Home Theater Forum and Systems banner

1 - 13 of 13 Posts

·
Premium Member
Joined
·
6,711 Posts
Discussion Starter · #1 ·
REW V5.13 has now been released.

Changes:
  • Reverted to requesting 512M RAM on startup, can be increased by editing roomeqwizard.vmoptions if required
  • Added optional frequency-dependent windowing of the impulse response, with default settings in the Analysis preferences.
  • Added a psychoacoustic smoothing option
  • Added an Equivalent Rectangular Bandwidth smoothing option
  • Improved the Gaussian approximations in the smoothing algorithms
  • Changed the spectrogram background colour to match other graphs
  • Added an option to normalise the spectrogram plot to the peak level at each frequency
  • Added an option to plot the peak energy arrival time curve on the spectrogram
  • If the peak of an imported impulse response is near (within 1%) of the start or end of the file, rotate the response by half its length. This allows import from packages that put negative time data in the second half of the response.
  • Export measurement as text will export unsmoothed, linear spaced measurements at their native resolution. This produces very large files, for many uses it will be appropriate to apply smoothing to a measurement before exporting it, in which case it will be exported at a multiple of the smoothing octave fraction.
  • Bug fix: When saving miniDSP filter settings on OS X the filename included the path
  • Bug fix: When starting up with ASIO drivers selected SPL meter calibration status could be forgotten
  • Bug fix: Cursor frequency was incorrect on room sim graph when the graph extended to 10 Hz
  • Bug fix: Graph cursor values did not update when the graph scaling changed
  • Bug fix: Air temperature spinner arrow buttons on Thiele-small parameters dialog did not change temperature
  • Bug fix: Trace arithmetic merge did not work if measurements being merged were imported responses with magnitude data only
 

·
Registered
Joined
·
261 Posts
Nice release :)

is it possible to adjust the FDW on an existing measurement? Perhaps I am being blind but I can't see the controls anywhere except the defaults screen.

EDIT: I am blind
 

·
Registered
Joined
·
21 Posts
Cheers John.

Bug: With multiple measurements open, when closing the measurements one by one in a reasonably quick fashion, starting from top to bottom, after closing 2 measurements, it is not possible to close the next measurement. At this point, some other measurements can be closed, others cannot. Basically a bunch of funky stuff happens.

edit: Win 8.1 64bit

This bug is present in beta 8 also. I actually came here to post about it in the other thread.
 

·
Registered
Joined
·
21 Posts
Bug: In the spectrogram, changing Mode, or Freq. Resolution:, or Window (ms), or Time Range (ms), resets the top and bottom axis limits.

edit: Nevermind, I should learn to RTFM.
 

·
Registered
Joined
·
261 Posts
I don't think FDW is respecting window ref time.

If I apply FDW to a measurement taken with a loopback then I get a severe rolloff at the HF

fdw_1.jpg

I then Estimate IR Delay/Shift IR and reapply the window to get

fdw_2.jpg

the 2nd one agrees with FDW seen in other apps (e.g. Holm).

sample measurement to demonstrate - View attachment fdw_test.mdat
 

·
Premium Member
Joined
·
6,711 Posts
Discussion Starter · #7 ·
In the file the window ref time was set to 2.9 ms, but the IR peak is at 3.3 ms. Setting the ref time to the peak time gives the expected result.
 

·
Registered
Joined
·
261 Posts
In the file the window ref time was set to 2.9 ms, but the IR peak is at 3.3 ms. Setting the ref time to the peak time gives the expected result.
OK thanks. I now notice the help says

If Add frequency dependent window is selected the window is applied after first applying the selected left and right windows. The FDW is centred on the window reference time - for best results this should be at the peak of the impulse.

so I guess I should read the manual properly (I see the forum software doesn't like a common acronym for that concept!) :help:

why is this required btw? perhaps I am misunderstanding what window ref time really means?
 

·
Premium Member
Joined
·
6,711 Posts
Discussion Starter · #9 ·
The ref time is just REW's reference for where the the windows should be located, end of the left window and start of the right window. An FDW gets narrower as frequency increases, so if it is not centred on the IR peak at higher frequencies the peak will be excluded which will mean HF is rolled off. I could have forced the FDW to always be centred on the peak, but there are reasons a user might want it elsewhere so using the ref time provides flexibility. With REW's default settings the ref time is at the peak.
 

·
Registered
Joined
·
261 Posts
I'm seeing a lot of hangs on startup with 5.13

The behaviour I'm seeing is that REW launches normally and if I mouseover a menu item immediately then the menu highlights but then it is just freezes and the GUI is completely unresponsive but it doesn't white screen in the way a windows app normally does. It does not respond to closing the window either (using the X close button) so I have to kill it.

All the log file says is

Aug 26, 2015 10:17:03 PM roomeqwizard.RoomEQ_Wizard main
INFO: REW V5.13 running JRE 1.8.0_45 32-bit on Windows 8.1

jstack wouldn't attach to the PID

process explorer reports that some thread is stuck on svchost.exe and that the corresponding stack is

ntdll.dll!NtAlpcSendWaitReceivePort+0xa
wow64.dll!Wow64SystemServiceEx+0x3cf
wow64.dll!Wow64SystemServiceEx+0x116
wow64cpu.dll!TurboDispatchJumpAddressEnd+0xb
wow64.dll!Wow64LdrpInitialize+0x23a
wow64.dll!Wow64LdrpInitialize+0x172
ntdll.dll!LdrInitializeThunk+0x12b
ntdll.dll!LdrInitializeThunk+0xe
ntdll.dll!NtAlpcSendWaitReceivePort+0xc
RPCRT4.dll!I_RpcSendReceive+0x1fa
RPCRT4.dll!I_RpcSendReceive+0xab
RPCRT4.dll!NdrServerInitializeNew+0x83f
RPCRT4.dll!I_RpcSendReceive+0x49
RPCRT4.dll!NdrSendReceive+0x2b
RPCRT4.dll!NdrGetBuffer+0xd9
AUDIOSES.DLL!DllGetClassObject+0x2535
AUDIOSES.DLL!DllGetClassObject+0x24bf
AUDIOSES.DLL+0x8954
wdmaud.drv!wodMessage+0x11cc
wdmaud.drv!widMessage+0x18a2
wdmaud.drv!modMessage+0x252
wdmaud.drv!wodMessage+0x4179
wdmaud.drv!wodMessage+0x3ebe
ntdll.dll!RtlInitializeExceptionChain+0x8f
ntdll.dll!RtlInitializeExceptionChain+0x5a
 

·
Registered
Joined
·
21 Posts
Bug: In all of the save commands that I have used (measurement as text, or whatever), you can overwrite some file the first time, by double clicking that file in the save as dialog.

The next attempt to overwrite the file with a double click, will close the save dialog box without the file being modified. Clicking on the previously saved file won't actually load the file in the preview pane. To defeat the bug, you can select some other file, then select and overwrite the file of interest without issue. On the next attempt, the bug has returned.

Manually entering the filename will overwrite the file, but won't defeat the bug on next attempt.

This bug has been present for some time.
 

·
Registered
Joined
·
1 Posts
Hello,

thankyou for this release. I really appreciate the option to normalise the spectrogram plot to the peak level at each frequency. Actually we had been discussing in another forum for quite some time when this would be applicable to waterfall spectra and I encourage you to add this feature to waterfalls in v5.14. In directivity plots it has been a standard for a long time and it helps tremendously to differetiate room modes from resonating structures in the room.

Thanks a lot for your good work!
rumpeli
 

·
Registered
Joined
·
21 Posts
EQ is struggling on these measurements. The screenshot is from an average of the two measurements.

https://dl.dropboxusercontent.com/u/34113196/Audio Stuff/REW/Left.mdat
https://dl.dropboxusercontent.com/u/34113196/Audio Stuff/REW/Right.mdat

edit: If I raise the target level to 78dB, two filters are created @ 495hz, each with Q=5 and gain = 1.9. This sort of defeats the Individual Max Boost don't you think?

I expected the attachment to be resized. What is forum etiquette regarding image sizes?
 

Attachments

1 - 13 of 13 Posts
Top