Hey folks,

Pending some business dealings, I'm probably going to fit myself up with a pair of HD600's (or maybe even 650's) pretty soon.

So, not knowing much about headphones until recently (and still not knowing much now), I understand you need an amp to power the HD600's (at least, to get the proper sound out of them, which I'd want to do).

How bulky are these things, and what's a good one for under 100 US (if one exists)?

Regards,
- Spike

Tough to do for under $100, unless you can find something used. I have seen Creek OBH-11 go for around $100 on Audiogon. Have you thought about building one? The Pimetata is a good option under $100

IMO basically all higher-end headphones need amplification, so you are on the right, albeit expensive, path.

Your price range very nicely solves the bulkiness issue since for under $100 all headphone amplifiers are positively tiny. Little Dot Micro+ is slightly smaller than a deck of cards!

For that price I'd say the most formidable amplifiers are Little Dot Micro+ (available on eBay) and PA2V2. The next price/quality mark after that is about $300 with great products such as HeadRoom Micro Amp and the Corda amplifiers by Meier Audio. See the Head-Fi Forums for updated discussion and reviews of headphone amplifiers.

I got the Behringer MiniAmp AMP800 (79.99 AUD) for my HD600 and I'm satisfied with it. The cMoy I had for my portable device was too weak for the HD600 imho.

A very good amp that mates well with the HD600s is the Rega Ear. In fact that is the exact combination that I have. It brings out the best from the laid back Senns. Check out head-fi for more info.

Nevermind, I forgot I was posting in Hydrogenaudio forums. I'm sorry

The most important thing for a headphone amplifier is the gain, and the background noise.

Then comes the frequency response under real-life load (not just resistive). This is given by its intrinsic frequency response and by its damping factor vs frequency.
Then the phase response under real-life load.
These data are interesting for power amplifiers, but I don't know if they are relevant with headphone amps.

Then intermodulation distortion according to the frequency, and harmonic distortion.

In my opinion, if you don't have at least the first data, you may as well flip a coin to choose your amplifier.
Unfortunately, these data are never published in a standardized way, and nearly never published at all.

Well said!!!

May I add..... buyer beware!!!

Headphone amplifier channel balance for low-level listening can be a major issue too, IMHO. Noise can be an issue both in terms of noise floor due to too high gain or just poor s/n ratio, and in terms of ground hum noise. Noise tends to be more of an issue with very sensitive low impedence phones like the MDR-V6s, especially if the gain is too high. An amp's ability to work around difficult ground-loop hum issues can be quite important, IMHO. These issues are what separate the good stuff from the just okay stuff, in my experience. A near-0 ohm impedance headphone amp will usually have the least impact on frequency response, though some phones are designed to be driven from an industry-standard 120-ohm jack. Some smaller battery-powered amps will have trouble driving very low-impedence phones without some small amount of bass rolloff (the ipod photo headphone out had this problem, as I understand it).

The biggest problem with high-impedence, low sensitivty phones like the HD600 and HD650 is usually not noise or channel balance at low levels but just getting them loud enough without pushing the amp's performance limits too far (resulting in clipping or distortion), and this is usually only an issue with battery-powered portable equipment. Most things that you plug in a wall will drive them very comfortably.

I use a Behringer UB802 mixer. The input gain is adjustable on two channels if you are a perfectionist. Or you can just use other channels with fixed gain. It drives my Senn HD580s and Beyer DT880s very easily, and my Sony MDR-V6s very quietly. My confidence in my UB802 increased when a friend took measurements on its big brother, the UB1202, and it did quite well for noise, frequency response, and distortion. Though I am not the expert on this stuff. The UB802 mixer also has bass, mid, and treble controls, and multiple inputs and outputs. And panning controls that you can use as a primitive (but continuously variable) crossfeed for old recordings with excessive or tacky stereo separation. And pretty flashing lights. Price: $60.

Expensive headamps edge on over into snake-oil-ville as price increases, IMHO. In many circumstances, a home receiver or home CD deck (with headphone jack) or computer sound card will provide solid amplification even with the HD600s or HD650s, IMHO. As to whether one could hear a difference versus an expensive headphone amp in an ABX test, I make no claims. I can tell you based on experience that any difference is not worth it to me, though. A $100 portable amp could be genuinely useful if you are going to use such hard-to-drive headphones as the HD600 or HD650 portably. Thing is, you could get a second pair of really nice and more practical headphones for portable use for the price of that portable amp.

If someone is thinking of paying $200 or $300 or more for the meager amount of amplification needed to drive a pair of headphones, I suggest taking the time to come to a better understanding of what you are paying for. If someone at hydrogen audio asserts you need to pay that much for transparent headphone amplification, I say prove it. What is it -- T.O.S. 8?

They are quite relevant for headphone amps. People often measure frequency response with particular headphones attached and the graphs are often very non-linear indeed for consumer hardware.

You can look into some threads at http://www.head-fi.org.

I own both the HD600 and HD650. You don't need an amp to power them. I do have a Grace Design headphone amp but before that, I just plugged the headphones straigt into a CD player jack, soundcard headphone jack, or into my laptop's headphone jack.

I do notice that you used the phrase "get the proper sound out of them..." ... if "proper" means loud, then yes, you'll need a headphone amp to drive them to loud levels. If "proper" means a pleasant sound at decent normal listening levels, then a special headphone amp is not required.

Unfortunately just 'loud' doesn't quite cut it. The problem with computer/other line-outs is that they generally have high-impedance output and expect a very high-impedance load, such as 30,000Ω amplifier input. When someone plugs a 60Ω headphone into a line-out expecting 30,000Ω load, the frequency response characteristics of the line-out may change considerably. Probably the single most common phenomenon is attenuation of lower frequencies since they usually have the lowest impedance in most headphones. So you can achieve a sound that is 'loud' but still lacks a major proportion of the spectrum. Also when volume (load) is yet increased with such an underpowered source, pretty much anything can happen to the frequency response curve, but it almost certainly won't be flat. And at this point we've drifted a long way away from 'hi-fi'.

So basically a headphone amp's most useful purpose is to perform impedance matching between line-out and a pair of headphones. If you already have a low-impedance output expecting a low-impedance load (such as high quality headphone output of a mixer table), then you shouldn't have major problems driving headphones and don't need a headphone amp.

I often listen to my HD600s plugged straight into my MAudio Transit (USB soundcard). While the headphones can't be driven to "loud" levels I find the playback very pleasurable indeed and have no problem with the bass response (unlike listening to my ER4Ps plugged into the Transit, that's pretty awful). I'd say try and live with the new headphones for a week or two and then make the final decision afterwards.

Wow. Consider yourself lucky. My setup is M-Audio Transit USB -> Little Dot Micro+ -> Sennheiser HD595. My experience was that this setup is completely unusable without an amp. To the extent that listening to any music actually hurt my ears since the treble was so loud and coarse. I still think that even with amp the Transit is slightly treble-y, but it's OK since it complements the famous Sennheiser Laid-back Sound pretty nicely.

That's very interesting.



This is pretty much how I find the Transit with my ER4Ps, this combination is nasty for me. My HD600s are nothing like this with the Transit, I find the set-up fairly neutral.

Here are the frequency responses of my amplifier and soundcard headphone outputs, with and without the Senheiser HD600 plugged into them :



Ampli : Arcam Diva A85
Soundcard : Marian Marc 2

I have the same amplifier and use my HD600s mostly with that via an Audiophile 2496 soundcard. The experience is just as nice as it is with the HD600s plugged straight into the Transit.

All applications I throw at that image complain that it's broken, even after reloading the page
I tried Firefox, Gnome Imageviewer and The GIMP... for some reason I have the same problem with a lot of png's produced by RMAA. Don't know what's wrong with the file...
Still I'd be interested to see it. Would you perhaps test/republish the file?

You can view the image with Openoffice Draw.

Thanks for the tip! It got me a little bit further: it would actually open the file. But... it shows me the sort of noise that you see on a TV that's not connected to an antenna signal. It's a bit green, too. That is, greenish random bits, no graph to be seen. All this even after the third download of the file (I cleared my browser cache inbetween).

Internet Explorer 6 works OK here.

He-he, that's funny! OOo 2.0.2 Draw opens just fine, but here is a working copy of Pio2001's incompatible original.

Interesting results!

THANK YOU for showing (by example) how to get this thread out of the land of fantasy and back to planet earth. Some earlier posts regarding subjective impressions were really bugging me and I thought were way outside the boundaries of TOS 8 and most likely totally unsupportable by objective measurement. But I thought I'd just let it go, being a nobody and a novice and all, and never having owned the sound card in question. It's really nice to see a "super moderator" setting a good example.

I have a UB802 and it does measure very well. I also tried using it as a headphone amp too but I didn't like it at all. I tried it with a Grado SR-80 and a Koss KSC-75 too but I found it to be quite thin and sterile sounding. It just wasn't coming out to me.

Why do you think the UB802 would sound "thin and sterile" even though it measures very well?

[edited slightly for clarity]

Many thanks for that, this one I can actually view. And, glad to find it's interesting to see too
(oh and now I can see this was also generated by rmaa - like all the other stuff I can't view... at first I was wondering why Digit-Life had removed all their measurement content.... lol)

Now, you said that your UB802 measures very well, but it sounds "thin and sterile" to you when you use it to drive your headphones -- your Koss KSC75s and Grado SR80s. So the obvious inference to me is that you have neutral amplification but you are not pleased with your headphones. So I tried to offer you some help, having some experience with headphones.

"Thin and sterile" is not much use to anybody.

Here, a refresher course for you, regarding TOS 8:

Hydrogenaudio is supposed to be an objectively minded community that relies on double-blind testing and relevant methods of comparison in discussion about sound quality. The usual "audiophile" speak of non-audio related terms which are completely subjective and open to redefinition on a whim, are useless for any sort of progression in discussion.

This rule is the very core of Hydrogenaudio, so it is very important that you follow it.

Here is a discussion explaining why

http://www.hydrogenaudio.org/forums/index....showtopic=11442


You can read how to easily perform double blind listening tests here :

http://www.hydrogenaudio.org/forums/index....showtopic=16295

I must add some important precisions to my measurment :

The amplifier used is a power amplifier for speakers with headphone output. A dedicated headphone amplifier should perform better.

I could not ABX the signal recorded with the green frequency response from the one recorded with the pink one. I tried to ABX an artificial bass boost with a similar shape, but since it is quite narrow, I fail when the difference is less than +/- 1 dB. But I didn't try with a pink noise.

if your strongest argument (since you keep repeating it) is that it should sound well simply because it measured well then sadly you've got it all wrong.

I have never heard of anything that measures well and sounds bad in a blind listening test.
Things that measure well and sound bad in casual listening test do so only because of psychological factors. People imagine them to sound bad, then the brain creates the illusion of a bad sound, that is really heard by the listeners, but that is generated, against their will, in their head, and not by the device listened to.

Sound quality must be evaluated by the way of blind listening test only. So if you think that a headphone amplifier that measures well regarding all problems relevant for an amplifier, that is frequency response, phase response, background noise, and distortion, can sound bad, try a blind comparison.

Not very easy, but possible. You need a jack splitter, a jack adapter in order to redirect the splitter output into the line input of your soundcard. You also need a software capable of measuring level.
Plug your headphones into the first amplifier, and set the volume to the desired listening level. The volume won't be changed after this setting.
Then, if you can, generate a sinewave at -6 dB with a software and feed the headphone amplifier with it. If not possible, play a CD. Then unplug your headphones, plug the splitter instead. In the first output of the splitter, plug your headphones, and plug the second output into your soundcard. Note the playback level. If you are using a CD, record a given part, select extremely carefully a precise part, and ask your software for its RMS level.
Then do the same with the other amplifier, until the recorded level is less than 0.1 dB away from the one of the other amplifier.

Once the voulmes equalized, remove the splitter, and get someone to help you blind testing. You can ABX your two amplifiers this way. Make sure that your headphone cable is fixed to somlthing in the middle, to avoid feeling it pulling your ear when it is plugged into an amplifier or the other, and make sure that your friend plugs the splitter into the other amplifier, so that you always hear one plug entering both amplifiers whatever the random choice.
Check also that your headphones don't click when they are plugged in. It may help you identify the amplifier without listening to it.

Let your friend use a coin or a dice to choose the amplifier. Perform some training.
When you are ready, try to identify the amplifier 10 times in a row. Be sure to distinguish training and testing. A mistake during testing is a mistake and can't be corrected. You can test more than 10 times, but don't adjust the total to suit your score. If you choose 10 trials and begin testing, stick with a total of 10.

Never plug speakers output into headphones or into line input It can destroy the amplifier !

Performing this test, you should be able to identify amplifiers whose frequency response is affected by the headphone impedance. Use pink noise for this. It's the easiest signal to recognize. You can also recognize the background noise of the amplifiers, and maybe the harmonic distortion if they are overdriven.

But you should not be able to recognize the thinness or sterileness of the sound if they are not related to a frequency response change. These "non measurable feelings" do not come from the signal delivered by the amplifier. It's all in our heads,and people sensitive to this can't control it. The proof is in the blind test, when, as should be the case, you attribute the characteristics of one amplifier to the other by mistake, because you are not aware of which is which.

If you nonetheless pass the test with a perfect score (10/10), it would be a very interesting result.

Indeed it would! Thanks a lot Pio2001 (and others). I love the idea of ABX's, as I wrote in my first post on these forums a few weeks ago.

Now. On the RMAA (or is it now RMA?) image, what are the x and y axis? I'm bad at interpreting the images, but I'm fairly knowledgeable about what the stats themselves mean. A basic itnerpretation would be helpful.

Also Pio, if you could give a comparison of what the amp and sound card you've used compare to, it'd help as well, because I've never heard of the amp, which isn't surprising since I know ntohing about them- but I've also never heard of the sound card.

If anyone could tell me if they've ABX'ed amps it'd be lovely. I certainly don't want basically guesses and inferences and random claims (IMO, head-fi.org) about the quality of amps. If I wanted that, I'd visit a manufacturer's website

Finally, I love the great discussion and VERY helpful wisdom that's been imparted not just on me, but all-round. I feel like I can make a much more informed decision, if not a better one in general.

Regards,
- Spike

I use a Behringer UB1202 mini-mixer board for headphone amplification. I find it is extremely versatile, in that it can amplify(without clipping) the AKG K340(very inefficient, voltage wise) on very dynamic opera/classical with zero clipping. [UB1202 uses a +18/-18 power supply] It can also amplify the MDR-CD3000(one of the most efficient headphones, low impedance) without any audible noise floor(assuming you set the gains correctly --- this is a mixer with multiple gain settings). The unit measures respectably; demonstrating no measured parameter would be suspect of being truely audible[as opposed to biased sighted listening where other factors come into play], and it's very affordable. I could not ask for better audible performance, as it just seems to do the job[amplify headphones] without problem(s). I have so far used the unit with the AKG K501, AKG K340, Beyer DT880, Koss Pro 4AA, Sony MDR-V6/MDR-7506 and Sony MDR-CD3000 and have not noticed any problem(s).

-Chris

Chris - good point! I too am very happy with the sound and flexibility of the pair of Behringer Euroracks that I use for the audio work that I do (primarily live-DAT post-processing), with both powered studio monitors as well as headphones (MDR-7506s, usually).

-brendan

Could you explain how do you measure with your headphones plugged in?

See Pio2001's post above.

Thanks. Just overflew that post and must have missed it.

Are the worse ratings when headphones are plugged in due to the headphones per se or just because the soundcard/amp output isn't supposed to drive both headphones and line-in at the same time?

The x axis is the frequency, from the lowest audible frequancy (bass) on the left, to the highest (treble) on the right. The y axis is the level of the signal. All frequencies should be at the same level. We can see here on the green curve that some low frequencies are 0.7 dB too loud. It means that the ampli output too much bass. However, 0.7 dB is small for an interval that goes from 50 to 200 Hz only.



The amplifier is an audiophile stereo integrated amplifier worth 1200 €, with a headphone output in front of it.
The sound card is a semi-pro 24 bits 96 kHz soundcard using one of the first 24/96 chip available for the mass marker. It was issued at the same time as the M-Audio Audiophile 2496, the Terratex EWX, the Echo Mia and the Hoontech DSP2496. It should be similar to them.



Here are some successful ABX tests :
http://www.provide.net/~djcarlst/abx_pwr.htm
http://www.soundandvisionmag.com/assets/do...erInterface.pdf

Why not? "Bad" doesn't have a defined value, so everything that one doesn't like may be counted as bad. Including flat frequency response.

BTW, that image opens fine in ACDSee 3.1. It's a bit strange that some more recent software (including Opera 9 latest build, which I use) can't show it right.

In order to clarifiy, I have never heard of something A that sounds different from something B in a blind listening test if the relevant measurments show no difference bigger than the threshold of hearing.

Interesting. I can't say I didn't expect a (much) larger difference between line-out and amplifier, but one can't argue with those results. I cheerfully retract all of my statements regarding the necessity of a headphone amplifier, but will remain a user of one nonetheless.

It would be interesting to perform the test with a considerably worse sound card (I'm thinking of motherboard/integrated here) to see if there's any significant difference between 'driving ability' of different line-outs.

Yes, very good point Nik. But, for audiophiles like us we wouldn't be using a motherboard sound (hopefully anyway!).

I'd be even more curious as to the latency incurred when NOT using an amp than how well the amp drives the headphones (i.e. does it run fine, BUT, use PC power to the point that audio latency occurs?). In fact, that question would be interesting to ask with regards to Pio's example. Does using the sound card to power them mean higher latency or other side effects? Headphones are but one partof the audio setup, of course



So, I take it the RMA graph Pio posted shows differences betwee nte 4 settings, but as far as differences between amp/phones and sound card/phones, the actual difference, although relevant, is minimal.

- Spike

Don't forget that the headphones tested are 300 Ohms. With lower impedance headphones, the problems are bigger.

I think the HD600s have an unusually large impedence hump centered at about 100 hertz (up to about 550 ohms) that accounts for the small (often inaudible) amplifier frequency response rise in the bass, particularly as amp output impedence rises well above zero:

http://graphs.headphone.com/graphCompare.p...pare+Headphones

For a near-0 ohm headphone out (maybe your sound card?), the effect of this impedence hump should be very small, I think.

Notice the much more even impedence curve of the AKG 501s (see the link above), just to pick another headphone out of the air. Would this not result in an even flatter amplifier response and be less amp dependant, even though the headphone impedence is less? I know that WmAx knows how to computer-model this stuff.

Very low-impedence (24 ohm or less) phones may cause the low bass to be a little rolled off with amps with inadequate output capacitors (mainly, some portables), if my sketchy memory and understanding are correct.

Here are some RMAA comparison results for my Audiophile2496 in loopback mode and the output from my Transit recorded through the Audiophile.

http://daefeatures.co.uk/tests/Audiophile2496Transit.htm

I'll keep my eyes open for a splitter so I can do further tests in the future. It would be interesting to see how my soundcard and amplifier combination compares with Pio2001s since I have the same headphones and amp.

I have the HD580, and I've tried it out on the Echo Indigo laptop soundcard and the HeadRoom Micro DAC and Micro Amp. Whether I use lossless or mp3, there is absolutely no difference in sound quality compared to the headphone jack of my Dell Inspiron 700m. Either my laptop is equipped with one of the best sound processing chips every made, or all the audiophile products are just a bunch of nonsense. Don't waste your money on an amp. Anyone who believes they need one is just kidding themselves, and just wants to spend a lot of money of expensive toys to inflate their ego.

But even considering the previous discussion there still is a legimate use for a headphone amplifier -- volume. If you have a very high-impedance, insensitive (as in low efficiency, nothing emotional) pair of headphones combined to a very low-power output, you might not get enough volume for comfortable listening. At this point, if you're not willing to change the source, a headphone amplifier is the only option. However, most line outputs I've seen seem to be able to drive headphones at very loud volumes.

BTW, did you get the $600 HeadRoom combo before or after you reached this conclusion?

This problem occurs mostly with classical recordings with very quet parts.

I got the HeadRoom Stack before reaching this conclusion. I had a lot of faith in the things the guys at Head-Fi were saying. But when I actually put things together just like they said, I didn't get anything better out of it. They're all nuts, if you ask me.

This is one way, but other ways are also informative. E.g. measuring and just listening.

Just measuring and listening can lead to false conclusions; I.E.; the measurements can influence one psychologically. Without specific perceptual research to coorelate with the measurements[or control trials of specific phenomena as they occur specifically, such as simulating and ABXing for oneself], they are of very limited use. Most(even those that design this stuff) do not refer to perceptual research or carry out control listening trials, though such is possible.

-Chris