The discovery thread!

[Hypops]

And for those that assume I am deaf or have no idea what to listen for, popping in a couple of db bass boost from the Gryphon or flicking on the cross feed from the PA10 stood out like dogs balls in a back and forth comparison. Any slight bass change or staging change for example between devices isn't lost in the test somehow, they are quite obvious if they exist.

I respect that you're committed to doing AB tests. But I wouldn't mistake those tests for objective proof. They're still a subjective judgment IMO.

Spoiler: Expectation bias

The critically important thing you're not controlling for in your AB testing is the impact of negative expectation bias ("I don't believe there will be a difference"). As is well documented, and as I'm sure you're aware, expectation bias cuts both ways equally. There is literally no way to avoid that bias if you are your own test subject. You don't expect a difference so you are much more likely not to hear one. You couldn't even have another person conduct the test on you (the proverbial "wife test") since you'd still know what the test is testing for. Your negative expectation bias would still be firmly in place. There are good reasons scientists don't conduct experiments on themselves.

For your AB test to be scientifically meaningful, you'd have to have at least two different subject groups: one group who are told that there is a difference and another who are told there is no difference. Then you'd have to run controls where each group is asked to compare two things that the experimenters know are audibly identical and then compare things that are known to be audibly different to regularly varying degrees. And then you'd conduct your AB test on each group. Otherwise all you're proving with an AB test is that you believe there is no difference, and your results are simply confirming your own expectation. That's not science. It's just proof that you're as human as the rest of us!

Different sources can sound perceptively different. Some sources sound very similar and can be difficult to distinguish from each other. Transducer sensitivity is obviously a factor as well.

Indeed. Harmonic distortion is real, measurable, and audible. So are things like jitter, crosstalk, etc. Different amps and other gear can sound audibly different. Whether or not those differences are big enough to matter to someone is a different question!

I'm one of those people who can hear differences between gear, but I don't care enough to start collecting amps and DACs. I have different devices for different use cases, not for their sonic differences. Once I reached a "good enough" performance baseline in each of those devices, I stopped shopping.

 

[BS5711]

I respect that you're committed to doing AB tests. But I wouldn't mistake those tests for objective proof. They're still a subjective judgment.

The critically important thing you're not controlling for in your AB testing is the impact of negative expectation bias ("I don't believe there will be a difference"). As is well documented, and as I'm sure you're aware, expectation bias cuts both ways equally. There is literally no way to avoid that bias if you are your own test subject. You don't expect a difference so you are much more likely not to hear one. You couldn't even have another person conduct the test on you (the proverbial "wife test") since you'd still know what the test is testing for. Your negative expectation bias would still be firmly in place. There are good reasons scientists don't conduct experiments on themselves.

For your AB test to be scientifically meaningful, you'd have to have at least two different subject groups: one group who are told that there is a difference and another who are told there is no difference. Then you'd have to run controls where each group is asked to compare two things that the experimenters know are audibly identical and then compare things that are known to be audibly different to regularly varying degrees. And then you'd conduct your AB test on each group. Otherwise all you're proving with an AB test is that you believe there is no difference, and your results are simply confirming your own expectation. That's not science. It's just proof that you're as human as the rest of us!

As I stated with the Clarinet test I actually did expect/hope to hear a difference because I was convinced that there was one based on normal listening.

No my tests are not hugely meaningful in a strict scientific sense and no they are not proof of anything except to me and yes there is something left on the table by way of doubt about possible very subtle nuances that I may have overlooked or simply could not hear. I presented them as my observations nothing more and I am aware of the limitations that you stated.

I did not suggest they were scientifically definitive in anyway way and I said several times that perhaps other folks might perhaps be able to pick up differences in the nuances at least.

However, doing a test within sensible limits of what is appropriate for a schmuck enthusiast at home is better than doing no tests at all and they proved to my satisfaction that if there is any difference that perhaps I might have missed or is beyond my hearing it is at a level that is immaterial to me.

For clarity, sighted comparisons always made me feel like there were differences, just like everybody else talks about, blind comparisons no.

The upshot was simply that I was encouraging others to try such tests because the experience was eye opening and humbling.

Without wanting to sound rude, instead pointing out the pitfalls of my efforts why not shell out $20 and try it for yourself then come back and report your observations ? If enough people actually took an interest in this stuff the hobby as a whole might be a little better informed. You know full well that I am far from alone in my observations, a notable YT channel presenter has the same opinion based on doing the same sort of tests as well as other measurements.

Respectfully ....

 

[Hypops]

As I stated with the Clarinet test I actually did expect/hope to hear a difference because I was convinced that there was one based on normal listening.

No my tests are not hugely meaningful in a strict scientific sense and no they are not proof of anything except to me and yes there is something left on the table by way of doubt about possible very subtle nuances that I may have overlooked or simply could not hear. I presented them as my observations nothing more and I am aware of the limitations that you stated.

I did not suggest they were scientifically definitive in anyway way and I said several times that perhaps other folks might perhaps be able to pick up differences in the nuances at least.

However, doing a test within sensible limits of what is appropriate for a schmuck enthusiast at home is better than doing no tests at all and they proved to my satisfaction that if there is any difference that perhaps I might have missed or is beyond my hearing it is at a level that is immaterial to me.

For clarity, sighted comparisons always made me feel like there were differences, just like everybody else talks about, blind comparisons no.

The upshot was simply that I was encouraging others to try such tests because the experience was eye opening and humbling.

Without wanting to sound rude, instead pointing out the pitfalls of my efforts why not shell out $20 and try it for yourself then come back and report your observations ? If enough people actually took an interest in this stuff the hobby as a whole might be a little better informed. You know full well that I am far from alone in my observations, a notable YT channel presenter has the same opinion based on doing the same sort of tests as well as other measurements.

Respectfully ....

I can get behind this. Yeah, if AB tests are your particular kink in this hobby, by all means! We've all got our own things. But, no thanks, AB testing really isn't my thing. I know their limitations, so I know not to trust any conclusions I might draw from them. It's sort of like learning the Tooth Fairy isn't real. The magic isn't there for me.

Now, show me some FR graphs... mmm. Yeah, just like AB tests, FR graphs have fundamental flaws. But I love to look at, compare, and talk about those graphs. Like I said, we all have our own particular kinks in this hobby!

 

[Argha]

Juzear Defiant
Cheapest A-Ranked IEM in my ranking list)
Although it might not be 10/10 on everything, but for the price of $99 , damn mommy! "Juicy Thick Ass Bass" alone overshot the value spectrum by a huge margin.

 

[mikaik]

I like to look at graphs after I've listened, not before. But like @Hypops said, we are all looking for different things. I prefer to listen because my free time is limited anyway, so I'd rather spend it listening to music than looking at graphs.
I'm a male, so I can't multitask!

 

[Vamp898]

Sources do sound different, scientifically proven by measurements. You just have to do the right measurements and/or look at them correctly but they do sound different.

The issue is, that people only record test sounds and (usually) only test at 100% Volume. I test gear by playing music and/or white noise and at different volumes.

I play the music/noise, record the outpt and use a Software called "DeltaWave" to check the difference (Delta) between the original music and the output from the source gear. So i know exactly what changed in the soundwave.

I do this in two ways. I record directly the output from the source and in a second test, i record the output with different receivers (9 Ω IEM, 32 Ω Headphone, 470 Ω Headphone)

That way i found out why my Walkman sounded different than my DAC/Amp and different to my DAC/Dongle (all 3 sounded different) and why exactly it sounds different.

I went to different sound science guys and nobody believed me, they said its impossible, one of the devices is broken and so on. But i did know they sound different, i could distinguish them in a blind test so i bought measuring gear myself and tested with that.

And the measurement did show it. Conclusion, nobody tested any of these devices in real-world conditions. Most testers just connect and let the Software do the rest and wait for the result. Turns out if you use the devices how you use them in the real world, they perform different than in those synthetic test cases.

What annoyed me was, that my 50$ Dongle performed best and close to perfection. Really really good. The 2000$ DAC/Amp performed worst and the 3500$ Walkman was somewhere in the middle.

During further testing, i also found out why. Most devices i tested using this method perform worst at low volume. The louder you make them, the better (objective, not subjective, so closer to the original thing) they sound. At 100% Volume for example, the DAC/Amp performed very good, pretty much perfect, but then i tested at real world listening volumes.

It had a completely different frequency response on -90db, -70db (where IEM are already very loud), -50db (where most Headphones are already very loud), -30db and 0db

Its almost perfect at 0db but sounds broken at -90db with a very strong roll-off in bass and lower mids and a boost in the upper mids and that did show up in the measurements.

After further investigation, it turned out that the volume control of this DAC/Amp was the reason for this. And its like that by design.

On the Walkman it turned out that Sony did not implement an output snubber which caused the signal to overshoot and so change the frequency response.

In these tests by just playing music, the best devices (objectively) have been those with the lowest power output, as you can use them at relatively high volumes and digital volume control used with 24/32bit up sampling (The free upsampling that comes with your OS, no special device!) where they perform closest to ideal.

So in my personal experience the louder you can make something, the better it will sound. The more power something it has, the less loud you can use it (unless you like to listen at 160db while blowing out your eardrum and the IEM/Headphone at the same time) which will lead to distortion and or change the frequency response due to analog volume control or other factors.

So sources do sound different, it can be measured and, in my experience, (objectively!) the less power something has and the simpler it is built, the better it sounds. The higher the output power and the more fancy the components, the more will the soundwaves differ from what is originaly in the song.

If you subjectively prefer that, is on a completely different piece of paper. But if you want objective best performance, in my personal experience:

- Digital Volume Control
- Upsampling to 24/32bit (the free upsampling your OS does is sufficient) to prevent rounding errors of the digital volume control
- Output power that allows you to use the device at high volume with your receiver without destroying your hearing

But even then, you have no guarantee. Without measuring in real world circumstances, you simply don't know and maybe you don't care. Maybe you like the sound with higher distortion, why not? But sources can sound different.

That said, i also tested a lot of sources so far that all performed 99,99999% identical. So sources can sound different, quite significantly, but at the same time, most sources are pretty much identical.

Example:
DAWN PRO <--> Walkman --> Small difference
DAWN PRO <--> TA-ZH1ES --> Big difference
DAWN PRO <--> M11 Plus LTD --> Small difference
DAWN PRO <--> M11 Plus ESS --> Identical
DAWN PRO <--> MIAD01 --> Identical
DAWN PRO <--> Tanchjim Stargate II --> Identical
DAWN PRO <--> SMSL SP200 --> Small difference
DAWN PRO <--> SMSL SP400 --> Identical

 

[Okcerg]

I had a $50 LEC plus a $8 code from Linsoul, so ended up getting FIIO KA15 for $37.
It's a great little toy! Very clean and detailed, leaning a bit towards warmth.
It has filters that I actually hear a difference in, I usually don't. And a 10 band PEQ, so it can have any sound you want.
It doesn't get warm.
Does anyone know what FIIO's non-oversampling filter does? Is it mimicing the sound of an R2R, or is it just skipping oversampling, making it a distortion fest?
Whatever's the case, I really like the non-os filter. To my ears it sounds "less digital" than the other filters.

And the best part: Yes, the wheels spin!

 

[perrew]

I do know my HP Z Book headphone out sounds like garbage so I am certainly not saying that no device ever built sounds different, just that based on my humble assessment modern gear is not as different as general everyday listening experiences and folks comments in places like this suggest is the case.

That's too bad, the 3.5mm out on my MBP M2Max is excellent. I have the Sony WM1A and a UA2plus and the MBP and they all sound fine. I believe Apple spent some time one doing a decent amp/dac so would be very interesting to see if it is worse than an Ifi/Clarinet.

 

[nikbr]

Damn, I'm grooving hard with the H60. It's tastefully rich.

Funnily but understandably, inserting them into my earholes was much harder now, in the morning, than it was yesterday, in late PM hours. Optimally, I'd need Divinus Wild Boars in S, not M, for these.

 

[Okcerg]

Aside from that, I LOVE THE BP. Even more so that the Spark/Pirole Resin

Looks like Fengru BP is another cheap banger!
What would you say is the main difference between Fengru BP and Fengru Spark/Pirole Resin?

 

[Vamp898]

Oh i forgot to mention. As long you use a modern DAC/Amp, you don't have to worry about Crosstalk.

This here is my 50$ DAWN PRO



And this here is the 3.5mm of my cheap Sharp Aquos Sense 7 that i bought 2 years ago for 300 bucks (an average smartphone with 3.5mm jack)



Absolutely nothing to worry about. But at 3.5mm, you have have lower output impedance (better for sensitive low-ohm devices) and you have half the volume, which means you can crank up the volume --> higher quality.

Give more love to 3.5mm. Yes i know, the 4.4mm crosstalk is, theoretically, not existing, but you probably will not be able to hear crosstalk at -110db, so enjoy your 3.5mm worry free and the benefits of lower output impedance and better THD+N with it.

 

[BS5711]

I can get behind this. Yeah, if AB tests are your particular kink in this hobby, by all means! We've all got our own things. But, no thanks, AB testing really isn't my thing. I know their limitations, so I know not to trust any conclusions I might draw from them. It's sort of like learning the Tooth Fairy isn't real. The magic isn't there for me.

Now, show me some FR graphs... mmm. Yeah, just like AB tests, FR graphs have fundamental flaws. But I love to look at, compare, and talk about those graphs. Like I said, we all have our own particular kinks in this hobby!

I just find it fascinating that you and others feel the need to post to poo poo such things with no personal experience of them.

It isn’t a kink, that comes across as a touch condescending but that’s OK. It is simply a little experiment I undertook, it is hardly like it will be part of day to day listening. I used it to experiment for a few days now it has served its purpose and is gathering dust in a cupboard.

Yes there are flaws, I never pretended there wasn’t, but at the very least it is a stepping stone to improved understanding. If you are not interested that is fine.

However, if something is profoundly different blind comparing versus sighted it seems some folks might be intrigued to better understand why rather than shut the idea down because of preconceptions as you and a couple of others have done.

You seem interested in details such as output impedance, harmonic distortion etc etc but shun the notion of perhaps better understanding psychology despite that it is a critical part of audio perception.

Even if you don’t put any importance on the outcome of such comparisons it seems that an intelligent person interested in technical details might at the very least be somewhat intrigued to experience the massive shift in perception when the only difference is knowing what you are listening to.

If you don’t believe that biases play such a big role then logically you would be able to be removed from bias and hear the same differences sighted versus not. If you know that isn’t the case how do you rationalise the bias isn’t playing a big role in your current perceptions and perhaps something like my blind findings are closer to reality than sighted comparisons ?

Look, at the end of the day I don’t care what you or anyone else does but I don’t understand the need to criticise but at the same time apparently have no frame of reference to criticise from except normal sighted listening that is at least as, and realistically far more, flawed than a blind comparison in terms of perception errors and simply repeating the usual criticisms.

As always, honestly and respectfully despite that I am sure it doesn’t come across that way due to the communication medium and previous history.

---

@Vamp898 …. I have no doubt that some sources can sound different and certainly with some transducers. However, my point is that any differences are very likely not the night and day that gets talked about so often and not to the extent that every man and his dog can easily pick up the differences at a casual listen. You had to go to the extent of detailed measurements to support your own feelings about what you heard, if the differences were as utterly profound as most people describe I doubt you would have felt the need to go to those lengths.

Considering the lengths you have gone to I would like to reiterate this statement from you if I may:

“ That said, i also tested a lot of sources so far that all performed 99,99999% identical. So sources can sound different, quite significantly, but at the same time, most sources are pretty much identical. “

And that has been precisely my observation with the few sources I compared and discussed previously

 

[MD Rohit]

Sendy Audio Aiva 2 – First Impression​

So, recently I got my hands on the new Sendy Audio Aiva 2 headphone and honestly, I was pretty excited the moment I received the parcel. Couldn't wait much, so I jumped straight into unboxing and started listening right away. Also clicked some decent shots of it to go with this quick first impression post.

Build and Design

First thing first, the build quality on this one is just really solid. For the price, it's quite impressive. The design feels premium, and the materials they used makes it look and feel high-end. When I hold it, there's that kind of solid-in-hand feeling that’s satisfying. The comfort is actually pretty good too – I was expecting it to be a bit heavier but nope, not that heavy at all. I’d personally give it a 5/5 for comfort for now. Though to be honest, I didn’t get to test it for really long sessions yet, so let’s see how that goes in the full review later.

Sound Impressions (so far)​

Alright, now the sound part. The Aiva 2 sounds really detailed. The treble is airy and has a nice extension – doesn’t feel sharp or fatiguing to my ears. Bass has that solid impact and also quite fast – doesn’t feel muddy or bloated, which is good. But when it comes to vocals, I feel like it’s missing a little bit of warmth. Not saying it’s bad or anything, just feels a bit dry or slightly cold maybe.

Maybe it needs a bit more burn-in, I believe. Sometimes these planars open up more after some hours of use. Haven’t really had enough time yet to test with all kind of tracks or setups, so yeah, gonna give it more time before making up my mind.

That’s all for now – just wanted to share my early thoughts. Will definitely post more updates and full review once I spend more time with it.
Stay tuned!

 

[Nimweth]

Ziigaat Lush review:
https://www.head-fi.org/showcase/ziigaat-lush.28072/reviews

 

[Ozboyblu]

Looks like it’s been released. All your questions answered. (shamelessly copied from Facebook update)

What an odd and unexpected choice. It looks like it does more than just convert USB-to-coax, but I can't guess what those options actually do ("standard," "enhanced I," "enhanced II," and "digital to peanuts"?).

Yeah, I don't really see the point. Who needs a portable coax output?

Which I do as well. But lots of daps already have coax/spdif out, or have usb to coax cable adapters that do away with another dongle attachment.
Might as well just get the Dethonray dongle so you also have the optical option. Probably cheaper too.

How much will it sell for? This COULD be a cheaper alternative DAC for the Clarinet. I already have the Dethonray DAC they sell for the Clarinet, but it's $50.00. So if it's cheaper than that, might be a good alternative DAC to unlock the Clarinet's coaxial goodness?

$29USD
https://store.hiby.com/products/hiby-uc10-usb-to-coaxial-s-pdif-digital-audio-converter