why don't we graph DACs ? ( i did !)

[Joachim Herbert]

I struggle for words.

 

[solderdude]

i was looking for a way to get objective data to confirm my subjective impression when switching dacs on my iem

The main reason for the difference in sound is the used reconstruction filter and probably also the output R of the headphone section of the DAC-amp combo.

From what I can see the difference should be audible which your attempt did show... mission accomplished.

Without a shadow of a doubt the ifi would be the most 'truthful' to the recorded music .... but .. you may well have a preference for one of them (in combination with the IEM)

 

[amanieux]

The issue still remains though, too many unknowns. How is the IEM fixed to the laptop?

the chain is simple Pc->DAC->iem(through bal cable)

 

[wwenze]

This is precisely why people who perform experiments have control samples.

Simply remeasuring 5 times with one DAC does not count as a control sample. You would have to redo the entire procedure for the first unit again. i.e. If you disconnect DAC A and connect DAC B, go out for a coffee and return to measure DAC B, you will have to redo the whole procedure when measuring DAC A again. Observe the variation between the 2 sets of DAC A measurements, record that as natural random variance. Then if DAC B measurements lie within the variation range, it does not count as a real difference.

The coffee part is half joking, sometimes stuff like that do make a difference. Like maybe when you open the door the airflow shakes your IEM slightly.

 

[amanieux]

The main reason for the difference in sound is the used reconstruction filter and probably also the output R of the headphone section of the DAC-amp combo.

From what I can see the difference should be audible which your attempt did show... mission accomplished.

Without a shadow of a doubt the ifi would be the most 'truthful' to the recorded music .... but .. you may well have a preference for one of them (in combination with the IEM)

when i switch dac i hear a difference but when i switched the filters on dawn 4.4 and heard no difference (graph above of different filter show a difference in the high frequencies my untrained ears cannot catch, unlike the difference in bass level between the 2 dacs that my ear can hear)

 

[amanieux]

This is precisely why people who perform experiments have control samples.

Simply remeasuring 5 times with one DAC does not count as a control sample. You would have to redo the entire procedure for the first unit again. i.e. If you disconnect DAC A and connect DAC B, go out for a coffee and return to measure DAC B, you will have to redo the whole procedure when measuring DAC A again. Observe the variation between the 2 sets of DAC A measurements, record that as natural random variance. Then if DAC B measurements lie within the variation range, it does not count as a real difference.

The coffee part is half joking, sometimes stuff like that do make a difference. Like maybe when you open the door the airflow shakes your IEM slightly.

you are right i am not ready to publish in a scientific journal the answer to the question i asked is that there are too many combinations when combining dac+iem, i was hoping for a simple answer to help us choose a dac for a given iem

 

[solderdude]

Consider that ears are not as discriminate as most people think they are.

Also when the treble is rolled off for most people this sounds like the sound has more body/bass and not as if it has less treble. Others may hear the same system as having less treble.

What you can do is measure the DAC with different filters (at least 1min of recording noise) and then overlay they plots. This will tell you if the filters are actually changing the sound in an audible way.
The 3 second files are way too short to really say something of 'value'.

 

[solderdude]

the chain is simple Pc->DAC->iem(through bal cable)

What thulle asked was how was the iem connected to the ADC (soundcard)
I just assumed you stuck the IEM in some tube and a small microphone in the other end and connected that to the mic input of a laptop/PC/tablet or interface.
That would at least explain the results.

 

[MaxwellsEq]

@amanieux at least you are trying, so well done for investigating and discussing.

When conducting experiments, you need to ensure you are only testing one variable at a time. You need to eliminate anything else that might contribute to variations in the thing you are measuring.

 

[Matias]

But in his case, isn't the variable only the dongle DAC being changed? The IEM is the same AFAIK. It just shows how a real world load, that is an IEM and not a resistor, behaves with those 2 DUTs. Of course it is not representative of other IEMs with those dongles, but for that particular test, I see it ok, although limited in use for generalizations with other IEMs.

 

[voodooless]

isn't the variable only the dongle DAC being changed?

We don't know that. That's why there was he question: how did you hook up the ADC/micriphone.

If he just put the EMI against the microphone by hand, then only the fact that he can't precisely match them up at the exact position could explain all the differences.

 

[DVDdoug]

One additional thought... Noise is randomness. (Pink noise is filtered randomness.) That makes it hard to get precise measurements, or to get the same-exact results twice.

You are getting obvious differences with your setup so that's not the main concern here but a frequency-sweep can give much more accurate-precise frequency response.

 

[amanieux]

We don't know that. That's why there was he question: how did you hook up the ADC/micriphone.

If he just put the EMI against the microphone by hand, then only the fact that he can't precisely match them up at the exact position could explain all the differences.

yes i did put the eartip on the mic hole on the laptop (seal comes from the silicon eartip elasticity), redid the sampling 3 times to make sure it was not changing when i reapply the eartip on the mic hole

 

[voodooless]

yes i did put the eartip on the mic hole on the laptop (seal comes from the silicon eartip elasticity), redid the sampling 3 times to make sure it was not changing when i reapply the eartip on the mic hole

That's not good enough. You need more hands to do this properly

The IEM needs to be at exactly the same spot for both DAC measurements. Removing it to switch the DAC and then placing it back is not precise enough.

 

[GXAlan]

That's not good enough. You need more hands to do this properly

The IEM needs to be at exactly the same spot for both DAC measurements. Removing it to switch the DAC and then placing it back is not precise enough.

+1

What you (@amanieux) can do is to tape the IEM and microphone to the laptop or table. Then you can manipulate the cables/wires without moving the mic/IEM interface/distance. It's a good thought to try to measure the output of two very different DACs. The reconstruction filter may introduce a ton of IMD.

 

[voodooless]

And while you're at it, use REW for measurements. Much better than noise!

 

[solderdude]

One additional thought... Noise is randomness. (Pink noise is filtered randomness.) That makes it hard to get precise measurements, or to get the same-exact results twice.

That's exactly why I suggested to run a long test (2 mins) but now I get why that would be difficult as he is pressing the IEM against an internal mic.
This can easily 'modulate' bass response.

I have done FR tests (with headphones) using white noise and a sweep and when you record the noise long enough (so one can get a smooth average) I get the exact same response as the sweep. A sweep, of course, is much faster to do and more reliable.

The reconstruction filter may introduce a ton of IMD.

That IMD will be dwarfed by the peaks in the noise so won't be visible.

And while you're at it, use REW for measurements. Much better than noise!

That may not be the case with these filterless NOS DACs.
It may depend on the speed of the sweep and the sampling frequency.
Averaging the noise (over minutes) may give a more accurate view of the average tonal balance because of the sampling frequency and sampling points.
You could also do a peak measurement over the noise signal.
When the sweep is low enough in speed, of course, the sweep will be more accurate.

 

[Matias]

One additional thought... Noise is randomness. (Pink noise is filtered randomness.) That makes it hard to get precise measurements, or to get the same-exact results twice.

Pink noise is perfectly OK for moving mic method MMM on room measurements.

That's exactly why I suggested to run a long test (2 mins) but now I get why that would be difficult as he is pressing the IEM against an internal mic.
This can easily 'modulate' bass response.

From my experience doing MMM measurements in room, after a few seconds and some 30 samples the curve is quite static and not changing anymore. I doubt that adding hundreds of samples during 2 minutes would change the curve much.

 

[voodooless]

That may not be the case with these filterless NOS DACs.
It may depend on the speed of the sweep and the sampling frequency.

Why would that be? You’ll should still see the droop. And probably it is not really filter less. Probably some emulation, and some are really strangely implemented.

 

[voodooless]

From my experience doing MMM measurements in room,

There is no room involved though