|
|
|
On December 09 2010 20:23 MrHoon wrote:Do earphones count? (I hope so) Hey guys I want a permanent new replacement for my Iphone Earphone Yes I know Iphone earphone fucking sucks asssss and has a lifespan of 1.5 months. The only reason why I like the iphone earphones is that you're able to use the mic, which is great because I'm a lazy man  Can you guys give me a recommendation? I've tried some 3rd Party Iphone Earphones from china but the sound quality is horrendous. I just really like the whole mic+volume up/down part.
Etymotic mc3, enjoy.
|
On December 09 2010 19:44 Fyodor wrote:Show nested quote +On December 09 2010 18:49 ieatkids5 wrote:On December 09 2010 18:13 Fyodor wrote:On December 09 2010 16:31 Blackhawk13 wrote:
someone should educate us on some of this terminology like mids, highs (I assume volume? lol), amps, dacs, which headphones would need them, etc
^__v For DACs, all headphones need a good DAC. They determine the quality of the signal to be amplified so the better it is, the better the sound you get in your headphones. No they don't, only certain headphones need them. And a number of others do benefit, but need? I don't think so. And then many more don't need them at all. I don't think you have any idea what a DAC does. Headphones can't make any sound without a DAC. I'm sorry but there's no discussion to be had here, the better the signal to the headphone, the better the sound. I can't imagine a headphone so terrible that wouldn't react to a better, cleaner signal.
semantics.
You said, "For DACs, all headphones need a good DAC"
I said, "No they don't, only certain headphones need them."
Most people wouldn't be able to tell the difference between plugging your headphones into an ipod and plugging them into a good rig. Yes, a converter is needed for the ipod, but it's inferior to a better one, but do people really need the better one? Perhaps it will make a better sound, perhaps the headphone will sound cleaner. But by how much? Enough to make the average person go "Wow! I can't believe I've been listening without a dedicated DAC for so long!" ?? No. That's why many people don't need what you said they need - a good DAC. For those who can tell the difference, whether they need a good one or not depends on their headphones.
|
On December 10 2010 00:58 ieatkids5 wrote:Show nested quote +On December 09 2010 19:44 Fyodor wrote:On December 09 2010 18:49 ieatkids5 wrote:On December 09 2010 18:13 Fyodor wrote:On December 09 2010 16:31 Blackhawk13 wrote:
someone should educate us on some of this terminology like mids, highs (I assume volume? lol), amps, dacs, which headphones would need them, etc
^__v For DACs, all headphones need a good DAC. They determine the quality of the signal to be amplified so the better it is, the better the sound you get in your headphones. No they don't, only certain headphones need them. And a number of others do benefit, but need? I don't think so. And then many more don't need them at all. I don't think you have any idea what a DAC does. Headphones can't make any sound without a DAC. I'm sorry but there's no discussion to be had here, the better the signal to the headphone, the better the sound. I can't imagine a headphone so terrible that wouldn't react to a better, cleaner signal. semantics. You said, "For DACs, all headphones need a good DAC" I said, "No they don't, only certain headphones need them." Most people wouldn't be able to tell the difference between plugging your headphones into an ipod and plugging them into a good rig. Yes, a converter is needed for the ipod, but it's inferior to a better one, but do people really need the better one? Perhaps it will make a better sound, perhaps the headphone will sound cleaner. But by how much? Enough to make the average person go "Wow! I can't believe I've been listening without a dedicated DAC for so long!" ?? No. That's why many people don't need what you said they need - a good DAC. For those who can tell the difference, whether they need a good one or not depends on their headphones.
I would add that the problem with audio systems is that what you hear is pretty subjective. I might like headphones A much more than headphones B, but my friend might think otherwise. So because someone might notice the change and someone might not, the most important thing for me is to always try the system before buying it.
Don't buy headphones XXXX just because someone else likes them. Try to listen to them and then you should know more if you want to spend the money...
|
On December 10 2010 00:58 ieatkids5 wrote:Show nested quote +On December 09 2010 19:44 Fyodor wrote:On December 09 2010 18:49 ieatkids5 wrote:On December 09 2010 18:13 Fyodor wrote:On December 09 2010 16:31 Blackhawk13 wrote:
someone should educate us on some of this terminology like mids, highs (I assume volume? lol), amps, dacs, which headphones would need them, etc
^__v For DACs, all headphones need a good DAC. They determine the quality of the signal to be amplified so the better it is, the better the sound you get in your headphones. No they don't, only certain headphones need them. And a number of others do benefit, but need? I don't think so. And then many more don't need them at all. I don't think you have any idea what a DAC does. Headphones can't make any sound without a DAC. I'm sorry but there's no discussion to be had here, the better the signal to the headphone, the better the sound. I can't imagine a headphone so terrible that wouldn't react to a better, cleaner signal. semantics. You said, "For DACs, all headphones need a good DAC" I said, "No they don't, only certain headphones need them." Most people wouldn't be able to tell the difference between plugging your headphones into an ipod and plugging them into a good rig. Yes, a converter is needed for the ipod, but it's inferior to a better one, but do people really need the better one? Perhaps it will make a better sound, perhaps the headphone will sound cleaner. But by how much? Enough to make the average person go "Wow! I can't believe I've been listening without a dedicated DAC for so long!" ?? No. That's why many people don't need what you said they need - a good DAC. For those who can tell the difference, whether they need a good one or not depends on their headphones.
I have $10 Koss KSC35's and they sound astronomically better on my Audio-GD. I'm just saying the source component of an audio rig is critical to Sound Quality. It will improve any headphone.
You talked about it but what "most people need" is such a boring point of contention that I don't even feel like saying anything about it.
|
At the end of the day how we hear sound is opinion and history based. We perceive sound in a learned manner meaning we gravitate to how we previously heard a song when rating heaphones, some people also see new sound types as better just from their personality. So in a way it's personal some people like more bass some people like clearer highs and lows. The best way to achieve this is to get a great sound card and mess with the EQ not spend money on crap that you can possibly try out unless you have friends that have it and live near by. In other words buying a dedicated dac is just buying a external sound card with crap stripped away, a quality sound card would feature just as good of a signal except have more features including things like a microphone or support a 6 channel set up.
|
On December 10 2010 02:31 semantics wrote:
a quality sound card would feature just as good of a signal
Absolutely no chance of that happening. For one the power circuit is shared with the rest of the computer and is not suited for quality audio. Normal PC power supplies are built with loads of 200W+ in mind while a headphone amp requires 1 or 2 watts at best. Second, the 6+ channel DAC circuits aren't optimized for Stereo headphone use. Third, sound card makers are forcing you to buy their worthless "audio processor" chips which occupy an embarrassing amount of the total cost of R&D and manufacturing.
Sound cards are a pretty terrible choice for driving headphones, their main advantage is that they pack a lot of features for a cheap price but they're pretty bad at everything they do. Dedicated components are really the only serious options IMO.
|
On December 09 2010 18:03 Fyodor wrote:Show nested quote +On December 09 2010 10:35 Myrmidon wrote:On December 07 2010 19:49 Fyodor wrote:On December 07 2010 16:33 Myrmidon wrote:On December 07 2010 13:56 Fyodor wrote:On December 06 2010 09:22 Demi9OD wrote:On November 13 2010 12:50 Fyodor wrote:On November 13 2010 12:44 Demi9OD wrote:
Weird this thread just popped up and I saw the DT990s I bought last week on the first page. I was using audio technica ath-m50's for about two years prior, but I don't need a closed design any more (no roommate) and I found their clamp force fatiguing after a couple hours. The 32ohm version works just fine out of my Soundblaster X-Fi but this NuForce Icon uDAC-2 has been tempting me. I'd recommend the Audio-GD Sparrow. Wasn't a fan of the first uDAC. Don't think it would well with the DT990 either. I ended up getting a Asus Xonar Essence STX to replace my X-Fi Music, instead of a dedicated DAC/Amp. I really like the Dolby Headphone setting for FPS gaming and movies, and have been having fun doing op-amp swapping to find my perfect audio signature for music. So far the OPA op-amps have been my favorites, OP2137 at the moment. I have tried 49720NA, LT1358 and LT1057, and OPA2137. lol I think my Audio-GD came with an OP2137. It's pretty decent. When I finally let my discrete opamp burn in there was no turning back though. Talking about OPA Earth here. Forgive my ignorance, but what's the technical explanation behind op amp swapping? Or any rationale at all? I have a degree in EE and am working on another, but my specialty has nothing to do with electronics, so I only have limited knowledge in this area. If you look at the SNR, THD, etc. of the expensive (well, above a couple dollars a piece is expensive as far as op amps goes) op amps used for audio buffers and current drivers, they're specced way beyond the threshold of human hearing. Slew rate, max voltage swing, etc. are usually more than sufficient, and the transient response looks great too. Many of the op amps used are actually often intended for other applications that have much more stringent requirements than audio. e.g. a lot of them pass frequencies much higher than 20kHz. Granted, those specs are what the manufacturer is listing under their controlled testing environment and implementation. Maybe performance is different in a different circuit, like in a complete amplifier device? If the exact implementation is that important (I wouldn't think it would be with many classes of op amps), wouldn't the amp circuit be designed for the stock op amp? Then wouldn't the circuit be suboptimal for a different op amp you are switching to? Or maybe humans are good at hearing certain types of transient distortions that don't show up in the traditional canned response and distortion tests listed on the specs? (Another, much simpler explanation maybe would involve concepts like placebo and group think, but I'm not that quick to dismiss something I don't have much experience or knowledge of.) That's an excellent question, probably the best question asked in my thread so far. I can't tell you much about the science but opamps do sound different from one another. Generally you want an opamp with a flat frequency response from 20 Hz to 20,000 Hz, audio is extremely sensitive to that. You also take out opamps with power/bandwidth ratings that aren't suitable for audio applications and you're not left with that many choices. Chips that are left have different SNR, frequency response, input/output impedance etc and you have to find which one works for you. Opamp switching is more about altering flavor though, some opamps might be more popular in audio despite that they don't have a perfectly flat frequency response. Did my best to talk to you about opamps but I'm not an engineer, you can tell me if I'm being a fool here, I'd be glad. Well, even if the requirements do rule out 99% (arbitrary figure--I have no idea how much it really is) of op amps, 1% still represents lots and lots of op amps. Anyhow, the question is not about why certain op amps are unsuitable. Many general-purpose or specialized op amps would sound absolutely terrible, because of unacceptable distortion levels, poor characteristics close to 0V, not being able to supply enough current, not having enough voltage swing, having too low slew rate, or whatever, as you say. In any case, the input and output impedance, gain-bandwidth product, SNR, IMD, output characteristics, operating range, etc. should be sufficient for a group of op amps that we'll call "suitable for audio." The question is about why one should favor one suitable-for-audio op amp over another. If you look at the specs, these devices (e.g. LM4652) are getting THD+N at around 0.0001% with reasonable output levels and loads. If there's another op amp with a THD+N at around 0.0001% (or even at 0.005% for that matter), are you going going to be able to hear the difference between those two? It's not just THD+N that looks ludicrously good beyond the threshold of audibility, but all relevant stats that are reported. Think about it this way: let's say we have one device that reproduces a signal that is 99.999% true to the original in some sense. If you have another device that reproduces a signal that is 99.999% true to the original, why would you want the first device over the second, or the second one over the first? How do you distinguish between the two? Even if the 0.001% "inaccuracy" occurs differently in the two devices, aren't they both already so close it doesn't matter? The above is a gross oversimplification, of course--just used as an explanation. I understand what you're saying and looking at the oversimplified specs it does seem silly to compare opamps that are theoretically perfect, but you're arguing within a vacuum. If you can show me that there exists at least a few ideal opamps with flawless specs like we enumerated then we'll have made some work here. I know these are well regarded opamps for a headphone amplifier or DAC: http://focus.ti.com/lit/ds/symlink/opa627.pdfhttp://focus.ti.com/lit/ds/symlink/opa134.pdfCan't make heads or tails of these spec sheets though. Don't how perfect they are (or aren't) for audio.
A note first about most audio-DG products: you say that your Koss KSC35 sounds better using one, so a good DAC must make a big difference.
As they advertise, their DACs are supposed to color the sound, i.e. make it different from the original by intentionally adding certain distortions, maybe doing some EQ too as well. Who knows. (They do sell some DACs that are supposed to be less colored, but I don't know the specifics or the extent of what they're doing.) Therefore, it's not at all surprising that your KSC35 would sound different with them, and probably better--if that distorted sound is your preference.
It's the same idea behind tube amps. In theory, they are supposed to distort in ways that are pleasing. Mostly, it's adding extra harmonic content (2nd, 3rd, 4th, etc. above the fundamental, and things like that) to the original signal, which creates a fuller or warmer sound, to most peoples' ears. Actually, this is similar in some fashion to an organ with all its stops.
There are many different ways to distort the sound intentionally or unintentionally. When you're looking for what sounds the "best", that's hard to define since that will be different for each person. What is a good DAC? One that sounds the best, or one that produces the cleanest output, closest to the source material?
The former is hard to define, while the latter is simpler to measure. Standard audio measurements are just measures of the distortion, i.e. the difference between the output and the input. Usually these tests are done with pretty simple inputs, like a sine wave at a single frequency.
+ Show Spoiler [about distortion measures] +The total harmonic distortion is a measure of all the output signal power that occurs at frequencies other than the test tone. If you play a 60 Hz signal, output at 90 Hz, 851 Hz, etc. would be considered distortion. (However, extra power at 120 Hz would be a distortion that does not sound bad at all, while extra power at an unrelated frequency like 140 Hz would be much more grating.) Frequently, the baseline noise is considered along with THD because you can't distinguish between the two. You can think of THD and other distortions as noise anyway.
The intermodulation distortion is another measure of nonlinearity. If you input a signal with frequency content at A Hz and B Hz into a perfectly linear system, then the output would only have frequency content at A Hz and B Hz. However, real-world systems often have significant output power at A + B Hz, A - B Hz, etc. -- sums and differences of the original input signals. This is known as the intermodulation distortion (IMD).
Both THD and IMD for the op amps you showed were vanishingly small. They don't report any other distortion measures, except maybe the step response graphs, which look really good. (Look at the time scale.) We're mostly interested in performance at small closed-loop gains, like G = 1. In this case, the frequency response of the OPA134 is flat until 2 MHz, way past the audio range. Other specs are similar to this.
So I wonder benefit a discrete DAC device has over the DAC section of a sound card. The high-end DAC chips likewise have high specs for SNR, THD, etc. like the op amps we're looking at. Both the DAC section in a high-end sound card and a discrete DAC device may use the same DAC chip. The discrete DAC device may just have better EMI shielding or isolation (does this matter, unless stray radiation is actually a problem?), better power supply filtering, and a nicer-looking case. The circuits are going to be pretty much what's on a sound card. What else am I missing?
And you'll notice all these devices have a power supply rejection ratio of like 100 dB. If you've got 50 mV peak-to-peak ripple on your supply rail for your DAC or op amps on a sound card, this means something like 500 nV peak-to-peak noise induced by the power supply on a sound card, which is not only way way below the threshold of hearing, but below the other noise in the system. Do you need super-duper power supply filtering on a discrete DAC device to get 8 mV peak-to-peak ripple on your supply rail, so you can have 80 nV peak-to-peak noise from the power supply noise?
Yes, there are a few small assumptions or details I'm glossing over, which is the point. I'm interested in hearing which parts actually make a difference.
edit: I actually wrote this post before I saw your last one about sound cards vs. DACs  good timing?
edit2: Just to give a reference point for noise figures, let's assume you are listening at an output level of 1V (pretty loud--this would be at or above 100 dB SPL on most headphones). Let's assume our output listening level is 100 dB then, though it will depend on that input level and the headphones in question. 1 uV = 1000 nV of noise would give you a SNR of 10^6 = 120 dB. The noise would be 100 - 120 = -20 dB, where 0 dB is the threshold of human hearing. This is why I questioned the power supply requirements earlier. Even if you have only 80-90 dB power supply rejection ratio (PSRR), that's still plenty enough without worrying about getting the cleanest possible power supply rails. Most high-end audio op amps seem to be specified around a noise level of 1-5 uV in the audio range.
|
Does anyone have any experience with the ATH-M50's especially for EDM, house, electro and such?
|
+ Show Spoiler +On December 09 2010 18:13 Fyodor wrote:
mids and highs refer to musical note regions. Highs are typically cymbals, glass shattering, etc. While lows are things like bass drums on a drum kit or notes from a bass guitar. Headphones don't always treat those regions equally.
When it comes to amps, headphones with high power requirements will need them. All headphones benefit from the clean signal a quality amp will offer, however.
For DACs, all headphones need a good DAC. They determine the quality of the signal to be amplified so the better it is, the better the sound you get in your headphones.
Thanks that clears things up =)
|
|
|
just a heads up, the sony website has a refurbished MDR V6 for $40, which is an absolute steal. I have these and I'm sure they're alot better than most gaming headsets at that price level. And for $40 it makes it a much better deal than the sennheiser HD555, HD 280 or the audio technica cans recommended here
here
|
On December 10 2010 06:12 hifriend wrote:
Does anyone have any experience with the ATH-M50's especially for EDM, house, electro and such?
Fuck it just placed the order, $242. Hope I like em.
![[image loading]](http://blog.alltechrelated.com/wp-content/uploads/2010/08/ATH-M50.jpg)
|
wow $242? price inflation in sweden D:
|
On December 10 2010 10:25 pikaaarrr :3 wrote:
wow $242? price inflation in sweden D:
Yeah it's so much cheaper on amazon. 
|
8748 Posts
the cable on my se530 got frayed. gonna have inearz.com fix em up with a custom fit (getting mold from audiologist), custom color/design, new (better) cables. probably gonna be a while before i get the finished product but gonna be sick! IEM's (in ear monitors) are so good for LANs imo because they block out all outside sound so well!
oh yeah i wanted to ask though... at MLG's, my earphones were maybe too good. i could hear a lot of static and could even hear sounds from other computers. they have some device with a huge volume knob and connections for headphones and microphones that sits on the desk, and i think they're all connected together, so that they can get the game sounds and microphone sounds on their stream from any pc they choose. but yeah this system had same major static/interference when i used my se530. is there anything i can connect my earphones to that i can then connect to this device that will fix this problem?
|
On December 10 2010 10:49 Liquid`Tyler wrote:
the cable on my se530 got frayed. gonna have inearz.com fix em up with a custom fit (getting mold from audiologist), custom color/design, new (better) cables. probably gonna be a while before i get the finished product but gonna be sick! IEM's (in ear monitors) are so good for LANs imo because they block out all outside sound so well!
oh yeah i wanted to ask though... at MLG's, my earphones were maybe too good. i could hear a lot of static and could even hear sounds from other computers. they have some device with a huge volume knob and connections for headphones and microphones that sits on the desk, and i think they're all connected together, so that they can get the game sounds and microphone sounds on their stream from any pc they choose. but yeah this system had same major static/interference when i used my se530. is there anything i can connect my earphones to that i can then connect to this device that will fix this problem?
Yeah I think your IEM are too sensitive for that equipment. You could get an impedance adapter to help when that happens.
Either that's the problem or their gear is just bad.
|
On December 10 2010 03:55 Myrmidon wrote:Show nested quote +On December 09 2010 18:03 Fyodor wrote:On December 09 2010 10:35 Myrmidon wrote:On December 07 2010 19:49 Fyodor wrote:On December 07 2010 16:33 Myrmidon wrote:On December 07 2010 13:56 Fyodor wrote:On December 06 2010 09:22 Demi9OD wrote:On November 13 2010 12:50 Fyodor wrote:On November 13 2010 12:44 Demi9OD wrote:
Weird this thread just popped up and I saw the DT990s I bought last week on the first page. I was using audio technica ath-m50's for about two years prior, but I don't need a closed design any more (no roommate) and I found their clamp force fatiguing after a couple hours. The 32ohm version works just fine out of my Soundblaster X-Fi but this NuForce Icon uDAC-2 has been tempting me. I'd recommend the Audio-GD Sparrow. Wasn't a fan of the first uDAC. Don't think it would well with the DT990 either. I ended up getting a Asus Xonar Essence STX to replace my X-Fi Music, instead of a dedicated DAC/Amp. I really like the Dolby Headphone setting for FPS gaming and movies, and have been having fun doing op-amp swapping to find my perfect audio signature for music. So far the OPA op-amps have been my favorites, OP2137 at the moment. I have tried 49720NA, LT1358 and LT1057, and OPA2137. lol I think my Audio-GD came with an OP2137. It's pretty decent. When I finally let my discrete opamp burn in there was no turning back though. Talking about OPA Earth here. Forgive my ignorance, but what's the technical explanation behind op amp swapping? Or any rationale at all? I have a degree in EE and am working on another, but my specialty has nothing to do with electronics, so I only have limited knowledge in this area. If you look at the SNR, THD, etc. of the expensive (well, above a couple dollars a piece is expensive as far as op amps goes) op amps used for audio buffers and current drivers, they're specced way beyond the threshold of human hearing. Slew rate, max voltage swing, etc. are usually more than sufficient, and the transient response looks great too. Many of the op amps used are actually often intended for other applications that have much more stringent requirements than audio. e.g. a lot of them pass frequencies much higher than 20kHz. Granted, those specs are what the manufacturer is listing under their controlled testing environment and implementation. Maybe performance is different in a different circuit, like in a complete amplifier device? If the exact implementation is that important (I wouldn't think it would be with many classes of op amps), wouldn't the amp circuit be designed for the stock op amp? Then wouldn't the circuit be suboptimal for a different op amp you are switching to? Or maybe humans are good at hearing certain types of transient distortions that don't show up in the traditional canned response and distortion tests listed on the specs? (Another, much simpler explanation maybe would involve concepts like placebo and group think, but I'm not that quick to dismiss something I don't have much experience or knowledge of.) That's an excellent question, probably the best question asked in my thread so far. I can't tell you much about the science but opamps do sound different from one another. Generally you want an opamp with a flat frequency response from 20 Hz to 20,000 Hz, audio is extremely sensitive to that. You also take out opamps with power/bandwidth ratings that aren't suitable for audio applications and you're not left with that many choices. Chips that are left have different SNR, frequency response, input/output impedance etc and you have to find which one works for you. Opamp switching is more about altering flavor though, some opamps might be more popular in audio despite that they don't have a perfectly flat frequency response. Did my best to talk to you about opamps but I'm not an engineer, you can tell me if I'm being a fool here, I'd be glad. Well, even if the requirements do rule out 99% (arbitrary figure--I have no idea how much it really is) of op amps, 1% still represents lots and lots of op amps. Anyhow, the question is not about why certain op amps are unsuitable. Many general-purpose or specialized op amps would sound absolutely terrible, because of unacceptable distortion levels, poor characteristics close to 0V, not being able to supply enough current, not having enough voltage swing, having too low slew rate, or whatever, as you say. In any case, the input and output impedance, gain-bandwidth product, SNR, IMD, output characteristics, operating range, etc. should be sufficient for a group of op amps that we'll call "suitable for audio." The question is about why one should favor one suitable-for-audio op amp over another. If you look at the specs, these devices (e.g. LM4652) are getting THD+N at around 0.0001% with reasonable output levels and loads. If there's another op amp with a THD+N at around 0.0001% (or even at 0.005% for that matter), are you going going to be able to hear the difference between those two? It's not just THD+N that looks ludicrously good beyond the threshold of audibility, but all relevant stats that are reported. Think about it this way: let's say we have one device that reproduces a signal that is 99.999% true to the original in some sense. If you have another device that reproduces a signal that is 99.999% true to the original, why would you want the first device over the second, or the second one over the first? How do you distinguish between the two? Even if the 0.001% "inaccuracy" occurs differently in the two devices, aren't they both already so close it doesn't matter? The above is a gross oversimplification, of course--just used as an explanation. I understand what you're saying and looking at the oversimplified specs it does seem silly to compare opamps that are theoretically perfect, but you're arguing within a vacuum. If you can show me that there exists at least a few ideal opamps with flawless specs like we enumerated then we'll have made some work here. I know these are well regarded opamps for a headphone amplifier or DAC: http://focus.ti.com/lit/ds/symlink/opa627.pdfhttp://focus.ti.com/lit/ds/symlink/opa134.pdfCan't make heads or tails of these spec sheets though. Don't how perfect they are (or aren't) for audio. A note first about most audio-DG products: you say that your Koss KSC35 sounds better using one, so a good DAC must make a big difference. As they advertise, their DACs are supposed to color the sound, i.e. make it different from the original by intentionally adding certain distortions, maybe doing some EQ too as well. Who knows. (They do sell some DACs that are supposed to be less colored, but I don't know the specifics or the extent of what they're doing.) Therefore, it's not at all surprising that your KSC35 would sound different with them, and probably better--if that distorted sound is your preference. It's the same idea behind tube amps. In theory, they are supposed to distort in ways that are pleasing. Mostly, it's adding extra harmonic content (2nd, 3rd, 4th, etc. above the fundamental, and things like that) to the original signal, which creates a fuller or warmer sound, to most peoples' ears. Actually, this is similar in some fashion to an organ with all its stops. There are many different ways to distort the sound intentionally or unintentionally. When you're looking for what sounds the "best", that's hard to define since that will be different for each person. What is a good DAC? One that sounds the best, or one that produces the cleanest output, closest to the source material? The former is hard to define, while the latter is simpler to measure. Standard audio measurements are just measures of the distortion, i.e. the difference between the output and the input. Usually these tests are done with pretty simple inputs, like a sine wave at a single frequency. + Show Spoiler [about distortion measures] +The total harmonic distortion is a measure of all the output signal power that occurs at frequencies other than the test tone. If you play a 60 Hz signal, output at 90 Hz, 851 Hz, etc. would be considered distortion. (However, extra power at 120 Hz would be a distortion that does not sound bad at all, while extra power at an unrelated frequency like 140 Hz would be much more grating.) Frequently, the baseline noise is considered along with THD because you can't distinguish between the two. You can think of THD and other distortions as noise anyway.
The intermodulation distortion is another measure of nonlinearity. If you input a signal with frequency content at A Hz and B Hz into a perfectly linear system, then the output would only have frequency content at A Hz and B Hz. However, real-world systems often have significant output power at A + B Hz, A - B Hz, etc. -- sums and differences of the original input signals. This is known as the intermodulation distortion (IMD).
Both THD and IMD for the op amps you showed were vanishingly small. They don't report any other distortion measures, except maybe the step response graphs, which look really good. (Look at the time scale.) We're mostly interested in performance at small closed-loop gains, like G = 1. In this case, the frequency response of the OPA134 is flat until 2 MHz, way past the audio range. Other specs are similar to this. So I wonder benefit a discrete DAC device has over the DAC section of a sound card. The high-end DAC chips likewise have high specs for SNR, THD, etc. like the op amps we're looking at. Both the DAC section in a high-end sound card and a discrete DAC device may use the same DAC chip. The discrete DAC device may just have better EMI shielding or isolation (does this matter, unless stray radiation is actually a problem?), better power supply filtering, and a nicer-looking case. The circuits are going to be pretty much what's on a sound card. What else am I missing? And you'll notice all these devices have a power supply rejection ratio of like 100 dB. If you've got 50 mV peak-to-peak ripple on your supply rail for your DAC or op amps on a sound card, this means something like 500 nV peak-to-peak noise induced by the power supply on a sound card, which is not only way way below the threshold of hearing, but below the other noise in the system. Do you need super-duper power supply filtering on a discrete DAC device to get 8 mV peak-to-peak ripple on your supply rail, so you can have 80 nV peak-to-peak noise from the power supply noise? Yes, there are a few small assumptions or details I'm glossing over, which is the point. I'm interested in hearing which parts actually make a difference. edit: I actually wrote this post before I saw your last one about sound cards vs. DACs good timing? edit2: Just to give a reference point for noise figures, let's assume you are listening at an output level of 1V (pretty loud--this would be at or above 100 dB SPL on most headphones). Let's assume our output listening level is 100 dB then, though it will depend on that input level and the headphones in question. 1 uV = 1000 nV of noise would give you a SNR of 10^6 = 120 dB. The noise would be 100 - 120 = -20 dB, where 0 dB is the threshold of human hearing. This is why I questioned the power supply requirements earlier. Even if you have only 80-90 dB power supply rejection ratio (PSRR), that's still plenty enough without worrying about getting the cleanest possible power supply rails. Most high-end audio op amps seem to be specified around a noise level of 1-5 uV in the audio range.
About Audio-GD, they are not reputed for colored equipment. Only some models offer intentional coloration and they are clearly marked as so. Kingwa always pushes the neutral parts over the other ones. My particular gear is known for being neutral.
About distortions that are "below the threshold of hearing". Perhaps it is justified to discard these but I've been thinking that we have talked about those in a vacuum. When thinking about an audio system, small distortions can add up when you think about all the different stages the signal goes through.I can't possibly begin to do the science on this though.
Maybe you think I'm being irrational because I don't like to think scientifically about audio gear. I'm a big computer geek too and specs matter a whole lot in that domain. When I got into audio, I threw that love of science out of the window pretty fast. When I buy gear I rarely keep it for long, I sell, trade, buy on a monthly basis. I always much prefer to listen as final judgment and it's treated me very well so far. I'm a musician too so I know how things are supposed to sound. If the science says that a few jellybean chips assorted by people who don't care strapped onto a computer power supply represent audio perfection well I might as well be religious about audio because that's not what I'm hearing.
I'd love to meet an actual audio engineer who has built headphone amps and DACs for a living. He could tell me what's what in audio but that hasn't happened to me so far. Although you are being helpful so far and I'm thankful for that.
|
Can someone tell me what headphone the guys in the GSL are using? They look like Sonys. Does anyone know for sure? Really appreciate it!
|
On December 10 2010 19:39 zak wrote:
Can someone tell me what headphone the guys in the GSL are using? They look like Sonys. Does anyone know for sure? Really appreciate it!
Sony MDR-XB700 I believe
![[image loading]](http://ecx.images-amazon.com/images/I/31geMTHwttL._SL500_AA300_.jpg)
|
|
|
|
|
|
EPT
![[image loading]](http://www.shure.com/idc/groups/public/documents/webcontent/prod_img_srh840_l.jpg)
