Resolution Questions

G29

Member
Thread Starter
Joined
Jun 20, 2019
Posts
100
More  
Main Amp
Quad Stereo Amp
DAC
OKTO Research DAC8
Computer Audio
i7-6700, Debian 12, custom VST3 feeding CamillaDSP, REW, RePhase, custom Driver Level DRC-FIR with 128-bit internal (i)FFTs, 8 - 64-bit 256K tap FIRs
Front Speakers
Heavily modified Magenpan 20.1 with FIR XOs
Subwoofers
DIY OB/Dipole
Other Equipment
Full Mark Levinson Stack
Is there any reason why the "IR Windows" "Window Ref Time (ms)" is 3 digits and the "Impulse" "Offset t=0" window is 4 digits ?

When zoomed in, the "IR Windows" with only 3 digits will sometimes show a value of -0.000 and not indicate how far off the 4th digit is from 0.0000.

Can both of these windows be made consistent with respect to precision at 4 digits to the right of the decimal point so they reflect each other ?



49471



Also, what is the precision of the Spectrogram. This is a 1024K TAP 176.4kHz linear phase All-Pass filter made in RePhase. When zoomed in, the black "Peak Energy Time" trace isn't flat and vertical.

The black high frequency trace appears early (not centered), the black mid-bass trace is centered and usually the black bass/sub "Peak Energy Time" trace frequencies are slightly late.

Is there a reason or configuration setting for this that I am missing for a synthetic linear phase "all-pass-filter" ?

Thanks much.


49470
 

Attachments

  • 3 4 Digit Precision.jpg
    3 4 Digit Precision.jpg
    152.9 KB · Views: 15
There is no need for high precision on the reference time, it is only used to position the IR window. The t=0 time defines the origin for phase so needs high precision.

The spectrogram looks oddly noisy, so there may be some issue with the setup there, or perhaps that is how the response has been designed. I'm not clear what you are getting at regarding the all pass filter. An all pass is for changing timing (phase) without changing amplitude. If it doesn't change timing it isn't a filter at all.
 
... The spectrogram looks oddly noisy, so there may be some issue with the setup there, or perhaps that is how the response has been designed. I'm not clear what you are getting at regarding the all pass filter. An all pass is for changing timing (phase) without changing amplitude. If it doesn't change timing it isn't a filter at all.

I was trying to model a baseline "perfect/ideal" linear phase model with flat amplitude (for learning purposes and software testing).

It was created in RePhase at 176.4kHz, Hann window, middle centering, use closest perfect impulse, 256K taps, 64-bit float IEEE mono wav file. 256K taps does have a delay of 743.039ms.

Is there a better way to create an ideal model ?

Thanks much.
 
Last edited:
Do you mean a perfect impulse? You can create an impulse by using "Make measurement from filters" without any active filters.
 
Do you mean a perfect impulse? You can create an impulse by using "Make measurement from filters" without any active filters.

How might one "Make measurement from filters" without any active filters.? What menu option/window should be used to get to this feature.

Thanks much.
 
Back
Top