Home Theater Forum and Systems banner

1 - 10 of 10 Posts

·
Registered
Joined
·
261 Posts
Discussion Starter #1
I picked up a focusrite scarlett 6i6 recently and ran a quick soundcard calibration to sanity check performance.

The FR looks normal but I don't recall seeing this sort of ripple in the IR before on other devices (or perhaps I haven't looked). Does this look normal? signal is REW -> focusrite ASIO -> line output -> line input -> REW

scarlett_cal_ir.png

scarlett_cal_fr.png
 

·
Premium Member
Joined
·
1,562 Posts
Yes, this is normal for the 2 Focusrite interfaces I have owned. On both units my IR looks the same as yours. The frequency response and phase has a small wave/ripple that causes this. The ripple is only maybe 0.1dB. This is inconsequential for any REW functions and can be ignored in my opinion. The REW calibration of the interface removes this ripple very well and provides a very flat response. REW does not add the correction into the IR however. If it did the IR would then look ideal. The appearance of the IR is not really a reliable indicator of system performance anyway. It can be very misleading. That is why we have all the other charts for analysis. I think the ETC and the Step response along with the IR still shows the minor influence of that uncorrected IR so that should be understood when looking at the those charts.

2i2 Ripple.PNG
 

·
Registered
Joined
·
261 Posts
Discussion Starter #4
Yes, this is normal for the 2 Focusrite interfaces I have owned. On both units my IR looks the same as yours. The frequency response and phase has a small wave/ripple that causes this. The ripple is only maybe 0.1dB. This is inconsequential for any REW functions and can be ignored in my opinion. The REW calibration of the interface removes this ripple very well and provides a very flat response. REW does not add the correction into the IR however. If it did the IR would then look ideal. The appearance of the IR is not really a reliable indicator of system performance anyway. It can be very misleading. That is why we have all the other charts for analysis. I think the ETC and the Step response along with the IR still shows the minor influence of that uncorrected IR so that should be understood when looking at the those charts.

View attachment 119306
Ok thanks for confirming this is to be expected. It seems a bit odd, I wonder what the source is?

fwiw I noticed it because i was doing some quasi anechoic measurements of a speaker I am designing and the IR looked unusually "jittery" in the early part of the impulse.
 

·
Premium Member
Joined
·
1,562 Posts
I don't know the source, but indeed it's too bad. It is a bit distracting to see on the IR even though with calibration it does not impact the main REW charts.

I am curious if the Steinberg's UR22mkII has a cleaner IR. If someone that uses that unit will post a loopback measurement (.mdat) of that unit at 44.1kHz, I would appreciate it (TRS to XLR loopback preferred).
 

·
Registered
Joined
·
261 Posts
Discussion Starter #6
I suppose an alternative correction would be to record that loopback sweep as an impulse, invert it and then convolve the measured response with that inverted response.
 

·
Premium Member
Joined
·
1,562 Posts
JohnM suggested the approach Here as one method. I have been meaning to try it as a one time experiment, but have not done it yet. It doesn't sound practical to use it regularly however.
 

·
REW Author
Joined
·
6,711 Posts
Oscillations like that in the IR are normal, they reflect the fact the response is not flat to Nyquist (which it can never be).
 
1 - 10 of 10 Posts
Top