Take a familiar piece of full range mono music - say a guitar or keyboard track and loop a few measures of it. Insert a 1-band eq in the track. Set the eq to a 3 dB boost and sweep the frequency through the spectrum. Do the same with a 3 dB cut. You will hear the sweep easily. Now do the same with 2 dB and 1 dB. Those will be pretty noticeable as well. Bet you get down to a pretty small fraction of a dB before you can't hear it.

Well is +/- 3db good enough for studio monitors? I heard that it's a standard to have +/- 3db so it must mean it's good enough for monitors.

Most current standards for nominal human hearing can only discern a difference in volume change of 3 db. That has been the typical standard for thresholds of perceptions for the 90 percentile human.

The vast majority of modern day humans can't hear anything above 16khz anyway...below 40Hz the sound is more pressure which using proper amplification is actually really felt rather than "heard".

With a reference monitor speaker system what you want is the spectrum to be as "flat" as possible throughout the target bandwidth so that you can accurately judge the various levels of the various frequencies and levels your music contains.
Your musical content varies both in frequency, time and level as it's being played back to your ears.
Sometimes it playing a 100Hz sound sometimes it's playing a 8khz sound.
if you have a speaker that will faithfully represent as many of those frequencies and not vary much and you can rely on that...you can use that as a known reference standard in order to "know" the sound you hear is accurate within say 3db of what your hearing.
You know there is no exaggeration of any particular frequency.
It also helps you decide if your meters are lying to you or how accurate they are.
Sometimes they can be more accurate than your ears....depending on the design.
Of course digital metering has little or no relative accuracy because they have latency in response speed to incoming signals.

Today's low cost "powered" or internally amplified "monitor" speakers are "relatively" accurate based on the costs of design and materials used.
It's relatively easy and cheap to maintain consistent amplification levels within the mid-band regions to drive a specifically designed speaker driver/cone assembly but requires a different amount of consistent power to maintain that within the low-band region which is why this typically begins to roll off below a certain point...neither the speaker coils/cones or the amplification circuitry is adequate enough to consistently maintain equal levels of power as you go lower and lower in the bandwidth.
The monitors your talking about here have trade-offs in design because of cost considerations.
They are "relatively flat" for a consumer grade design.
A trained experienced mastering engineer would probably hear a difference on those speakers.
Of course that depends on the accuracy of your particular hearing and ear training too.
An untrained ear probably wouldn't be able to "hear" the difference and all of this is subjective anyways.
In the case of $150-$500 consumer grade speakers...it's not that critical.

Once the speaker system begins to roll off you no longer have a consistent reference that you can trust.
The lower in frequency you go the lower the power level becomes and the speaker no longer can produce a consistent level.
So now everything you perceive at that low frequency is inaccurate relative to everything else.
Your reference standard no longer works for you and....you over compensate with your mixer and boost the low end so it sounds louder and more balanced.
Now go put that music on another speaker system....hmmmm it's all muddy?...or hey there's no low end? It sounded good to me on my "monitors" in my bedroom!....(sorry....I couldn't help myself).

Speaker systems that are fully designed to be "flat" approaching say +/- 1db of variation from 20-20k will typically be very expensive passive speaker assemblies using very expensive highly accurate and powerful amplifier designs to maintain accurate power levels across that bandwidth.

For most hobbyist who use the 4", 5" or even 8" powered "monitors" costing $150-$500 in acoustically untreated listening rooms +/- 3db from 200hz-16khz is pretty amazing....actually technically speaking an amazing engineering bang for your buck...but at that point it really won't make much difference to you other than it won't be perfect by any stretch of the imagination....but that's all relative anyway!
So.....
Does it sound good to you?
How do you know it sounds good?
What does that even mean exactly?
Is that better than this? Yea maybe, I think so, I guess?
Does it sound good to you...Yea I think so
Did you hear that?....no did you?....
Can you hear me now?

What can you base any of that on....no two humans on this planet have the same hearing or ear construction.
Some people have complete loss of hearing at specific frequencies and don't even know it...some people are tone deaf in one or both ears and can't differentiate between subtle frequency changes anyway
A large percentage of people today in modern society can't really hear that well to begin with because of so much environmental sound pollution and exposure periods.
It's very probable you might be in that category.
The odds are if you listen to and are exposed to loud levels of sound over long periods of time there's a really good chance your hearing is already damaged to some degree or another.
Unless you've listened to sound on devices that have been designed and engineered faithfully to reproduce sound accurately in a professional environment and you've done that day in and day out you have nothing to base your hearing on.
You also have nothing to base your hearing against....it's totally subjective which is why people typically answer this with...
"Go listen to them yourself and see which ones you "think" sound better to you"....hmmm.
If they don't sound good to you when you do that... then what?....hmmmm...maybe that's why it's so hard to decide?...??
With the exception of maybe underwater speakers which you can tell (always a trip listening underwater...no air to deal with)!
I think people should first:
have their hearing tested. It should be a prerequisite.
I mean why bother with any of this or worry about it if your hearing is crappy to begin with?
Then listen to some music you absolutely know personally inside and out.
Your absolute favorite music...stuff that absolutely moves you inside on an emotional level...makes you cry whatever....at a modest level on some really "good" set of closed headphones.
Listen to the detail. Then go listen to speakers or "monitors" you can afford in a half decent room....
Just enjoy your music for what it is....to you
OK I'm definitely ranting now...I'll stop...too much tea!
Rock on!

Can you hear difference on few db?

Quote:
I know that my current speakers is from 65hz to 23kHz +/- 3db but if my speakers were +/- 2db or even +/- 1db would you hear a difference? I kinda wonder if anything 3db or lower is basicly very flat response for speakers. I just don't know if you can hear the difference. I think only a machine can detect the difference right?

...........................................................................

No, you believe the manufacturer. What you ought to know: they are cheating you big time.
Those cans are not giving you 65 Hz at -3 dB. They ripple over the whole range more then 3 dB up and down.

And yes, I hear that. With some material even below .05 dB.
What you don't hear correctly you cannot tweak correctly.
What you also hear is if the woofers have 4", 5" or 8", 10", 12"...
Don't think of your bass as properly existing. If you are just listening to music, ok.
But if you plan on doing serious mixing on it...please, consider some better monitors.