LOL, smack all you like Joe. But ABOVE the belt, please! The Fletcher Munson curves are an indicator of human hearing AT LOW LEVELS, and have to do with how loud a sound must be to be perceived (in a particular, broad frequency range), not about "high q" peaks at specific frequencies! Different can 'o worms! Consider yourself SMACKED!

Has anybody here noticed the SUPERB equalization capabilities (separate "bass", "loudness", and "bass boost" sections, as well as treble) in the Archos Jukebox Recorder 6/10/20? Plus the thing has a shockingly good headphone amp for a portable device! I love mine! This one will drive people who say mp3 can't sound good nuts!

I tried your wavs...

Later I'll do a frequency analysis to see exactly how far the 3kHz frequency was boosted: 3dB? 6dB? 12dB?

For demonstration purposes I bet that peak is bigger and sharper than most anything that real headphones will have

On my un-EQed EX70s I can't hear any tones standing out (except a general 'bass boost'), and those are hardly what you'd call balanced cans...

And yes! I'm calling your method primitive Even if I do hear 'tones' in the white / pink noise, I'd never know for sure what frequency it is. With my tone sweeps I can tell just from the time in the wav when I hear a peak / dip!

Back in those bad old days with my Aiwa HP-X225 I had a log like this:

Frequency(Hz) Response(dB)
0 – 50 +10
58 +9
80 +3
114 0
161 -2
227 -3
454 +2
641 +5
826 +8
900-1700 +10
1950 +8
2450 +9
2900 +3
3675 -2
4975 +2
5700-6300 (abrupt change)
6300-9000 ~+20
9000+ ~+10 (but uneven)

(extract from my old epinion review)
Now I know that because the Fletcher-Munson curve that I used for normalization does not match my own ear's hearing characteristics well, the relative responses are not accurate for large intervals. (e.g. 454 vs 4975) But for small intervals this does not matter--and look at how close some of those entrys are!

2450 +9
2900 +3
3675 -2
4975 +2
5700-6300 (abrupt change)
6300-9000 ~+20

I dare you to come up with something like this using white noise!
(2 HUGE peaks and a huge trough, bundled so close together that...)

Quote:



Anyone who says that, doesn't understand what the Fletcher Munson curve is. There an FM curve for 0dB (very quiet), for 120dB (earsplittingly loud), and everything in between...


(damned worst graph I've ever seen... should have uploaded the one I stored in my hard disk--but not using my own computer right now)

SMACKS below the belt---NYAAHH!

You've already admitted that white / pink noise can't be used to spot broad swells and dips in frequency response, only sharp peaks / troughs. My last post just showed how much better you can do with a tone sweep to deal with sharp peaks / troughs.

I'm trying to construct an FM curve tailored for my own ears at 70dB. It'll only be for one loudness level, but once I've constructed it--one 1min20s sweep and I can literally *hear* the frequency response curve of a new phone going through my ears! EQ will be a cinch!

...

My NEX II mp3 player has a 5-band EQ... beat that!
But mine is also a broken unit that hisses out of the headphone jack as loudly as a snowstorm on 8 ohm phones...

Your own curve shows exactly what I said...the Fletcher Munson curves are ALL gradual, with gently sloping response from octave to octave, and NO high q peaks (or dips). LOOK AT THE DAMN THINGS! Don't just say I'm wrong, prove it, or shut the hell up! Your own curves prove me RIGHT!

As for your "sweeps", you completely missed my point, Joe. My interstation hiss test is CHEAP (free), reliable, repeatable, and requires NO special equipment/software/etc. And it isn't just MY test! Respected audio journalists (JOURNALISTS, not the subjectivists who write for "flat earth society" subjectivist rags) have used this method for as long as there has been fm!

And speaking of PRIMITIVE, the largely discredited Fletcher-Munson curves are more than a half-century old! We understand FAR more about human hearing now than in the damn 50s!

The peak, by the way, is NINE db, less than a quarter of an octave wide, centered on 3khz. This is very much like the peaks found in MANY headphones/speakers. If you can't hear it, Joe, well...that speaks volumes about your hearing acuity, Brother! I can easily hear it on every pair of 'phones at my disposal...even on my cheapest computer speakers! As for how you're to tell this peak from others present in your phones/speakers, to do that I provided a white noise and pink noise sample both with and without the peak! The difference you hear (or a NORMAL person hears) between the two files is the f@@@ing peak, Joe! That is the ONLY difference. Because there is a sample (of both pink and white) with and without the peak, it is easy to spot the peak! Just as, when comparing various headphones while listening to fm interstation hiss, it's easy to spot peaks which are present in one pair of 'phones and not the other! If you really can't hear this, Joe, I would suggest keeping it to yourself...as it calls your audio judgement into question, because the rest of the world CAN! (KNOCKOUT!)

For those as deaf as Joe, I have created an additional file which should make it easy for ANYONE to hear the peak. Click on the bottom-most link, and you will hear 30 seconds of white noise with the 3khz peak cycling on and off every five seconds. Can't hear that? Then you ARE deaf! (sad, but true!) Click on this link and enjoy! Noise samples with and without peak at The Production Room Dot Net

Joe I DID NOT admit that white and pink noise can't be used to spot "broad swells and dips". Again you've misquoted me, and responded to the misquote rather than to what I said! SMACK!!!! I said that ALL 'phones deviate from flat response in some way, and that the narrow, sharp (high q) peaks are far more offensive than SMOOTH trends over octaves. I DID NOT say that the smooth trends over octaves ("broad swells and dips" as Joe puts it) aren't audible with this test! I said that they're less important ANYWAY! It's far easier for your ears to hear through a smooth curve resulting in a midrange "lift" over several octaves, then a smooth return to the reference or "nominal flat" level than to a curve that's flat except for a huge, narrow peak over a fraction of an octave in the midrange (especially in the midrange, because it's the range of the human voice...and no coincidence that it's the range where our ears are most sensitive. That's why I chose 3khz for my "peak"...it's in the range where everyone's ears but Joe's are most sensitive!)

For the last time Joe, don't misquote me! Respond to what I actually said, not what you twisted it into, or you'll get a really big SMACK!

Quote:



This makes very good sense, although if you "know" your headphones, you might simply mix so that it sounds like your headphones "should" sound.

Regarding the above, I always wondered why studios (during the mid to late 80's) used the excruciatingly bright Yamaha NS-10 monitors, and ended up with bright mixes. This does not seem to correlate to the above, and I always wondered why. I would assume that if you used a bright speaker, you should end up with dull sounding mixes (when played back over "flat" speakers).

Excellent question, Beagle (about how studios can use such bright monitors and still end up with too-bright mixes). I honestly believe that it's the result of hearing loss. Most engineers have mixed for too many decades, at too-high spls, that they simply can't hear how bright their mixes are. Sad, but I suspect probably true. Ever been inside a recording studio during mixdown? A jet plane landing 20 feet away is a whisper in comparison! (only a slight exaggeration!)

Hey Joe, thanks for that chart! Last time I took a hearing test, the hearing dude said that I don't have that dip around 4k. That's the same chart the guy showed me and then showed my hearing superimposed over it. I was trying to explain it to someone in another thread and I don't quite think it came accross.

Quote:



You are joking, right? Today's pop and rock recordings have established an all time sonic low point in fidelity. Unless all this compression and distortion is designed to mask the insipid quality of the music itself. And analog equipment has a much broader bandwidth than most studio digital gear. Compare Aerosmith's "Just Push Play" with "Rocks" or "Toys In The Attic" and ask yourself have we really gone forward or backwards in recording rock and roll in the last 25 years.

Beagle's right! Today's pop and rock recordings are HORRID in comparison to those made in decades past, with almost NO dynamic range. Not a subjective observation, but an objective one! Rip some tracks (from your cds) to your 'puter's hard drive, then look at the waveform(s) on an editor such as Cool Edit 2000. Note that the level meter doesn't really "sweep" over a broad range. Instead, it just "sits there and wiggles" (when playing a track), encompassing an average dynamic range of perhaps 5-10db. WOW, this digital gear sure is wonderful!

Actually it is (wonderful). Beagle's right that a well-tweaked open reel deck at 30ips (or even 15ips) goes octaves higher than most digital studio gear. But it (the open reel deck) also has higher distortion (although most of it of the very cool sounding, euphonically pleasing variety), wow and flutter, noise, dropouts, you name it. Good digital gear beats analog hands down in every one of these categories. And while high frequency extension may be greater with analog (than most digital), the ability to record high frequencies at HIGH LEVELS (right up to 0dbfs in digital, or tape saturation/self erasure in analog) is not among analog's strengths! Even so, a well engineered analog recording is INFINITELY more listenable than most of today's digital NIGHTMARES! The old saw "Garbage In, Garbage Out" certainly applies! Or as I've put it through the years, gear doesn't do production, PEOPLE do!

I'll take a great producer/engineer and a 25 year old analog multitrack over a "Johnny come lately" with today's latest/greatest 24/96 gear ANY TIME! Some of the all-time best, most natural sounding recordings were made in the 50s, when stereo was new, and recording equipment glowed in the dark (because of the vacuum tubes!)

Quote:



Amen, Mike....

I feel sorry for hebjam, all he wants to know is what kind of headphones you recommend, and along comes big bad mike.


Anyways, I still feel that the HD-280's are the best headphone for hebjam because they are the best headphones he could get for his budget. Here's blr's review

http://www.head-fi.org/forums/showth...d280+and+25080

I've been doing alot of looking around, and eveyone who has bought the HD-280's (recently) has said nothing but good things about them.

Oh and by the way, I can hear the 280's down to about 17hz where they audibly drop off and below that they are still able to rumble till 9hz or so. By comparison, my grado sr-60's, which have a low of 20hz cannot reproduce audible undistorted bass below 19hz or so.

Quote:



Well, he did get some answers and suggestions...

Now we're talkin' turkey. It's a learning experience and an eye opener to review these things from time to time. It amazes me how we scream for state of the art equipment to play back **** recordings. It ain't gonna help matters, we gotta get to the root of the problem first. These discussions and displays of pure passion certainly go a long way toward understanding the root of the trouble.