something between your mic and the speaker that's corrupting the impulse......there's no clear separation between the first reflection and the speaker pulse. I'd expect the speaker impulse to go to zero in the first couple ms. I doubt the M2 wg is that bad in the time domain.
I pretty much figured that out, since posting those sweeps. To be fair, I did tell you guys that it was probably bad data. And if you look at the first picture I posted on this thread you'll see what was happening. The room still looks just like that. I didn't realize that I needed to pull all of the furniture etc away from the speaker before starting. Don't worry, I'll eventually do this the right way, after I've exhausted every other option. And thanks to everyone who's trying to explain these things to me. I need it.
Originally Posted by Don C
I remember when I started taking measurements for Project May... my initial measurements were pretty much useless... there is a learning curve.
Widget
We all start somewhere! I started by using a radio shack spl meter and "designed" my crossovers with in room, ungated, measurements. I was totally clueless....
Today I tried to collect more data, and do a better job than last time. I moved most of the furniture out of the room and put the speaker near the center. I elevated the speaker by putting the woofer on top of another speaker, the putting the horn on top of both. I had about 40 inches to the floor from the base of the horn, and about 8 feet to to the wall that the speaker was facing.
I measured and set the mic 1 meter from the horn throat screen.
here's what it looked like:
The gate was set to 5ms. The red is with no eq applied, the blue is with the standard M2 eq settings, and the green is what I had after manually adjusting the settings for a while. I moved the filter that JBL had set at 3.1 khz, down to 1.7 khz, and removed filters at 9 and 11 khz that were not helping, and made a few other changes I didn't record. The result is more neutral than I had before, but also duller. It needs to have a target curve adjustment to perk it up a bit. My room is pretty dead sounding, intentionally. I have carpet and drapes, and something soft hanging on most of the walls. So it's no surprise that the room response curves down a bit at the top.
The diaphragm seems to sufffer from the usual break up at 4kHz. I don't know what is the standard JBL eq, but perhaps you could try a bandpass filter w/ Q= 1.0 (which you can adjust up or down), centered at around 17 o 18Khz with db bump of your choice (most likely around 12db)
Measured with 2452-SL?
This is what the filters look like now. There are some others in an input filter, but they are small.
Yes, these are 2452H drivers, very recently purchased new, but with used from ebay 16 ohm coated diaphragms installed, because they sound better.
I also fooled with the delay for a while, using a method I've seen described elsewhere in this forum. I set the phase invert on, and adjusted the delay up and down looking for the biggest null at the crossover frequency. Using that method, the best delay for this speaker is zero.
Hi Don_C,
Interestingly, the shown data about M2 behavior differ much relative to the measurements our forum member RoB
http://www.audioheritage.org/vbullet...?600-sebackman
has shown us, using 2451Be drivers.
http://www.audioheritage.org/vbullet...l=1#post373896
I think that you have to enhance frequencies over 8kHz, as JBL have done
Regards
ivica
I really need to work on a bigger computer screen, and remember to decrease the scale so that I can see the variation better. When I was looking at the green graph on the laptop it looked close enough to perfect that I decided to leave it alone. Later I changed the scale and looked on a desktop PC, and the need for further improvement is pretty obvious. I'm not going to do this again this week though.
for the few that care: BLU-160 wasn't immediately recognized by London Architect (with laptop ethernet address to same subnet).
I turned off all the windows networking crap except Ipv4 and fixed the device speed to 100M/duplex rather than autonegotiate.
Tools like nmap then found the BLU-box (no crossover cable needed). There's a status web page if you point your browser to the
BLU-box IP address.
Then fired up LArch... bingo. I expect that some of this was power/boot order dependent, but now it works and I can play.
Proper output card and terminal block connectors ordered.
Config I'm setting up should allow for analog -and- digital I/O (4 in, 4 out of each).
Horns are due next week
I was noticing that every time I turned the system on it sounded different from the last time I worked on it. This turned out to be my bad habit of bypassing the filters using the blue buttons along the right side of the eq screen. When I would cycle the power the individual filters would be active again. So it works better to set the gain to zero instead of using the bypass button. It remembers that setting. A firmware bug?
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote