[SenyorC]

Ifi's former lead designer said something about how the stock power supply isnt enough for planars, i wonder if a high power linear psu would actually benefit..

I'm not going to doubt anything he says as he has forgotten more than I will ever learn.

All I can say is that I haven't had any issue with the planars I have. I don't have the Susvara or the HE6 or anything like that but with the Arya, HEkse etc. I can't say I have had any issues.

I think it is going to depend on what planar headphones they are, what kind of music you listen to and at what levels you listen to said music.

 

[Assationater]

I'm not going to doubt anything he says as he has forgotten more than I will ever learn.

All I can say is that I haven't had any issue with the planars I have. I don't have the Susvara or the HE6 or anything like that but with the Arya, HEkse etc. I can't say I have had any issues.

I think it is going to depend on what planar headphones they are, what kind of music you listen to and at what levels you listen to said music.

Lots of amps these days of like 10w per channel, its crazy. But personally i havnt really heard a night and day difference from dongle power to discrete 5w.

 

[123peter]

The years pass, my children grow and the channel imbalance in the IFI Zen series is still going strong.

 

[Assationater]

The years pass, my children grow and the channel imbalance in the IFI Zen series is still going strong.

i dont think its just the zen series, i assume the diablo 2 has it too

 

[123peter]

Diablo 2 too?
You've got to be kidding me, it's as contagious as Covid.

I know that I am 101% sure that there is no channel imbalance in IFI and there is no need to start
the potentiometer from the 9 or 10 o'clock position is IFI AUDIO PRO ICAN SIGNATURE - this one runs smoothly and precisely.

One day there will be an IFI Zen that will have a balanced potentiometer this is what it is written in ancient books.

 

[GlenAppleton]

Now, the XSpace doesn’t just add extra treble, although we do get a bit of a boost as we move past the 1kHz mark. The XSpace also drops the presence below that mark. There is also a sensation of things being more spacious, more “spread out” if you will. This could just be based on the change in frequency response but I did find on some other iFi devices that the XSpace made no difference to the frequency response on the graph but still gave a sensation of “space”, so maybe there are other things going on that aren’t just related to frequency response.

From what I understand (and I could be wrong), XSpace (formally 3D on the OG Zen CAN) also adds just a bit of channel bleeding between the right and left channel. This gives it more of a "spacious" feel as if you were listening to speakers since both ears would hear both speakers, just at different levels.

 

[SenyorC]

From what I understand (and I could be wrong), XSpace (formally 3D on the OG Zen CAN) also adds just a bit of channel bleeding between the right and left channel. This gives it more of a "spacious" feel as if you were listening to speakers since both ears would hear both speakers, just at different levels.

That is something that I actually forgot to measure on this device.

 

[GlenAppleton]

That is something that I actually forgot to measure on this device.

Understandable, since it's not at all a common feature among headphone amps.

 

[Thorsten Loesch]

In the case of XSpace, this is what we get:

This suggests one channel is wired/set out of phase/wrong polarity. With both channels in phase (same polarity)the response should be flat.

Thor

 

[SenyorC]

This suggests one channel is wired/set out of phase/wrong polarity. With both channels in phase (same polarity)the response should be flat.

Thor

Is this something done on purpose with XSpace?

I ask because when XSpace is off, nothing is out of phase. And the XSpace on this device is the same as that on the HFM Signature that I measured quite some time ago (allow slight differences for difference in equipment):

Yet there have been other devices where XSpace did actually measure the same as flat (the Go Bar was one of them IIRC)

Edit: yes, here is a measurement of IEMs connected to the Go Bar. There are two lines in the graph, one with XSpace on and the other off:

 

[Thorsten Loesch]

Is this something done on purpose with XSpace?

No. The curve you measured can only be gotten IF the two channels are out of phase.

I doubt it is inside the Zen Can. Miswired cable? Incorrect setting in the Software?

I ask because when XSpace is off, nothing is out of phase. And the XSpace on this device is the same as that on the HFM Signature that I measured quite some time ago (allow slight differences for difference in equipment):

Interesting. Again, you can only get this if one channel is out of phase. Or perhaps if you use decorrelated noise (where each channel has in effect a different signal).

Yet there have been other devices where XSpace did actually measure the same as flat (the Go Bar was one of them IIRC)

XSPace, 3D etc. are all the exact same form of crossfeed. It is my design based on work we did in the late 1980's at the RFZ in Berlin. They all should measure identical.

Conclusion, some where in the chain to the input of the Zen Can one channel is polarity inverted, or the test signal is not "corelated" introducing frequency dependent crossfeed.

Thor

 

[SenyorC]

No. The curve you measured can only be gotten IF the two channels are out of phase.

I doubt it is inside the Zen Can. Miswired cable? Incorrect setting in the Software?

Note that the measurement is only of one channel and all measurements were done with the same settings and cable, without disconnecting or reconnecting anything etc.

The flat, XBass and XSpace measurements were all taken one after the other. The same with the Flat, HFM and XSpace on the older device.

Cable is fine. I don't discard incorrect setting in REW but it is the same settings used on all the devices.

 

[Thorsten Loesch]

Note that the measurement is only of one channel and all measurements were done with the same settings and cable, without disconnecting or reconnecting anything etc.

So the second channel had no signal?

That in effect makes the second channel have a different signal than the one being measured.

So you are measuring the crossfeed. The inactive channel "feeds" a "no signal" signal to the opposite channel and in effect as a result shows a frequency response change in the channel tested. If you measure the second channel you will see the complementary signal there. That's how crossfeed works.

Here an example (this in not the same crossfeed as in 3D/Xspace) including expected measured result:

Measure with the same signal on both channels and you will see a flat response.

Invert the signal between the two channels to measure the crossfeed instead.

All this can be done in REW (I also use it, though mainly for speakers and headphones, I use RMAA and others for Audio Gear).

Thor

 

[SenyorC]

So the second channel had no signal?

Correct, sweep was only sent to left channel.

So you are measuring the crossfeed. The inactive channel "feeds" a "no signal" signal to the opposite channel and in effect as a result shows a frequency response change in the channel tested. If you measure the second channel you will see the complementary signal there. That's how crossfeed works.

Makes total sense.

My only couple of questions here are (please excuse my ignorance, I only know just enough to be dangerous )...

In the case if the higher ranges being boosted (above 1kHz), I guess that is part of an EQ, as there is no second channel signal to "get" that extra signal level from?

My second question is regarding the measurement of the Go Bar, which I don't have a direct line measurement of (at least I can't find one on my PC at the moment) but that measurement is done in the same way. Left channel frequency sweep only and only left channel measured. In this case, there is no difference in the measured response. Should this not exhibit the same response as on the other devices?

BTW, thank you for responding to all of this, I always appreciate learning!

 

[Thorsten Loesch]

In the case if the higher ranges being boosted (above 1kHz), I guess that is part of an EQ, as there is no second channel signal to "get" that extra signal level from?

As I mentioned, this is not your cooking "Bauer" crossfeed. The coefficients are based on actual generalised HRTF and will look different.

ALL Crossfeed IS an equaliser, it alters the frequency response of the "side" signal in Mid-Side (M/S) even if it is applied in the two stereo channels and not after a matrix conversion from L/R to M/S and then after equalising the "S" portion back to L/R).

The equalisation applied to the "Side" signal in my design is what it needs to be to provide the correct coefficients for the HRTF to simulate stereo speakers.

My second question is regarding the measurement of the Go Bar, which I don't have a direct line measurement of (at least I can't find one on my PC at the moment) but that measurement is done in the same way. Left channel frequency sweep only and only left channel measured. In this case, there is no difference in the measured response. Should this not exhibit the same response as on the other devices?

It should, unless something was not measured in this precise same way.

BTW, thank you for responding to all of this, I always appreciate learning!

NP

Thor