or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › DT880 600ohm BS
New Posts  All Forums:Forum Nav:

# DT880 600ohm BS - Page 9

Quote:
Originally Posted by xnor

230V here, but prolly still not enough to make them really shine. This little beast needs a nuclear reactor!

Alright, enough fun me thinks. ;)

Pah, you're all a bunch of amateurs who clearly don't know what they're talking about. For true audio fidelity, you need to start with a synchrotron...

though from a pseudo-science view, it does make more sense than a nuclear reactor

Thanks to the help of several head-fiers such as xnor and odigg, you can actually calculate the FQ deviations with a higher Zout if you have the impedance curve data available.  Here is the information for the HD650 - xnor, feel free to correct me if I'm wrong.  The HD650 does have a humpy impedance curve and this should be interesting.

1 Volt applied to 300ohms with a 15ohm output (damping factor of 20) produces:

dB = 20*log (v1/v2)

Voltage 1 = baseline impedance of HD650

1V/315 = 0.00317A

0.00317A * 300ohm = 0.9524V

Voltage 2 = impedance peak of 475ohm @100hz

1V/490 = 0.00204A

0.00204A * 475ohm = 0.9694V

dB = 20*log (0.9694/0.9524)

dB = +0.15dB @ 100hz

Therefore, with a damping factor of 20, you have an entirely inaudible FQ increase of 0.15db @ 100hz.  Let's see what happens when you decrease the damping factor to 1 (300ohm Zout)

Voltage 1

1V/600 = 0.0017A

0.0017*300ohm = 0.5V

Voltage 2

1V/775 = 0.00129A

0.00129*475 = 0.6129V

dB = 20*log (0.6129/0.5)

dB = +1.8dB @ 100hz

Therefore with a damping factor of 1, you have an entirely audible 1.8dB boost @ 100hz.  Very audible, but given the wonky FQ response of most headphones anyway (+/- 5dB), it won't ruin your day.  I won't lie, lower Zouts produce a flatter FQ response.

Quote:
Originally Posted by Shike

Quote:

Originally Posted by maverickronin

So you don't really need more than a damping factor of 1 for a typical headphone then?

Depends what you consider typical.  If you look at the impedance graph at headroom if it's a flat line it's considered resistive and higher output impedance will just make them less efficient without showing huge FR changes.  Headphones like the Sennheiser HD650 which varies greatly based on frequency on the other hand will deed a damping factor of roughly 1/20th or better at said frequency to minimize or mitigate impact of it.

Quote:
Originally Posted by Armaegis

Pah, you're all a bunch of amateurs who clearly don't know what they're talking about. For true audio fidelity, you need to start with a synchrotron...

though from a pseudo-science view, it does make more sense than a nuclear reactor

OMG - I can't believe I forgot about the synchotron.  You're right - you'd absolutely need that.

I recently tried running my DT880/250Ω directly out of an iPod. At lower volumes, the sound was quite muffled. But once I cranked it up past a certain point, the clarity started coming back.

Now correct me if I am wrong on this. It is my understanding that the voltage supplied from an amp will affect how loud the headphones will sound. And the current supplied will affect the clarity / distortion, and pretty much all other aspects besides volume.

Supposing that's true, it seems that on lesser amps (such as the iPod built in one), adequate current is not supplied until the volume gets cranked up pretty loud. This could be why many people prefer listening at loud volumes. And on a 600Ω headphone, the volume pot must be cranked up higher than with a 250Ω one. So, my informal hypothesis is that this means the 600Ω headphones will have more current available than a 250Ω one at the same volume level. And the increased current makes it sound better. But on very good amps, there is already lots of current available even at the lower range of the volume pot. In such cases the 250Ω and 600Ω would have less difference.

How's that?

whats with updating the whole OP?

No joke its uncool.

Threads are not tech-papers, they are conversations. It is supremely uncool to edit info out of old posts even if to add new info elsewhere.

Quote:
Originally Posted by wavoman

(Of course ABX has the huge advantage that you can do it without a partner; but if you have a helpful partner, there are better ways IMO).

QFT.

Quote:
Originally Posted by maverickronin

Actually I just thought of another reason why the 600s would sound better.  FWIR most of the people on here who recommend the 600s over the 250s seem to listen to them on OTL tube amps.  Since such amps typically have quite high output impedance, a higher impedance driver will provide a higher damping factor.

Is a higher damping factor better for a Beyer?

Quote:
Originally Posted by Yoga Flame

I recently tried running my DT880/250Ω directly out of an iPod. At lower volumes, the sound was quite muffled. But once I cranked it up past a certain point, the clarity started coming back.

Most pdap's (which indcludes the iPod) do the volume control in the digital domain. Digital domain volume controls work by skimming bits off of the signal. Once you drop below a certain threshold of bits (in a multibit system...) digital is simply worthless.

Measure the THD of a 1Khz sine wave recorded 70db below peak in 16/44 :) After you do this try to tell me that digital is great....

What you describe with current output and whatnot isnt correct for why pDAP's suck: its the volume control method.

RockBox has a Dither setting - one would assume it applies to volume changes - latter iTunes dithers

also many DACs are "24 bit" delta-sigma type and could play undithered, volume reduced 16 bit audio without "bit loss" for 48 dB of attenuation

the player's output could be measured with better soundcards - or cheap motherboard soundcard with quiet amplification to determine if simple truncation @16 bits is used

Edited by jcx - 9/2/10 at 8:17pm
Quote:
Originally Posted by jcx

RockBox has a Dither setting - one would assume it applies to volume changes - latter iTunes dithers

also many DACs are "24 bit" delta-sigma type and could play undithered, volume reduced 16 bit audio without "bit loss" for 48 dB of attenuation

the player's output could be measured with better soundcards - or cheap motherboard soundcard with quiet amplification to determine if simple truncation @16 bits is used

Rockbox's dither setting is well explained in the Rockbox manual being that all it really does is noise shape.

"bit loss" or lack of dither is not imo the 1st most likely explanation of "muffled sound at low volume" - simple Fletcher Munson Loudness effect shows that perceived frequency response is absolute SPL dependent - low level listening will seem to lack both high and low frequencies without "smile" eq

esp when talking about 96 dB/mW 250 Ohm cans and ~ 1 Vrms output portable players

sensitive iem are another story - attenuation is really desirable between iem and any source

the RockBox dither setting is explained as changing from the default truncation behavior to noise shaped TPF dither

what is not explicitly stated is where in the signal processing chain - the proper location for dither is after all processing including eq and volume - if all audio signal processing is done with higher res internal representation - say 24 or 32 bits

adding dither immediately after the codecs which can output up to 29 bit would be wrong if further processing happens before output to DAC

another poor option would be to use dither repeatedly between processing blocks (eq, volume) if each bolck output only 16 bits despite higher wordlength internal math

Edited by jcx - 9/3/10 at 6:51am

Quote:
Originally Posted by jcx

"bit loss" or lack of dither is not imo the 1st most likely explanation of "muffled sound at low volume" - simple Fletcher Munson Loudness effect shows that perceived frequency response is absolute SPL dependent - low level listening will seem to lack both high and low frequencies without "smile" eq

Although there are ways around the problems associated with high attenuations in the digital domain, I dont think that anyone uses them in a portable device. I reserve the right to be wrong, and some days I think I own the patent on being wrong, but I dont think today is one of those.

Anyways, whether the poster above is hearing something related to the Fletcher Munson curves &nonlinearities of the ear or particularly bad effects of high digital attenuation could be tested very easily.

Compare the following:

A: Headphone out of portable player set for high attenuation into an input of an amplifier (or buffer) bypassing the volume control. This isolates the headphone output opamp from the load, but maintains the sound of the digital control.

VS

B: The line out of the same portable player into another input of the same amp (or buffer) but through an analog volume control.

Volume match, and ABX.

This arguably leaves the possibility that the opamp in the portable player's headphone amp is affecting the sound, BUT a quick glance at the datasheet for whatever part is used will no doubt show outstanding performance into a several-Kohm load.

I agree that most IEMs need attenuation after the output of pretty much anything. I built this a few years ago as a fun little project:

I would think compared to expanding compressed audio formats and multiband eq the extra processing cost of properly applying dither should be small - that doesn't mean anyone's doing it, or doing it right in the consumer DAP world

direct measurement of dap output with a good quality 24 bit soundcard should easily capture the correlated quantization distortion of 16 bit truncation/rounding in the dap if it is there - I need to bring my Juli@ back home and put my test software back together

Edited by jcx - 9/3/10 at 6:28pm

Quote:
Originally Posted by nikongod

I built this a few years ago as a fun little project:

Wow. That's certainly more than meets the eye

What's it do exactly? Transformer based pre-amp?

Quote:
Originally Posted by Yoga Flame

Wow. That's certainly more than meets the eye

What's it do exactly? Transformer based pre-amp?

1kohm ct:8ohm transformer to provide a ~20db stepdown to IEM's

The switch between the jacks selects between 1K or 500ohm primary.

The switch on the side is for design obfuscation to get less prejudiced responses when I can get people to listen to it with their IEMs.

Quote:
Originally Posted by nikongod

1kohm ct:8ohm transformer to provide a ~20db stepdown to IEM's

The switch between the jacks selects between 1K or 500ohm primary.

The switch on the side is for design obfuscation to get less prejudiced responses when I can get people to listen to it with their IEMs.

Very nice. So with this even as the voltage/volume gets attenuated, the amount of current available is not lost. I wonder if something like that would make a DT880/250Ω sound like a DT880/600Ω.

Now about digital attenuation. I have no trouble accepting that an iPod or laptop soundcard discards digital bits to lower the output volume. But what about say a FiiO E1 or E5? They use digital volume control buttons, but they must be operating in analog because their input is the iPod analog line out. So if a \$20 E5 can do that, maybe an iPod does something similar too?

Quote:
Originally Posted by Yoga Flame

Now about digital attenuation. I have no trouble accepting that an iPod or laptop soundcard discards digital bits to lower the output volume. But what about say a FiiO E1 or E5? They use digital volume control buttons, but they must be operating in analog because their input is the iPod analog line out. So if a \$20 E5 can do that, maybe an iPod does something similar too?

Indeed, there are several digitally controlled analog attenuator. The problem for getting one of them into a DAP is that its another chip and another 3cm^2 of board space which is more than enough reason for the marketing department to complain about "no longer having the smallest thing" and the accounting department to complain...

With digital attenuation everything can be done with chips you need for basic functionality. The accountants are happy and the marketing department is happy.

However of course the problem of digital attenuation is that the noise level stays the same. There's an article from an electronic engineering magazine that goes into this and why analog is preferred sound quality wise that I'll find later.

New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: Sound Science