Do you believe in the famous "burn-in"?

Off-Topic
42

Why only 2 options? I want to say “always believe” + “changes technicalities” + “already at 20h”.

I think it can be proven with measurements, I think Tyll has proven it with measurements even though his comments on his own graphs seem less convinced, and I think no one who claims to have disproven it with measurements can be believed if they haven’t tried it with cheap headphones that undergo the least testing at the factory, like Superlux older models (prime target would be the 668B). You test expensive headphones made out of the best materials, you’re not really giving the hypothesis a chance, IMO. I’d like to see 50-ish hours of frequency sweeps, and multiple measurements like FR, IR, SR and CSD to take an experiment seriously.

43

Interestingly, I’ve just bought 007mkii Stax and as I was paying, the chap said “don’t worry about running them in as that’s all done at the factory”
So Stax think it’s a thing apparently.

44

As does any other industry including but not limited to automotive, shipyards, electronic components, home appliances, construction equipment, etc.

45

I’ll take a swing at this! :grin:

I got into IEMs about a year ago (dang HBB :rage: ).

Of all the IEMs I’ve owned, or spent extensive time with, my Galaxy Bud + come the closest to the Harman Curve. This includes the Mangird Teas, Fir Audio VxV, Sennheiser ie900, Sony IER z1r and, my current daily driver, the UM Mest Mkii.

Any measurement system MUST have a standard. We don’t just say that something is “long” we say it’s 10 inches long and is LONGER than something 9 inches long. The same applies to “measuring” sound. Whether it’s THD, SINAD, Sensitivity, Frequency Response, Impedance, etc., if we insist on measurement we MUST do so against an objective standard. The problem, as it currently stands, is establishing an objective standard, like inches and pounds. And that’s all the Harman Curve (or any FR curve) is: inches and pounds. We aren’t anywhere close to an objective standard for sound. And this doesn’t even account for the uncontrolled variations in technique that occur among the many reviewers who do measure for FR.

Frankly, when it comes to music, my Galaxy Bud + sound like @ss compared to any other IEM I’ve owned. I only use them for listening to TV at night.

Anyone who accepts the Harman Curve (or any other curve) would have to account for why the Galaxy Bud + doesn’t sound as good as others that don’t adhere to the Harman Curve. That they don’t sound as good, I hope, is not controversial to this discussion.

(To be fair, the Galaxy Bud + don’t sound awful - just not as good as the others. As an old fart, I can remember the bad old days of the '70s and '80s and I would have walked over corpses to get something as good as the Galaxy Bud +)

46

Of all the things we could possibly measure in headphones, speakers, IEMs, etc, I think frequency response is probably the least interesting to me. Sure, we could hunt for a pair of cans with a truly “flat” FRC, or something that closely matches any “target curve” if we were so inclined. That’s only going to tell us how they respond in “from the factory” configuration, on “reference” source gear, measured on the rig of whichever reviewer did the measurements. Change the measuring device, dac, amp, pads, etc, and you’ll change the curve. That’s before even mentioning EQ.

For millennia, science was a semi-codified process of using our senses to observe the world around us. What’s worse, is the truth that the placebo effect can cause real physical manifestations. To a great extent, we hear what we expect to hear in much the same way we see what we expect to see (there’s a lot of new evidence coming out suggesting a huge percentage of our sensory perceptions are very heavily influenced or even completely controlled by the sum of our previous experiences). Further, with different ear shapes, ear canal shapes, etc. we don’t all hear the same sound the same way.

It seems to me that the best measurement tool we have is probably still our ears. Yes, leaning on them poses some challenges. Eliminating, or accounting for all of the various biases and psychological influences is difficult to do. However, things like FRC, minute variations in timing, sound waves partially interfering with one another, etc. we probably won’t come up with more “objective” tools that can tell us which factors interact with one another in what way to make the resolving ability of one pair of headphones sound superior, but in a very small, intimate stage, while a second pair (of equivalent open / closed level design) could sound smoother, less resolving, but in a grand, wide, deep stage.

TL;DR - sound reproduction is a problem where a large number of variables interface and influence one another. It’s still too complex for the basic measurements we’re using to tell us much that is very useful. Our ears are still the best tool for the job.

1 Like
47

It’s also interesting (and should not be left unchallenged or taken as a given) that we humans have this “urge” to convert sound into a Cartesian graph which displays amplitude and frequency in a VISUAL format.

The fact is that humans, like all primates, are primarily visual creatures. A significant portion of our neural apparatus is geared towards vision (unlike dogs/wolves, which are geared towards smell). We humans traded increased visual distance, due to the height of our eyes above the ground, for a reduced sense of smell - since most scent molecules are heavy and fall to the ground.

What does a frequency response mean to someone born blind? Teaching mathematics and geometry to those born blind poses serious challenges (Google it for more info. An interesting topic). Can a blind person not appreciate (or judge) the quality of sound of a transducer because they cannot see the FR? I actually heard on YouTube a reviewer (I won’t mention his name) say that he had a hard time judging a particular IEM because he didn’t have a graph…

Would anyone take seriously food recommendations reduced to X vs Y graphs measuring salinity, acidity, sweetness, sourness, umami, etc.? Certainly not. The reason we pay rapt attention to visual graphs is because we are visual animals.

Edit: I prefer to use my vision for it’s evolutionary adapted purpose!

1 Like
48

First off, this is a completely different topic to the one I was responding to (above). As noted wavetheory wasn’t clear in his criticism, but if I had to guess I would say he was alluding to microphones being an insufficient device/technology for measuring sound.

I don’t think you are saying anything controversial, but I also feel like you are mixing the terms somewhat. Feet and inches are the units of measurement. Saying you should be 6 foot tall at least to join the army is a standard. We could describe you relative to that standard, but we would still measure you in feet and inches (or centimetres), and we would use a measuring tape (the measurement device).

A response target is a standard, expressed in 2 dimensions. Unlike the example above which is expressed in a single dimension (height).

You don’t enjoy an earphone which allegedly conforms well to a response target. Ok, I would say this is a common occurrence, which could be explained by many factors (eg unit variance, measurement / comparison problems, preference). I personally, don’t prefer Harman tuned headphones.

With regards to your other points, THD, SINAD, sensitivity, impedance again are not units of measurement, but dimensions upon which a piece of audio equipment could be measured. Some of these are quite predictive of performance. Sensitivity measurements, for example, can tell you if your amp can make them sufficiently loud.

This thread is about the concept of burn-in. I would expect performance changes related to burn-in is a measurable phenomena. In fact from what evidence I have seen on these measurements, we can often measure an effect, but find that the effect is inaudible. This would lead me to believe that our measurement devices are quite sufficient (for the subject at hand).

49

I would expect that food scientists formulating mass produced snacks, soft drinks etc would use this kind of information. And then they could establish through subjective measurements how those technical measurements relate to preference. With that that they could attempt to make some relationships, improve the product, test again… repeat.

50

The good audio products gets tested, so they notice possible flaws or production errors. Maybe even figure out what was wrong and in what component, if there are more similar incidents. They know there is flaw in production or material. They will fix or change that.
Products should perform at optimum levels before consumer buys them, so no changes happen in sound.
It something would happen, it’s not in the product you bought.

51

Well, more on this. I was on the Focal site and saw this while in the Elegia product page:

"RUNNING-IN PERIOD

  • These headphones require a running-in period in order to operate at their best.
  • They will run themselves in naturally after several hours of listening time.
  • If you want to speed up the process, we recommend playing very bassy music through them for at least 24 hours at a relatively high volume.
  • This will stabilise the speaker drivers and allow you to get the very best out of your headphones."

So there you go. As previously mentioned, I did notice a significant improvement after several hours. I definitely think that in some cases, it is relevant, and in others, not so much.

1 Like
52

In my experience it depends on the product. Some brands you don’t need to worry about burn in like Schiit which does that as a means of quality control inspection before shipping their products. Sometimes burn in varies like with Dan Clark products; sometimes it is barely noticeable and other times it is a huge leap. I have experienced a night and day difference with burn-in like with my Meze Liric and my friend’s Meze Elite after just 10 hours. 100 hours of burn-in though is preposterous and just nonsense.

To me it is a case by case basis. Besides if after 10 hours of burn-in I am not impressed then I’ll take advantage of the retailer’s return policy. All I can say is I wish the poll had a case by case option.

53

Hello,
As always, it’s controversial.
Among my experiences that I can contribute are.
When I bought new headphones, they always worked a little worse in the beginning than afterwards.
It often took 20-50 hours before it got better.
The only headphones that sounded great right away were the Lcd 2 C.
I took them out of the box, plugged them in and they sounded magical.
All the other headphones I bought, Fostex, Denon, Aeon R/T, always took me a while to get used to.
The Denon D2000, which is now 10 years old, still runs better than my former Tr X00 - every time I put it on.
With consumer electronics I see it a little differently.
Most of the electronic devices were already satisfactory after 20 hours.
Recently I bought the Ifi Hip Dac 2.
Ifi also said to give it 100 hours.
It was so that the sound was a bit hard at the beginning.
After more than 50 hours I can say that it has improved for me, it is no longer hard, and the sound is also softer and somewhat deeper than before.
Subjectively, it wasn’t much, but it was audible, so something did happen.
With capacitors like Vcaps and the like, it’s a completely different story.
When I converted my Feliks Euforia to Vcaps, the 400h were really necessary.
During that time, I really went through a rollercoaster ride.
Sometimes it was ear fatigue, sometimes too sharp, sometimes it scratched occasionally or lacked flesh.
After 400 hours I can say that it is excellent and worth the rollercoaster ride.
And would do it again.
In the end, it has to be said that everyone should find their own measure of how they feel about burn-in.
Based purely on observation, 20-50 hours is usually enough for consumer electronics, and it’s not wrong to stick to this for headphones either.
It is not wrong and contributes to the cause.
It might be interesting at first, but now I hate it.
There is not really a recipe for doing a scientific burn in.
Everyone does it as they see fit.
I have developed my own method of doing something that I think is right for the product.
There are now also many manufacturers who incorporate this into their production processes.
Queststyle, for example, runs their amplifiers in for 100 hours and avoids amplifier failures.
Zmf also tests their headphones before shipping, even Denon with their D9200 do something similar.

2 Likes
54

Okay, I have one right now:

TW: talk about burn-in (reader discretion is advised)

So I’m lurking in the Head-Fi watercooler because dudes talking about $4000 IEMs like they’re as accessible as $100 sets is weirdly fascinating. Someone was talking about the Campfire Trifecta and was like “it’s a 3DD, you need 150 hours of burn-in to get the sound where it’s supposed to be, and there are people selling it after their first listen. That’s weird.”

What I think is weird is a company selling a $3400 IEM that isn’t ready in it’s “best, as intended” state upon box opening. Nobody has said if CFA encourages or suggests this burn-in period (companies like Dunu do), and if they haven’t done so and it’s just what enthusiasts suggest, I’d drop this point dead.

But if CFA recommends 150 hours of burn-in (that’s a week, for the people not doing math), why are we accepting a TOTL $3400 IEM can be shipped to customers and it’s not ready to be listened to “as intended” until you run it for a week. Someone argued that it must be too much to manage logistically because they had a run of 333 units, and to burn those units in for a week, they’d have to charge more to figure out how to do it.

Yeah, and? It’s a $3400 IEM. If you charged $3800 with burn-in assumed and baked into the price, was anybody who bought it new for $3400 going to honestly not buy it at $3800 now?

Much like with Dunu selling 300 SA6 Ultras but only having 100 production models completed when they launched, this kind of half-assery bothers the hell out of me.

Or am I just crazy and unreasonable?

6 Likes
55

But that’s like folks talking about £500k + cars they probably have 10+ of them…as for burn in I like the ride from the get go tbh ie 90% mental and the rest… let the drivers flex :muscle::muscle: :stuck_out_tongue_winking_eye:

2 Likes
56

I can’t take 99% of the discussions in that thread seriously, and that’s a big part of why I post here and only ever lurk Head-fi. Don’t get me wrong, people can do whatever they want with their money and involvement in the hobby but I can’t go more than 3 posts at a time without rolling my eyes at the assertions made. And that’s coming from someone who laid down the money for the Storm (which is by FAR the largest single purchase I have ever made for a hobby).

@Ohmboy I am absolutely in the same boat regarding “burn-in” of anything; I expect that it shipped to me as it was supposed to be, and taking the time to get to know gear is like half of the fun IMO! I can’t fathom running something on noise instead of just enjoying your music for whatever period of time you’re “supposed to” burn them in, what a waste of time that could be spent enjoying them or quickly moving on if they don’t impress!

3 Likes
57

“Burn in”…I think sometimes even with IEM’s this can make sense…heres what our @WaveTheory says regarding certain sets which I can get behind :+1:…

1 Like
58

I kinda would like to have a Detailed, Rail Road Track type of explanation how a planar-magnetic driver gets “burn in”.
I do actually get the pad wear… ear driver distance changes a bit and this is something i tested with factory new HD6XX vs. used HD6XX. 2 months production difference, same factory.
You get the same sound characters by little pressing the new headphones inwards = ear driver distance gets the same. But they still sounded very very very similar. Would say, the same.

Now back again to the very very little planar-magnet… what moves or changes while using it so that the sound changes or gets better?
I would bet money… nothing.
I still think the term is just f–ked up. Burn in.

1 Like
59

And that’s cool :+1:

1 Like
60

Uh oh! My name has been invoked on a controversial issue.

For clarity, I use the term “burn-in” for what goes on in the brain, and I use “break-in” for what physically happens with a driver. Both happen. The one with the larger impact is less clear and varies with each piece of gear, as best I can tell. It’s also really hard to pin down brain effects because while all human brains physiologically have the same mechanistic operation (we are all of the same species, after all), our experiences cause nearly limitless differences in how those physiological mechanics manifest.

Monty-python-god GIFs - Get the best GIF on GIPHY

Right…

Some amount of break-in is unavoidable. Moving parts are going to be stiff, particularly when they’re new, and will require some flexing before they settle out into what their “normal” or “target” behavior is going to be. The field of material science has known this for decades, maybe even centuries. I don’t know why audiophiles like to forget this.

Does it make an audible impact? Sometimes! The case of the DCA Expanse is a good one. It’s actually quite measurable even in the frequency domain. My friend Lachlan at Passion for Sound (check out his channel, it’s very good) measured an Expanse brand new out of the box, and then after break-in:

image
image1920×1080 104 KB

That image shows up around the 27 minute mark of his interview with Dan Clark.

At first blush, that may appear to be a tiny difference. If you isolate any 1 frequency, more often than not, it is a tiny difference. But our ears do not isolate 1 tiny frequency when we listen to music. You have to look at the relationship between fundamental frequencies and their integer-multiple overtones called harmonics. Our ears are going to pick up on the patterns that all of those create together. In the graph, look at the difference at around 250 Hz; it’s about 1dB. Alone, that might be barely audible for a small percentage of even highly experienced audio professionals. But, now look at the first 2 integer multiples: 500Hz and 750Hz. There’s about a half dB difference at both of those places, all with the broken in being higher than the brand new. Any string pluck or vocal note that has a fundamental frequency close to 250Hz is going to sound about 2dB more forward on the broken-in unit than when it was new. A 2dB difference is now getting into a range most humans can detect.

So, yeah, break-in happens. Here’s an example.

What this little diatribe does not answer is “should we tolerate this?” That’s for you to decide with your wallet.

Cheers :beers:

5 Likes
61

I think my rant got misunderstood: I wasn’t going off on “does burn/break-in exist?” That wasn’t what was relevant to me.

My issue was coming from the idea that a company believes their IEM drivers work better after break-in. I remember one Dunu IEM recently (I don’t remember if it was Talos, SA6 Ultra or something else) that the Dunu rep said Dunu encourages and suggests a 50-100 hour break-in for that IEM.

The issue I’m raising is as such: if a company feels break-in is important for the “ideal or as-intended” operation of their IEM, then why aren’t they breaking the drivers in, in a manner of their own control and then shipping to customers so the IEM arrives in a “ready to use properly at open-box” status?

3 Likes