REW issue, or MINIDSP issue?

[robbnj]

Hello all, me again. Tinkering, learning, and avoiding the 50MPH winds and negative double-digit temps outside.

I know the majority of users are taming bass frequencies with their setups, but I'm happy with what I've achieved, and am now spending some more time mucking about with my main speakers.
Getting better at taking measurements, interpreting results, learning how changes will sound, etc.

An odd thing that I noticed before, and confirmed tonight: After running my measured curve through EQ, saving the coefficients to a txt file, and importing into my MINIDSP software, it seems that some of the settings are being changed somewhere along the line.
Below is an example of an EQ curve I created for my right main speaker, selectively in the 200-6,000Hz range. If you look closely, REW suggests a correction of -6dB at ~2KHz, but when I look at it in the MINIDSP, it is showing as -3dB.

Same thing for the ~3.2KHz frequency. REW suggests a gain of over 8dB, but REW interprets that to be 5dB.

Anyone have any idea of what might be causing this? My MINIDSP is the 2x4, running the 2x4ADV software. REW version 5.3.9.
'Wondering if I should be manually entering the REW results instead of exporting the coefficients, and importing to the MINIDSP.

Cheers, and keep warm (to all my northern neighbors).

 

[Keith_W]

The reason behind that discrepancy is because REW is showing you individual PEQ's, but your MiniDSP is showing you the combined effect of multiple PEQ's.

In REW, click on "generate measurement from filters" and compare that with your MiniDSP. Or, in your MiniDSP, examine each individual PEQ. It should be the same.

The other possibility is that you generated biquads for the wrong model of MiniDSP, but I doubt if this was the case (if that had happened, the resultant PEQ's would be at different frequencies). But it doesn't hurt to check.

 

[JStewart]

@robbnj, I believe Keith is correct and the wrong equalizer has been selected. The 48kHz plug-ins equalizer is selected in the screenshot you posted. The MiniDSP you have operates at 96kHz.

 

[robbnj]

The reason behind that discrepancy is because REW is showing you individual PEQ's, but your MiniDSP is showing you the combined effect of multiple PEQ's.

In REW, click on "generate measurement from filters" and compare that with your MiniDSP. Or, in your MiniDSP, examine each individual PEQ. It should be the same.

The other possibility is that you generated biquads for the wrong model of MiniDSP, but I doubt if this was the case (if that had happened, the resultant PEQ's would be at different frequencies). But it doesn't hurt to check.

Gotcha' on the DSP showing the combined effect. I'll pay closer attention to that as I explore more and try the "generate measurement" suggestion.
I'm pretty sure I chose the right MINIDSP settings, but the other response above has me wondering.

I'll have to get th serial numbers off the units and see exactly what they are.

 

[robbnj]

@robbnj, I believe Keith is correct and the wrong equalizer has been selected. The 48kHz plug-ins equalizer is selected in the screenshot you posted. The MiniDSP you have operates at 96kHz.

Being older ones, I thought they operated at 48KHz, but I'm honestly not sure.
As I mentioned above, going to check serial numbers and see exactly what I have, versus what I assumed I have.

I will say that I subsequently keyed in the filters manually under "Basic" mode, did a quick sweep, and the response smoothed out nicely. I had to quit before I could do any real listening tests, but a quick portion of a known song sounded pretty good.
More testing tonight.

 

[Keith_W]

If you generate a PEQ for 2kHz at 48kHz, and implement it at 96kHz, it will double the value of the PEQ. So instead of a 2kHz PEQ, you now have a 4kHz PEQ. MiniDSP makes many models, some are 48kHz, some are 96kHz. I can't remember which models are which without having to look it up.

But comparing the PEQ's in REW with the PEQ's in your MiniDSP device console, it seems to me that the frequencies are all correct. Which is why I doubt if the PEQ's were generated for the wrong sample rate.

I am pretty sure that PEQ #3 (2004Hz / -6dB / Q=2.416) is at -3dB because of the neighbouring PEQ #4.

 

[JStewart]

If you generate a PEQ for 2kHz at 48kHz, and implement it at 96kHz, it will double the value of the PEQ. So instead of a 2kHz PEQ, you now have a 4kHz PEQ. MiniDSP makes many models, some are 48kHz, some are 96kHz. I can't remember which models are which without having to look it up.

The old 2x4s were 48kHz the newer 2x4HD is 96kHz For some reason I thought I ready 2x4HD in the OP. I just looked back at the OP that's not the case. Sorry guys.


.