Page 30 of 58 FirstFirst ... 20282930313240 ... LastLast
Results 436 to 450 of 864

Thread: Project M2 DIY Thread

  1. #436
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    I overlooked the fact that you could enter exact values from the keyboard

    The LF curve looks very smooth here (the 254Hz EQ peak is completely smoothed out), but it is hopefully only a visualization artifact.
    It would probably be good to measure the actual output of the DSP to check this.

    You could also check the IIR implementation quality by playing a 21Hz tone, rigth on the EQ peak, looking for fullrange (quantization) noise using a pair of headphones.
    Not all biquad implementations are born equal in this regard, but I doubt QSC would fail that test.

  2. #437
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65

    Measurments

    The measurements are in, well from one M2 at least.

    This took a bit longer than expected, mainly because I borrowed a minidsp OpenDRC-AN from a friend so that I could see how the Q-SYS DSP compared to it, that meant that I had to wait a few weeks for the OpenDRC-AN to become available.

    The REW files are here in my dropbox

    The three sets of measurements are:

    Q-SYS DSP + FIR 4096 taps + PEQ
    Q-SYS DSP + FIR 6144 taps
    OpenDRC-AN + FIR 6144 taps

    The Q-SYS DSP's output looks fairly close to that measured with the minidsp which is exactly what I was hoping for. Although a trained eye may see something I don't.

    The measurements were taken with the M2 set to large and Room EQ off.

    Name:  Screen Shot 2017-10-25 at 8.33.43 PM.jpg
Views: 1975
Size:  63.7 KB
    Name:  Screen Shot 2017-10-25 at 8.38.02 PM.jpg
Views: 1810
Size:  78.0 KB

  3. #438
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    It would be more accurate and repeatable if you could measure directly at the output of your openDRC and QSC.

    Still, there are strange things showing here:
    - openDRC and QSC have opposed polarity (not a big deal, but worth noting, especially if you try side by side comparisons)
    - LF response low-pass does not looks as sharp as it should, with only -20dB at 3kHz compared to 750Hz, whereas it should be around -70dB, but this is probably due to the sweep range limits
    - openDRC and QSC without IIR EQ have the same LF response and lack of "boost" around 40Hz
    - HF response is rising on the QSC, or the openDRC one is falling (probably not), either way there is a 6dB shelving difference in response (see attached screenshot)

    What rephase settings did you use for the openDRC and QSC 6144 taps (without PEQ)?
    Could you do measurements of the outputs of your processors, including "flat" ones without any correction?
    Also please use a complete sweep (10Hz-20kHz) for both measurements, or an MLS.

    Name:  HF.jpg
Views: 1883
Size:  96.2 KB


    Name:  LF.jpg
Views: 1866
Size:  88.1 KB

  4. #439
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65
    Quote Originally Posted by pos View Post
    It would be more accurate and repeatable if you could measure directly at the output of your openDRC and QSC.
    Not sure what you are requesting, I'm pretty much a newbie to audio measuring. Could you briefly explain what you mean by 'directly at the output'.

    Quote Originally Posted by pos View Post

    Still, there are strange things showing here:
    - openDRC and QSC have opposed polarity (not a big deal, but worth noting, especially if you try side by side comparisons)
    - LF response low-pass does not looks as sharp as it should, with only -20dB at 3kHz compared to 750Hz, whereas it should be around -70dB, but this is probably due to the sweep range limits
    - openDRC and QSC without IIR EQ have the same LF response and lack of "boost" around 40Hz
    - HF response is rising on the QSC, or the openDRC one is falling (probably not), either way there is a 6dB shelving difference in response (see attached screenshot)
    No doubt some if that is due to the deficiencies in my measuring methodology.

    Quote Originally Posted by pos View Post
    What rephase settings did you use for the openDRC and QSC 6144 taps (without PEQ)?
    Didn't make any changes in rePhase, just exported the file then loaded it to the DSP.

  5. #440
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    Quote Originally Posted by lesliew View Post
    Not sure what you are requesting, I'm pretty much a newbie to audio measuring. Could you briefly explain what you mean by 'directly at the output'.
    You are currently measuring the output of the speakers with a mic.
    What you should do in this situation is measure the output of the DSP, as this is what you want to measure, and leave the speaker/room/mic out of the equation.
    So you would simply connect the analog output of the DSP (either QSC or openDRC) to the line input of your soundcard, and unplug the amplifier, speaker and mic.

    Didn't make any changes in rePhase, just exported the file then loaded it to the DSP.
    But was it the rephase preset that I send you for your specific use (ie with provision for an additional IIR EQ point) or the more generic one found on the google doc?
    The openDRC and QSC without EQ would require the settings from the google doc.

  6. #441
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65
    Quote Originally Posted by pos View Post
    You are currently measuring the output of the speakers with a mic.
    What you should do in this situation is measure the output of the DSP, as this is what you want to measure, and leave the speaker/room/mic out of the equation.
    So you would simply connect the analog output of the DSP (either QSC or openDRC) to the line input of your soundcard, and unplug the amplifier, speaker and mic.
    OK, I'll work on this.

    Quote Originally Posted by pos View Post
    But was it the rephase preset that I send you for your specific use (ie with provision for an additional IIR EQ point) or the more generic one found on the google doc?
    The openDRC and QSC without EQ would require the settings from the google doc.
    The QSC 6144 & OpenDRC 6144 were both generated from the rephase settings contained in the google doc. No changes were made.

  7. #442
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65
    New measurements DSP output dropbox link

    Hopefully I did a better job of it this time.

    Name:  Screen Shot 2017-10-26 at 10.35.00 PM.jpg
Views: 1929
Size:  72.5 KB

    Name:  Screen Shot 2017-10-26 at 10.33.08 PM.jpg
Views: 1737
Size:  69.9 KB

  8. #443
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    This is much more usable indeed.

    A few remarks:

    • There appear to be some (measurement?) artifacts at 2kHz, 3kHz, 4kHz, etc..., showing as peaks in the HF measurements and even more so in the stopband of the LF measurements.
    • The bypassed measurement of the QSC clearly shows the shelving up high. Are you sure you don't have a filter of maybe some sort of compensation somewhere in your processing chain?
    • The LF 4096+EQ measurement looks spot on
    • the openDRC FIR is not the one from the google doc as it does not contain the high Q EQs around 200Hz. This version is to be used with M2 "clones" that do not have the exact same dimensions as the original


    If you decide to use the 4096 +EQ version for LF, then use the 4096 one for HF as the delays are matched.

  9. #444
    Senior Member
    Join Date
    Nov 2008
    Location
    Gothenburg, Sweden
    Posts
    761
    Sometimes when I measure with REW I also gets artifacts, like the smaller ones of the ripples seen here. Then I measure again, and it is not there anymore.
    If it happens, it is often the first run.
    I suspect my soundcard to be the perpetrator, but I do not know.
    Actually, my previous dsp, dbx4820, often gave strange LF-curves the first run after startup of new settings, seemed like it had to "optimize" or adapt some algorithm or something.

  10. #445
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65

    Cool

    Quote Originally Posted by pos View Post
    This is much more usable indeed.

    A few remarks:

    • There appear to be some (measurement?) artifacts at 2kHz, 3kHz, 4kHz, etc..., showing as peaks in the HF measurements and even more so in the stopband of the LF measurements.
    • The bypassed measurement of the QSC clearly shows the shelving up high. Are you sure you don't have a filter of maybe some sort of compensation somewhere in your processing chain?
    I do have a couple of minidsp DDRC-88A dirac boxes in the chain, they where definitely set to bypass but I guess that could be a possible cause of those peaks. I will try to take the 88A's out of the chain and see what happens. I have an Altitude 16 on order so I wont be using these much longer anyway.

    I also had to use a Beringer UCA-222 external sound card on my Macbook Pro Retina since apple dropped the line-in on these models, probable but less likely to be the cause of the spikes.

    Quote Originally Posted by pos View Post


    • The LF 4096+EQ measurement looks spot on
    Hooray to me, I got something right. So that little peak at 250Hz is nothing to worry about ?

    Quote Originally Posted by pos View Post


    • the openDRC FIR is not the one from the google doc as it does not contain the high Q EQs around 200Hz. This version is to be used with M2 "clones" that do not have the exact same dimensions as the original
    I'll take a second look at that, possible user error

    Thanks for putting up with me, its a good learning experience.

  11. #446
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    Quote Originally Posted by lesliew View Post
    Hooray to me, I got something right. So that little peak at 250Hz is nothing to worry about ?
    This peak is part of the correction, and should also be there on the openDRC measurement.
    Are you sure you used the settings provided in the google doc for the openDRC?

  12. #447
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65
    Here we go with another set of measurements, I still have not managed to track down the cause of the spikes, it does seem strange that they are occurring at exactly 1k,2k,3k,4k etc, it could be my Denon 7200 introducing these spikes, I'll worry about it if they are still there when I install the Altitude. I did remove the Dirac boxes from the signal chain and that made further improvements to the measurements. I also re-uploaded the correct FIR measurements to the OpenDRC for comparison.


    Name:  Screen Shot 2017-10-31 at 11.07.17 PM.jpg
Views: 1671
Size:  80.9 KB
    Name:  Screen Shot 2017-10-31 at 10.51.29 PM.jpg
Views: 1728
Size:  77.0 KB



    REW file Here

  13. #448
    Senior Member
    Join Date
    Nov 2008
    Location
    Gothenburg, Sweden
    Posts
    761
    Maybe it is the noise floor showing up as spikes? If something is wrong in the setup with gains? (just a theory)
    If you measure flat EQ-settings (i.e no EQs at all) and on lower gain-level, does it look good?

    Hmm, strange the peaks at perfect numbers. Maybe some sample-rate problem? Different input and output soundcards used?

  14. #449
    Senior Member pos's Avatar
    Join Date
    May 2007
    Location
    France
    Posts
    2,628
    Quote Originally Posted by lesliew View Post
    REW file Here
    The measurement is noisy but everything looks good

  15. #450
    Member
    Join Date
    Jul 2017
    Location
    Florida
    Posts
    65
    Quote Originally Posted by bubbleboy76 View Post
    Maybe it is the noise floor showing up as spikes? If something is wrong in the setup with gains? (just a theory)
    If you measure flat EQ-settings (i.e no EQs at all) and on lower gain-level, does it look good?

    Hmm, strange the peaks at perfect numbers. Maybe some sample-rate problem? Different input and output soundcards used?
    I've had to splice and join a few cables just to get these measurements done, they were only temporary cables so they don't look very pretty being trailed all over the place, the raised noise floor could be a result of some less than perfect connections. I did notice that when I took the measurement at a lower-gain there was definitely more noise in the signal.

    At least I have the DSP side of things right now, I will be getting a new processor in a few weeks time, I will go through this exercise again with balanced cables end to end just to be certain.

    All of these measurements were done using a macbook pro with external sound card, I have a tablet PC but that has a single 'USB C' connector so I would still have to use the same external sound card if I was to try that, the only other option is to borrow my sons laptop, pretty sure that still has all of the legacy i/o ports.

Thread Information

Users Browsing this Thread

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

Similar Threads

  1. JBL LN3 Project, *Official Thread
    By Nightbrace in forum Lansing Product DIY Forum
    Replies: 58
    Last Post: 07-13-2016, 03:08 PM
  2. Fundraising request; donations to Lansing Heritage Project May loudspeaker project
    By mikebake in forum Lansing Product General Information
    Replies: 7
    Last Post: 04-21-2011, 12:37 PM
  3. Project May, similar privat project?
    By Flodstroem in forum Lansing Product DIY Forum
    Replies: 14
    Last Post: 02-21-2007, 03:42 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
  •