[marcussmj]

I'm really not sure that his methods are all that accurate. Unless the equipment can take in the sound in the same way as a human ear, with an input that represents a human ear, I don't think he would be measuring what we all hear. He makes his assumptions on measurements but he is a hobbyists, just like the rest of us. My understanding is that he uses a microphone setup with his phone.

He could certainly be right, but then he would be going against what many of us have heard. In which case then Singapore got a bad batch of earphones.

I'm inclined to question his conclusions, and not put a lot of stock on his graphs. His ranking system is not very indicative of the SQ of some of the earphones he has tested, IMHO, and the placement of earphones on that ranking system seem to me to be highly subjective. If he is that loose with his methodology as his chart's ranking system, then I really do not trust his conclusions.

I do agree that brand loyalty can lead to nowhere, but that is not the case here.

I should also add that this is not a case of shooting the messenger, I'm sure he is doing his best with his resources. I also think it is a Herculean task, but why try to bring pseudo-scientific testing with amateur equipment to a hobby based on listening and individual preferences? One is setting themselves up for a ton of disagreement.

No the list is subjective as if it was not implied already.

 

[Carlsan]

On the other hand, things being what they are, there should be some cheap Solaris's on the for sale section for those of you that have waited on buying...

 

[Penguino]

Yeah that can't be it, because if there were variances in chamber volume to the point where it affected sound, you could never match 2 sets. This is something else.
If you rule out faulty measurements, burn in, tip variances etc etc.
To my eyes the measurements look like different tips, or even different testing rigs. If that is not the case, only campfire can explain what might be going on.

If this was a true QC issue, both left and right channels would have an imbalance.
Here we have huge variances in FR response, but both sides still match.
Really really weird.

Well it is possible I guess that they could do batches of BA assemblies and measure those to match them. But I'm not sure how effective that would be in practice. I'd like to clarify that I'm not claiming this is the case I am merely intrigued and the idea came across my mind.

 

[marcussmj]

To my eyes the measurements look like different tips, or even different testing rigs.

Did you not read the reddit post on his testing methodology?

 

[Penguino]

Yeah that can't be it, because if there were variances in chamber volume to the point where it affected sound, you could never match 2 sets. This is something else.
If you rule out faulty measurements, burn in, tip variances etc etc.
To my eyes the measurements look like different tips, or even different testing rigs. If that is not the case, only campfire can explain what might be going on.

If this was a true QC issue, both left and right channels would have an imbalance.
Here we have huge variances in FR response, but both sides still match.
Really really weird.

Also Crin does the upmost to ensure that his measurements are consistent. Same rig, same tips, same insertion depth, ect. The issue lies within the units, not the person measuring it in this case.

 

[Kitechaser]

Did you not read the reddit post on his testing methodology?

Yes I did. The question I am asking is if the huge variances in FR response is because of poor quality control, how is that both left and right channels still match on these sets?
Wouldn't the first sign of deviance be an obvious imbalance?
How to you eff up an FR response by 15db unintentionally, and still have a pair that matches?
Weird...

 

[jeffreyw311]

Good channel matching doesn’t mean there isn’t variances in manufacturing. It’s whats done with boutique speaker drivers, you merely match two drivers that measure similarly and sell them as a pair.

The channel matching from Crinacle’s samples suggest that left and right IEMs that measure similarly are being paired together since channel matching there is close but not perfect. That’s not anything new, we’ve seen this with Audeze (iSINE10s especially, pairs always have the same overall sound signature but not perfect channel matching or consistency), AKG, Beyerdynamic and so forth.

 

[Kitechaser]

Was that the experience with the Andromeda? Huge variances?
Campfire has been doing this for years.
Like I said earlier it's either Campfire is knowingly pairing drivers with variances up to 15db, or something is up with the measurements.
Only @KB can settle this.
Or we all wait till more review measurements that confirm this.

 

[jeffreyw311]

Was that the experience with the Andromeda? Huge variances?
Campfire has been doing this for years.
Like I said earlier it's either Campfire is knowingly pairing drivers with variances up to 15db, or something is up with the measurements.
Only @KB can settle this.
Or we all wait till more review measurements pop up.

Andromeda implementation is extremely simple, consistent manufacturing would have been pretty easy with those. I’ve done some DIY things with BA drivers that have ended up consistent enough.

Hybrids have had these problems because the internal volume is pretty important and evidently something that is sometimes difficult to get right. Crinacle’s custom 64Audio N8s had this problem where his sample had extremely bloated bass, which was quickly resolved by the manufacturer. It’s one of the main reasons I have a makeshift measurement rig because it’s a fantastic sanity check when I feel something isn’t right.

Doing it for years doesn’t really mean much. Beyerdynamic have made headphones since before the Cold War yet they’ve had a long history of poor consistency and channel matching with certain product lines.

 

[Kitechaser]

Andromeda implementation is extremely simple, consistent manufacturing would have been pretty easy with those. I’ve done some DIY things with BA dri

Hybrids have had these problems because the internal volume is pretty important and evidently something that is sometimes difficult to get right. Crinacle’s custom 64Audio N8s had this problem where his sample had extremely bloated bass, which was quickly resolved by the manufacturer. It’s one of the main reasons I have a makeshift measurement rig because it’s a fantastic sanity check when I feel something isn’t right.

Doing it for years doesn’t really mean much. Beyerdynamic have made headphones since before the Cold War yet they’ve had a long history of poor consistency and channel matching with certain product lines.

If that is the case they how do you match two sets with different internal volumes that are so sensitive to even minuscule changes?
We are back to the original problem.
Going in circles.
On a custom shell, what you are saying makes sense. This is a universal shell, not much different than the andromeda.

 

[jeffreyw311]

If that is the case they how do you match two sets with different internal volumes that are so sensitive to even minuscule changes?
We are back to the original problem.
Going in circles.
On a custom driver, what you are saying makes sense. This is a universal shell, not much different than the andromeda.

You measure left and right channels and pair ones that measure similarly. Look at Crinacle’s measurements again past 500hz. They have similar sound signatures but not perfect channel matching. It isn’t his measurement methods that are bad because the same samples with same tips measure identically.

It is exactly like boutique tube resellers who measure and pair tubes into pairs that have similar characteristics. The two tubes aren’t the same but they both have very similar behaviours that few people would notice a difference between the left and right.

 

[Kitechaser]

You measure left and right channels and pair ones that measure similarly. Look at Crinacle’s measurements again past 500hz. They have similar sound signatures but not perfect channel matching. It isn’t his measurement methods that are bad because the same samples with same tips measure identically.

It is exactly like boutique tube resellers who measure and pair tubes into pairs that have similar characteristics. The two tubes aren’t the same but they both have very similar behaviours that few people would notice a difference between the left and right.

So where are you proposing the lapse in quality control is?
Variance in chamber volume, or selection of drivers from the get go?

 

[echineko]

So where are you proposing the lapse in quality control is?
Variance in chamber volume, or selection of drivers from the get go?

I would say the lapse in QC is with CFA, because apparently they released these as is for sale.

FWIW, when I was in Singapore last week I had a chance to spend time with the so-called "best sample" of Solaris available domestically. Tested with my WM1Z for 30 minutes or so, comparing with some other IEMs I had with me too.

I could understand why people like it, though I would take my SE5U and the VE8 demo at the store I auditioned them at, just not really my thing. But I can see why people appreciate them. The question on variance however, makes me think this could be another situation as a certain planar headphone manufacturer that will remain nameless, I might have liked the sound, but couldn't really recommend it as I wouldn't know what people will end up with.

 

[Kitechaser]

I would say the lapse in QC is with CFA, because apparently they released these as is for sale.

Based on what? Three measurements in Singapore?
The left right channel on the pair he dislikes match up very very closely.
If the volume of the chamber was of issue, that would also throw off the DD because of the air expansion.
I am still trying to figure out, how you have variances of upwards of 15db and still ship out almost identical matching sets.

This is getting fishy.. just saying.

 

[echineko]

Based on what? Three measurements in Singapore?
The left right channel on the pair he dislikes match up very very closely.
If the volume of the chamber was of issue, that would also throw off the DD because of the air expansion.
I am still trying to figure out, how you have variances of upwards of 15db and still ship out almost identical matching sets.

This is getting fishy.. just saying.

What are you saying, exactly? I'm not sure I follow. And did you even read the part about repeated measurements being consistent?