Need Amp/Dac for Sennheiser 6XX

[ShabbirKAN]

Hello there,

 

I know this topic might be redundant but I have been immensely confused reading different threads on Amps/Dacs. I am completely new to the whole audiophile equipment world and need some recommendations. 

 

I have the Massdrop 6XX and wanted to buy an amp/DAC to run it. I tried connecting it to my current speakers (Bose Companion 2s) and my computer but it doesn't get loud enough and the sound quality doesn't seem right. 

 

What would be a great starter amp and DAC that would come in around $300-400 CAD.

 

I have saw posts about Schitt MAGNI 3 and MODI 2. Someone also wrote about Objective 2 and STANDALONE ODAC from JLabs. 

 

This is for my desktop setup! Thanks for the help!!!

[Max_Settings]

I would go with the Audio-GD R2R-11 if you can go through the process of ordering it. If not the Schiit Magni 3 and Modi 2 Uber will work perfectly fine.

[an actual squirrel]

On dacs, I think the topping d30 is the best at this price point. It uses an xmos usb chipset, so it has lower jitter than the odac and modi 2. The odac is obsolete and has been replaced by the ol dac. The modi 2 was measured at audiosciencereview and found to have jitter and noise issues.

 

On amps, both the magni 3 and objective 2 look fine. Though I have heard anecdotally that the magni 3's noise floor goes berserk when changing volume. But the default layout for the objective 2 is horrible for cable management, and the way they change the layout is pretty janky. So, idk, pick your poison.

[Dackzy]

O2 clips like crazy especially with high impedance headphones. The O2 is probably one of the worst amps you can get. The D30 is awful, sure it measures well, but holy fuck it is a warm mess, it has so little detail. The modi 2 is extremely sensitive to bad power from a USB port, which is why I recommend the Uber version of it. For amps the only cheap option is really the magni 3, it has a low output impedance (which is good), it has a lot of power, but it is a colored amp that is over on the warm sounding side. If you want something less colored, then you can look for a used magni 2 Uber. 

You can also give audio gd nfb 11.28 a look, it is a tad more revealing than the R2R version, but they are both fairly colored sounding DAC and amp combos. Audio GD seem to have a problem with their potentiometers. I have personally seen a lot of their combos suffer from something related to the potentiometer. 

 

You could look around on the used market. A lot of people in this hobby buy most of their gear used, so no shame or anything like that in buying used.

In the past 6 months or so I have seen a lot of good deals on Meier Audio Corda gear, but I don't know if the deals are also good where you are.

[an actual squirrel]

The power curves for the o2 are available from multiple sources. On an hd 6xx, you'll run into discomfort from the volume being too high well before the amp clips. This is why power isn't too important in headphone amplifiers, and you should focus more on having low noise and low output impedance imo.

[blue2kid3]

If you wanna go with a easy amd multi option set up and maybe run some speakers down the line an AVR works great, almost endless options of hook up plenty of power and some of the best DACs on the market. I run mine via HDMI pass only audio to the AVR. This might be to much for you but is great to hook up with a record player, stream music from my pc or phone to listen to music via desktop speakers or headphones, and all the avaliable codecs.

[Spuriae]

10 hours ago, an actual squirrel said:

On dacs, I think the topping d30 is the best at this price point. It uses an xmos usb chipset, so it has lower jitter than the odac and modi 2. The odac is obsolete and has been replaced by the ol dac. The modi 2 was measured at audiosciencereview and found to have jitter and noise issues.

Why is XMOS inherently better than the CM6631A or SA9023? Hint: Amir wasn't measuring the inherent jitter of the chipset. The cheaper Behringer UMC204HD performs objectively better despite also being an XMOS USB implementation.

 

10 hours ago, an actual squirrel said:

On amps, both the magni 3 and objective 2 look fine. Though I have heard anecdotally that the magni 3's noise floor goes berserk when changing volume. But the default layout for the objective 2 is horrible for cable management, and the way they change the layout is pretty janky. So, idk, pick your poison.

Both do the same thing noise-wise. That's what cheap volume pots sometimes do, and both amps use the same Alps RK09 pot. Magni might be a bit worse while adjusting at low levels because the noise is amplified, but the O2's implementation between the stages has more noise, all of the time, at normal listening levels.

 

I do wish someone would put out a nice cheap chip amp to compete with the O2. OPA1612's and LME49600's are cheap-ish nowadays, and it wouldn't be too difficult to make something a lot better in the spirit of the O2. Not to mention the board layout could use some improvement.

 

4 hours ago, an actual squirrel said:

you should focus more on having low noise and low output impedance imo.

TBH any op-amp is going to have low noise and output impedance; that's what negative feedback does.

[Guest]

@an actual squirrel ... it was you that had your Magni make scratching noises when you quickly turn the volume knob back and forth, right? With very quiet recordings, I also hear it. My solution = play it LOUD. Seriously, I don't know why they make that quiet (yet neurotically irritating) noise (my old 1980s clock radio also did it in the end).

 

@ShabbirKAN ... welcome to the forum, eh. The DACS and headphone amps you and I buy come from the states. And we get hosed for duties, taxes, and meh-exchange rates.  It's interesting the O2 (from JDS Labs?) is problematic. I went the Schiit route; however, if I were to start again, I'd go with the JDS Labs The Element. I love its knob and the asthetics (expensive though).  You're going to love those Massdrop cans! If the clamping gets too much, GENTLY flex the steel headband. It works (I have a Charlie Brown head and finally got relief).

 

@Dackzy suggests the used headphone market. Now that I have a solid listening rig, I'm going to try that option for weird headphones.  Kijjiji for used gear (I might even try Bunz for trades).  That means road trips to Scarborough, Mississauga, and Etobicoke (crap, I might even trek out to Brampton [comic Russel Peters' stomping ground?]  or hazard a trip to Markham). 

[an actual squirrel]

7 hours ago, Nimrodor said:

Why is XMOS inherently better than the CM6631A or SA9023? Hint: Amir wasn't measuring the inherent jitter of the chipset. The cheaper Behringer UMC204HD performs objectively better despite also being an XMOS USB implementation.

 

Both do the same thing noise-wise. That's what cheap volume pots sometimes do, and both amps use the same Alps RK09 pot. Magni might be a bit worse while adjusting at low levels because the noise is amplified, but the O2's implementation between the stages has more noise, all of the time, at normal listening levels.

 

I do wish someone would put out a nice cheap chip amp to compete with the O2. OPA1612's and LME49600's are cheap-ish nowadays, and it wouldn't be too difficult to make something a lot better in the spirit of the O2. Not to mention the board layout could use some improvement.

 

TBH any op-amp is going to have low noise and output impedance; that's what negative feedback does.

I don't know how usb receivers work, but I've seen enough results from different dacs to note that the xmos ones are low in jitter. I also don't know how flash memory controllers work, but I've seen enough ssd reviews to know that samsung ones are top tier.

 

The o2 has dc blocking capacitors at input, so it's volume control tends to be better isolated from noise than others. Though I don't know how much of an issue it is in the magni 3. It may be inaudible on audiophile headphones due to their lower sensitivity. o2 in general has lower noise than the magni 3. Usually the more powerful amps are more noisy unless they are much more expensive.

 

The RME ADI-2 has a bunch of opa1688 opamps, which are like updated 4556 from the o2. It's pretty powerful and has good noise performance... though it too isn't as powerful as a magni 3 lol.

[spwath]

19 hours ago, an actual squirrel said:

On dacs, I think the topping d30 is the best at this price point. It uses an xmos usb chipset, so it has lower jitter than the odac and modi 2. The odac is obsolete and has been replaced by the ol dac. The modi 2 was measured at audiosciencereview and found to have jitter and noise issues.

 

On amps, both the magni 3 and objective 2 look fine. Though I have heard anecdotally that the magni 3's noise floor goes berserk when changing volume. But the default layout for the objective 2 is horrible for cable management, and the way they change the layout is pretty janky. So, idk, pick your poison.

Ok, amir is about the last person I trust.

Here is some things on why, and what's wrong with the d30

https://www.superbestaudiofriends.org/index.php?threads/modi-2u-and-d30-dac-deconstructing-amirs-hack-job-part-ii.6449/

[an actual squirrel]

28 minutes ago, spwath said:

Ok, amir is about the last person I trust.

Here is some things on why, and what's wrong with the d30

https://www.superbestaudiofriends.org/index.php?threads/modi-2u-and-d30-dac-deconstructing-amirs-hack-job-part-ii.6449/

The audiosciencereview measurements are accurate as far as I can tell, but I think he didn't like the conclusions, so decided to dedicate a thread to trolling.

[Spuriae]

3 hours ago, an actual squirrel said:

I don't know how usb receivers work, but I've seen enough results from different dacs to note that the xmos ones are low in jitter.

On an asynchronous implementation, the processor reclocks the signal to a local audio clock. All of the chipsets I mentioned do this. The jitter limitation is not the chips. Cheaper solutions like TI's do have measurably worse jitter.

3 hours ago, an actual squirrel said:

I also don't know how flash memory controllers work, but I've seen enough ssd reviews to know that samsung ones are top tier.

You should learn. It can be hard to defend your arguments when they aren't your own.

3 hours ago, an actual squirrel said:

The o2 has dc blocking capacitors at input, so it's volume control tends to be better isolated from noise than others.

Noise is by definition not DC. A DC blocking capacitor is intentionally selected to not attenuate within the audio band. Heck, many designs take advantage of the fact that noise goes through capacitors to use common-mode rejection.

3 hours ago, an actual squirrel said:

Usually the more powerful amps are more noisy unless they are much more expensive.

As far as signal to noise ratio goes, the opposite is often true. For actual high performance op-amp designs, constant sources like input noise and resistor Johnson noise dominate. Since noise is constant, the more powerful amps have better SNR. This is also why THD+N charts for high NFB amplifiers often have the best performance right before clipping.

3 hours ago, an actual squirrel said:

The RME ADI-2 has a bunch of opa1688 opamps, which are like updated 4556 from the o2. It's pretty powerful and has good noise performance... though it too isn't as powerful as a magni 3 lol.

The OPA1622 is the best cheap option going forward, though the price of the OPA1612+LME49600 combo has come done quite a bit. That's the best conventional negative feedback option available at the moment as far as I'm aware. OPA1688 can deliver decent current, but its 8nV/sqrt(Hz) input disqualifies it from serious low noise applications. Its biggest attraction is its price.

1 hour ago, an actual squirrel said:

The audiosciencereview measurements are accurate as far as I can tell, but I think he didn't like the conclusions, so decided to dedicate a thread to trolling.

John Atkinson, Marvey, Atomicbob, and Jude, all of whom have been doing measurements for a while now, measure the Yggy. All of them get similar results. Amir later tests it and his are significantly worse. Which one of these results is most suspicious?

 

Aside from that, Amir is just plain wrong about so many things (fundamentals like linearity, ENOB, jitter, frequency/time domain translation, as well as specifics like part names, measurement availability, layout analysis) that it's often hard to take him seriously.

[an actual squirrel]

2 hours ago, Nimrodor said:

On an asynchronous implementation, the processor reclocks the signal to a local audio clock. All of the chipsets I mentioned do this. The jitter limitation is not the chips. Cheaper solutions like TI's do have measurably worse jitter.

You should learn. It can be hard to defend your arguments when they aren't your own.

Noise is by definition not DC. A DC blocking capacitor is intentionally selected to not attenuate within the audio band. Heck, many designs take advantage of the fact that noise goes through capacitors to use common-mode rejection.

As far as signal to noise ratio goes, the opposite is often true. For actual high performance op-amp designs, constant sources like input noise and resistor Johnson noise dominate. Since noise is constant, the more powerful amps have better SNR. This is also why THD+N charts for high NFB amplifiers often have the best performance right before clipping.

The OPA1622 is the best cheap option going forward, though the price of the OPA1612+LME49600 combo has come done quite a bit. That's the best conventional negative feedback option available at the moment as far as I'm aware. OPA1688 can deliver decent current, but its 8nV/sqrt(Hz) input disqualifies it from serious low noise applications. Its biggest attraction is its price.

John Atkinson, Marvey, Atomicbob, and Jude, all of whom have been doing measurements for a while now, measure the Yggy. All of them get similar results. Amir later tests it and his are significantly worse. Which one of these results is most suspicious?

 

Aside from that, Amir is just plain wrong about so many things (fundamentals like linearity, ENOB, jitter, frequency/time domain translation, as well as specifics like part names, measurement availability, layout analysis) that it's often hard to take him seriously.

That's not how asynchronous usb works, that's asynchronous resampling. Asynchronous usb is when the dac controls when data is sent from the computer. It's kinda like vsync in gaming, where the monitor controls the frame timing from the gpu. I thought you were asking me what about the xmos processor architecture makes it so good, which I can't answer because I don't know about chip design. I think you can just infer from various results that it is a relatively good design.

 

When DC voltage goes through the volume control, which can happen when it is in motion, it modulates to AC to create audible noise.

 

I'm not talking about signal to noise ratio, I'm talking about absolute noise. 

 

John Atkinson's results were the worst. He had glitch, high distortion, and high jitter. His reviews are just always diplomatic. Amir had very high thd+n on one sample, but that same bug was on atomicbob's initial results from 2015. It seems like a quality control issue or lingering hardware bug. Other than that, the results were pretty similar, just tests run with a little bit different parameters between them and presented a little differently. The main difference is that Amir is like a tiger mom that demands perfection, while headfi and atomicbob are more in the "everyone gets a trophy" camp. They argued over whose way was the right way, but the results for all of them were mixed at best I felt.

 

Amir knows his stuff, but he's from an electrical engineering background, so I think he's not always familiar with the conventions used in audio. I don't think that important; I don't think it's that hard to understand the reviews or what the tests are. I think people are mainly threatened by him because he's popularizing judging dacs by performance, and companies have gotten lazy and don't want to compete like this. It's easier to just throw on a bunch of gimmicks.

[Spuriae]

7 hours ago, an actual squirrel said:

That's not how asynchronous usb works, that's asynchronous resampling. Asynchronous usb is when the dac controls when data is sent from the computer. It's kinda like vsync in gaming, where the monitor controls the frame timing from the gpu.

You're correct; that's how async USB works. However, async USB audio also implies timing based purely on local clocks (technically it's possible to not do that, but why?). Common adaptive USB solutions recover the clock from the USB stream and not directly from a dedicated audio oscillator.

7 hours ago, an actual squirrel said:

I thought you were asking me what about the xmos processor architecture makes it so good, which I can't answer because I don't know about chip design. I think you can just infer from various results that it is a relatively good design.

XMOS is the best, for reasons unrelated to pure USB conversion. Their wealth of documentation and cheap built-in DSP solutions have been a godsend for the small-scale audio electronics community. That they also have an excellent USB implementation is icing on the cake. If you don't care about additional processing or already have your own solution, there's no need to pay more for the extra features when a CM6631A works just as well.

8 hours ago, an actual squirrel said:

When DC voltage goes through the volume control, which can happen when it is in motion, it modulates to AC to create audible noise.

Consider that in the O2, the signal path goes gain stage –> potentiometer –> coupling capacitors. All DC current (which has incidentally also been amplified) must go through the potentiometer, as it is the only path to ground for the signal. Additionally, placement right after the gain stage means higher levels and more shot noise than the conventional input configuration.

8 hours ago, an actual squirrel said:

I'm not talking about signal to noise ratio, I'm talking about absolute noise.

Easy, just use smaller resistors. No, I'm not joking. In a good low noise design, noise is a constant dominated by the input noise of solid state devices and resistor noise. If you select for components with low current noise, this level will not change significantly with power.

8 hours ago, an actual squirrel said:

John Atkinson's results were the worst. He had glitch, high distortion, and high jitter. His reviews are just always diplomatic. Amir had very high thd+n on one sample, but that same bug was on atomicbob's initial results from 2015. It seems like a quality control issue or lingering hardware bug. Other than that, the results were pretty similar, just tests run with a little bit different parameters between them and presented a little differently. The main difference is that Amir is like a tiger mom that demands perfection, while headfi and atomicbob are more in the "everyone gets a trophy" camp. They argued over whose way was the right way, but the results for all of them were mixed at best I felt.

0-crossing glitches are the norm with R2R conversion as the MSB flips. This is nothing new. Amir is the only one surprised by its existence.

JA's highest jitter sideband was 20db lower than Amir's.

Only Amir has a 60hz dominant hum, compared to every other tester. It's worth noting that everyone else got a -130db 120hz hum despite measuring different samples with different test setups.

 

If jitter is a purely digital phenomenon, why do his jitter measurements change significantly when the output stage is changed? Food for thought.

 

This thread has some specific comparisons. Open JA's measurements and Jude's in another tab and compare.

8 hours ago, an actual squirrel said:

Amir knows his stuff, but he's from an electrical engineering background, so I think he's not always familiar with the conventions used in audio.

Bits isn't an audio-specific term. Linearity isn't an audio-specific term. Simple digital filters should be no mystery to EE's.

11 hours ago, an actual squirrel said:

I don't think that important; I don't think it's that hard to understand the reviews or what the tests are.

Apparently Amir doesn't think that's very important either, given how often he misinterprets the results.

11 hours ago, an actual squirrel said:

I think people are mainly threatened by him because he's popularizing judging dacs by performance, and companies have gotten lazy and don't want to compete like this. It's easier to just throw on a bunch of gimmicks.

He's not just a joke at head-fi and SBAF. He's a joke at hydrogen.audio too. It's a seriously impressive accomplishment to get Jude, Marvey, and everyone at hydrogen.audio to agree on something. It's not subjectivists that are unhappy with him for measuring things, it's objectivists unhappy because he's so often wrong (or more likely, intentionally misleading) that he's giving objectivity a bad name.

 

[an actual squirrel]

3 hours ago, Nimrodor said:

JA's highest jitter sideband was 20db lower than Amir's.

Only Amir has a 60hz dominant hum, compared to every other tester. It's worth noting that everyone else got a -130db 120hz hum despite measuring different samples with different test setups.

 

If jitter is a purely digital phenomenon, why do his jitter measurements change significantly when the output stage is changed?

Stereophile tested 44.1khz sampling rate, audiosciencereview tested 48khz sampling rate, and they had different volume settings. With power supply noise, one site checked it while playing a 12khz square wave, others happened to use a 1khz sine wave. Jitter is a timing problem, so different analog stuff can affect the clocks. For me, it's like if linustechtips and hardwarecanucks found different fps results for a game despite using the same graphics card. I'm not going to try to convert one of their results to the other, I just compare it to the other results with the same methodology. On stereophile, you can see the results for the bryston bda 3 and the benchmark dac 3 and see they are much better than the Yggdrasil.

 

3 hours ago, Nimrodor said:

Bits isn't an audio-specific term. Linearity isn't an audio-specific term.

Watts isn't a computer processor specific term, but Intel and AMD have different ways to define TDP.

 

3 hours ago, Nimrodor said:

He's not just a joke at head-fi and SBAF. He's a joke at hydrogen.audio too.

 

At headfi, everything is awesome and no one can do anything wrong. At sbaf, schiit is beloved. At hydrogenaudio, nothing matters, and everything (within reason) sounds about the same. That's those communities in a nutshell. asr says there are good and bad dacs and gave negative reviews to schiit. Hence why they're hated by all 3.

[Spuriae]

45 minutes ago, an actual squirrel said:

Stereophile tested 44.1khz sampling rate, audiosciencereview tested 48khz sampling rate, and they had different volume settings. With power supply noise, one site checked it while playing a 12khz square wave, others happened to use a 1khz sine wave.

Mains hum doesn't care about your test tone. If the noise isn't random, an FFT will measure it accurately. That aside, even when running the same tests or using the same setup as other testers Amir's results were significantly different; see the J-test comparison for both jitter and mains hum.

47 minutes ago, an actual squirrel said:

Jitter is a timing problem, so different analog stuff can affect the clocks.

The only thing that ostensibly changed was the output stage. How does the output stage affect the clocks?

1 hour ago, an actual squirrel said:

For me, it's like if linustechtips and hardwarecanucks found different fps results for a game despite using the same graphics card. I'm not going to try to convert one of their results to the other, I just compare it to the other results with the same methodology.

However, if one site consistently shows significantly different performance than all of the others, that's a good reason to call into question the validity of that site's results. Especially when the reviewer has a history of artificially cherry-picking for desired outcomes.

1 hour ago, an actual squirrel said:

On stereophile, you can see the results for the bryston bda 3 and the benchmark dac 3 and see they are much better than the Yggdrasil.

I'm not trying to argue that the Yggdrasil measures well, I'm trying to point out that Amir's measurements are flawed. Delta-sigma vs. R2R is a completely separate debate.

1 hour ago, an actual squirrel said:

Watts isn't a computer processor specific term, but Intel and AMD have different ways to define TDP.

Watts are a unit. Watts means watts; TDP refers to two different measurements here. Bits are a unit. Bits means bits; different measurements can produce different numbers of bits. However, Amir's "bits" are a made-up unit distinct from bits.

1 hour ago, an actual squirrel said:

At headfi, everything is awesome and no one can do anything wrong. At sbaf, schiit is beloved. At hydrogenaudio, nothing matters, and everything (within reason) sounds about the same. That's those communities in a nutshell. asr says there are good and bad dacs and gave negative reviews to schiit. Hence why they're hated by all 3.

That's a nice set of straw men you've got there.

[ShabbirKAN]

Have any of you looked into whats offered on Massdrop? 

 

Specifically the Massdrop CTH + SDAC DAC/Amp or the Massdrop Liquid Carbon X + SDAC DAC/Amp? I have been reading more and more about O2 and Schitt and it seems they aren't the best options. Realistically speaking I'm not going to crazy in terms of upgrades. I don't think ill become a crazy enthusiast. 

Also, what is there in terms of using a setup for both headphones and speakers? I currently am using the Bose Companion 2s and wanted to know if there was an under 400 option that could connect these in. This thread got very very technically which is making it hard for me to understand what to do. 

 

If yall can simplify your thoughts I would probably understand more. 

23 hours ago, PrometheanCat-1970 said:

@an actual squirrel ... it was you that had your Magni make scratching noises when you quickly turn the volume knob back and forth, right? With very quiet recordings, I also hear it. My solution = play it LOUD. Seriously, I don't know why they make that quiet (yet neurotically irritating) noise (my old 1980s clock radio also did it in the end).

 

@ShabbirKAN ... welcome to the forum, eh. The DACS and headphone amps you and I buy come from the states. And we get hosed for duties, taxes, and meh-exchange rates.  It's interesting the O2 (from JDS Labs?) is problematic. I went the Schiit route; however, if I were to start again, I'd go with the JDS Labs The Element. I love its knob and the asthetics (expensive though).  You're going to love those Massdrop cans! If the clamping gets too much, GENTLY flex the steel headband. It works (I have a Charlie Brown head and finally got relief).

 

@Dackzy suggests the used headphone market. Now that I have a solid listening rig, I'm going to try that option for weird headphones.  Kijjiji for used gear (I might even try Bunz for trades).  That means road trips to Scarborough, Mississauga, and Etobicoke (crap, I might even trek out to Brampton [comic Russel Peters' stomping ground?]  or hazard a trip to Markham). 

thanks for the advice and yeah the ordering from the US already has me stressed cuz of exchange rates.